Summary
The right to access information requires high quality oversight. Studies of effectiveness of Access to Information (ATI) legislation tells a clear story: the benefits of greater transparency and access to information can only be realised when this system is actively enforced. To be effective, the whole system of ATI review and appeal has to be designed as a system of cultural change. The system has to use limited resources in a strategic way to reform cultures of unnecessary secrecy in government that protect corruption and inefficiency in public life.
Building on a comprehensive picture of appeal systems and processes across Europe, this report argues for the value of specialised oversight bodies (Information Commissioners), who have independence from government and the power to compel compliance from authorities. In countries that use a system of internal review, better monitoring and interventions are necessary to ensure this system enhances rather than detracts from access to information.
Summary of recommendations:
- Better investment in the resources, capacities and independence of Information Commissioners improves the quality of the ATI regime, attacks corruption, and strengthens good governance.
- Specialist Information commissioners are preferable to general ombudsman, bringing more specific knowledge, and are a more suitable structure to shepherd the access to information regime.
- The power to enforce decisions is a required tool for driving culture change in public authorities.
- Systems of internal review should be replaced by commissioner-led systems of appeal, where information commissioners have understanding of appeals across the entire system, and can use internal review as a strategic choice, rather than a hurdle before an appeal can be considered.
- In general, oversight bodies and civil society rarely have high quality information about full workings of the ATI system. We argue that better quality statistics are a valuable tool in demonstrating the value of the system, and in allowing targeted focus on problems.
Fundamentally, good ATI regimes are important because of the effects they have in society, strengthening anti-corruption and good policy-making approaches. Better oversight is a cost-effective way of unlocking these wider benefits. This report explores how technical details of how the oversight system works are important in achieving these overall objectives.
Introduction
The majority of countries across Europe have laws that give individuals the right to access information held by public authorities. Each of these laws - referred to as Access to Information (ATI), Freedom of Information (FOI), or Right to Information (RTI) laws - operate slightly differently. There are differences in the scope of authorities covered, the forms of information available, and how the laws are regulated. Access to Information laws are part of the framework of institutions, customs and norms that shape how effective the implementation legislation can be.
This report focuses on the appeal and oversight process. A decision by a public authority to deny access to information is not always the end of the story. If the decision is challenged, the local system might mean it is reconsidered by a higher authority, judicially reviewed through the courts, overruled by an Information Commissioner, or criticised by an ombudsman. The effectiveness of the local system of oversight is a key factor in the effectiveness of the overall ATI regime.
Comparable institutions and situations in different countries help provide evidence for the effectiveness of different approaches. While being sensitive to local circumstances and institutions, well funded, independent Information Commissioners, with the power to make binding decisions, are a key part of making ATI systems effective. Examining the problems of systems of internal review of complaints, this report explores a commissioner-led system of appeal, where internal review is one of several possible strategies allowed by the oversight body. Alongside this, gathering statistics is an essential but often absent part of overseeing an ATI regime that enables better oversight and more targeted and effective interventions.
To get to this point, we have developed some new labels and terminology to explain different approaches taken in different jurisdictions. While the process of accessing ATI from a citizen point of view is similar across different European countries, how that system interfaces with deeper systems of administrative process and accountability vary, and fit into a number of different patterns. This report first explores the general context of access to information in Europe, and the evidence base for the importance of the oversight system. It will then look at different forms of oversight, and processes of appeal, before turning to arguments about which approaches are more effective.
This factual content of this report is a synthesis and condensing of previously published work and data. The main information source on different methods of appeal was the 2018 book "The Laws of Transparency in Action: A European Perspective"1, which contains detailed descriptions of ATI systems across Europe, including the appeal structures. This information has been supplemented with information from additional publications, including some written by those running Alaveteli sites; and background interviews with European transparency campaigners.
Notes ():
1: Dragoș, D. C., Kovač, P., & Marseille, A. T. (Eds.). (2018). The Laws of Transparency in Action : a European perspective. Palgrave Macmillan.
This report also makes key use of statistics published by Access Info/Centre for Law and Democracy's RTI Rating to explore comparisons between European countries specifically on oversight and appeal processes.
The report was written by Alex Parsons. Project scoping and background interviews additionally involved Rebecca Rumbul and Jen Bramley. The report was edited by Myf Nixon. We are grateful to the feedback of Helen Darbishire on an earlier version of this report.
Access to Information in Europe
Access to Information (ATI) is a term for the right to access government information. This concept is commonly referred to as Freedom of Information in English-speaking jurisdictions; the preferred term and acronym in international contexts tends to be Access to Information (ATI) and this term will be used in this report. In Sweden, the concept dates back centuries, but most European systems have been formalised (and sometimes revised several times) in the last fifty years.2
Notes ():
2: Worthy, Ben, Freedom of Information in Europe: Creation, Context and Conflict (June 24, 2019). Contested Trade-Offs: Transparency and Secrecy in European Democracies. Forthcoming , Available at SSRN: https://ssrn.com/abstract=3409066 or http://dx.doi.org/10.2139/ssrn.3409066; Dragoș, D. C., Kovač, P., & Marseille, A. T. (2018). The laws of transparency in action : a European perspective.
ATI legislation defines the scope of the openness regime. This will either be through an explicit list of public authorities that the law applies to, or requirements that describe the bodies that the law applies to. ATI may result from a single law, or a collection of laws and regulations. A law may say that only certain kinds of information should be released, or reverse this and say that only certain kinds of information should be withheld. Special rules may apply to specific kinds of information (such as information about the environment), or to specific kinds of requesters (such as special regimes that require faster response times for journalists). An ATI law might be prescriptive (for example it may state that requests are only valid if submitted on paper in writing) or unconcerned about the form a request for information should take (accepting requests by Twitter or phone). Some jurisdictions may require a fee for access to information, or for practical costs such as copies. ATI laws can also apply to supranational bodies (such as the EU) and cross-border institutions (such as shared cross-border bodies in Ireland).
An ‘ATI regime' includes both the laws that create rights, and the set of institutions and practices that make those rights practically accessible. ATI regimes are often inspired by each other and share common features, but they are also a product of the specific local context of their creation. A new ATI regime can represent an iteration on older laws, a solidification of existing custom, or be a deliberate break with the past. A legislative change may be the end result of a long-running legislative process, demands from civil society, implemented by a new government based on promises made when out of government, or required by a new constitution. While presenting a similar face to requesters, ATI regimes have a local character, as they interact with existing legal and cultural attitudes around official secrecy.
Federal or devolved systems can have different sets of ATI laws and institutions that apply depending on the jurisdiction. For instance, Scotland (UK) and Flanders (Belgium) have different ATI rules and institutions from the federal/national governments. In Germany, 14 of the 16 Länder have ATI legislation, and this legislation is not uniform. As a result, the law and regulation on Access to Information may not be uniform across a country. Making ATI more accessible means being sensitive to the local conditions and where the levers for change are. Examples of ‘best' practice may not be easily applicable in all jurisdictions, but greater knowledge of similar regimes is helpful in providing more information for local actors to understand the current situation, and argue for change.
Oversight is vital to successful Access to Information
A key idea behind ATI laws is that too much secrecy in government leads to bad government, and bad government makes the lives of citizens worse. Exposing more government information and decision making makes it more likely that poor or corrupt processes or officials will be detected, and in the longer run, increase the effectiveness of government by driving out poor or corrupt practices. Access to Information is part of a framework of open government policies designed to achieve this goal.
A focus on anti-corruption and poor performance also makes it clear that there will be public officials with a very strong incentive to resist ATI requests (because it would expose their corrupt behaviour or poor performance). With this in mind, enforcement of ATI laws is important for the simple reason that officials are more likely to comply with the law if there are effective penalties for not doing so.
Those who believe in the value of ATI need to be concerned with oversight because of a strong evidence base that ATI's benefits are greater (or even only detectable) when there is effective enforcement of the law. In a 2021 meta-analysis of 56 studies, Can Chen and Sukuma Ganapati3 found that transparency tools had a detectable effect at reducing corruption, but that this effect was much stronger in conjunction with fiscal transparency and e-transparency. In a 2016 study, Krishna Vadlamannati and Arusha Cooray4 used panel data from 132 countries and found that ATI laws in general were associated with improved bureaucratic efficiency, and this was even higher in countries with ‘strong' ATI laws. In this case, 'strong ATI' meant that it scored highly on the Access Info/Centre for Law and Democracy RTI Rating, which includes measures of the strength and powers of the oversight body. In a 2014 study, Adriana Cordis and Patrick Warren5 examined different ATI regimes across the United States, and found both that switching from a weak to a strong FOI law immediately increased conviction rates for corruption, and that this rate declined in the long run. They argue this showed an initial increased detection of corruption, followed by a reduction in actual corruption. The most significant component of their measurement of "weak" and "strong" was a change in the civil and criminal penalties for violating the ATI law. This validates a relatively common sense view of ATI laws: if they have strong sanctions they are more effective.
Notes ():
3: Chen, C., & Ganapati, S. (2021). Do transparency mechanisms reduce government corruption? A meta-analysis. International Review of Administrative Sciences. https://doi.org/10.1177/00208523211033236
4: Vadlamannati, K. C., & Cooray, A. (2016). Do freedom of information laws improve bureaucratic efficiency? An empirical investigation. Oxford Economic Papers, 68(4), 968–993. https://doi.org/10.1093/oep/gpw008
5: Cordis, A. S., & Warren, P. L. (2014). Sunshine as disinfectant: The effect of state Freedom of Information Act laws on public corruption. Journal of Public Economics, 115(March), 18–36. https://doi.org/10.1016/j.jpubeco.2014.03.010
The strongest validation of the importance of oversight bodies comes from a 2022 study looking at aid transparency. Dan Honig, Ranjjit Lall and Bradley Parks6 used transparency regimes implemented by funders of aid projects as a way of testing whether the features of those regimes had an impact on the success of the final project. Looking across 20,000 aid projects supported by 12 donor agencies across 183 countries, their main conclusion was that the oversight and enforcement mechanism was very important, and that "the adoption of ATI policies by agencies is associated with better project outcomes when these policies include independent appeal processes for denied information requests but no improvement when they don't". This dataset presents a rare opportunity to understand the impact of effective transparency not just on general measures of perceived corruption, but on the good running and success of projects. Honig and colleagues suggest that this effect happens through a "shadow of the future" mechanism (supported by interviews), where the possibility of future disclosure means that more work is done earlier to avoid design and implementation problems, leading to better results. This study provides key support for the value of ATI policies while providing a substantial caution: in the absence of a real enforcement mechanism, they make no difference.
Notes ():
6: Honig, D., Lall, R., & Parks, B. C. (2022). When Does Transparency Improve Performance? Evidence from 23,000 Projects in 148 Countries. American Journal Of Political Science, 00(0), 1–21. https://doi.org/10.1111/ajps.12698
The recurring theme of this evidence is that ATI laws are not a magical fix, but as part of a wider framework can drive progress. ATI cannot work alone, and the enforcement and oversight of the process is vitally important if it is to achieve its goals. Getting the oversight right has potential to be very good value for money, unlocking reduced corruption and improved projects throughout the supervised system. Within this question, there is a lot of debate about the exact form this oversight and appeal process should take.
Different forms of oversight and appeal
A review process is most effective at ensuring the law is followed where it is cheap, fast and accessible. Appeals processes allow requesters to challenge the decisions of public authorities, but there are several different forms of appeal process, and they vary in the effectiveness and cost to the requester.
Despite facing the same problem, very few countries in Europe have exactly the same form of appeal process or oversight body. As ATI oversight interacts with the existing legal and administrative structure and processes, there is always a local character to how Access to Information is regulated. That said, there are variations on common approaches that can be bought together to build a better sense of what is effective across multiple systems.
A common part of the appeal process is some form of administrative review of the decision. Laws that specify an external appeal process can require that administrative review is attempted first. In other cases non-binding mediation or advice from an Information Commissioner can play a role in the internal administrative review. Access Info/Centre for Law and Democracy's RTI Rating shows that 62% of European countries covered by the rating have some kind of free access to administrative review.7
Notes ():
7: Analysis of data downloaded for European countries from rti-rating.org.
In Europe, there are three main types of administrative review: internal review,supervisory review, and assisted internal review.
In an internal review, the original authority has a process for reconsidering their opinion after a complaint. This is likely to be formalised and escalated to a higher official inside the same organisation. In the UK and Ireland this is called an 'internal review', in the EU a ‘confirmatory application' and Sweden 'an appealable decision'.
In a supervisory review, a superior or supervisory authority to the original public authority evaluates the original decision. This generally connects to an existing hierarchy of public authorities. For instance, the appeal process for Hungary8 and Czechia9 refers appellants to existing systems of supervisory bodies, while Denmark's first point of appeal is the boards of appeal system that is the first step for general administrative redress in the Danish system.10
Notes ():
8: Petra Lea Láncos (2018), Freedom of Information in Hungary: A Shifting Landscape : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
9: Stanislav Kadečka, Jan Brož, and Lukáš Rothanz (2018), The Laws of Transparency in Action: Freedom of Information in the Czech Republic : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
10: Pernille Boye Koch, Rikke Gottrup, and Michael Gøtze (2018), Special Report: Transparency on a Bumpy Road— Denmark : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
In an assisted internal review, the original authority conducts an internal review in light of an opinion provided by the oversight body. For instance, in France11 and at the federal level in Belgium12, the Information Commissioner (CADA) will consider a complaint, and provide their opinion to the original authority. The authority may or may not change their mind on the basis of this advice, and the decision of the oversight-body is non-binding (if potentially informative about the prospects of a future legal challenge).
Notes ():
11: Yseult Marique and Emmanuel Slautsky (2018), Freedom of Information in France: Law and Practice : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
12: Stef Keunen and Steven Van Garsse (2018), Access to Information in Belgium : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
Some systems do not have any form of internal review: for instance in both Slovenia and Croatia a requester can appeal directly to the Information Commissioner, while appeals in Ukraine go directly to the Administrative Court.
There are two broad questions that shape the oversight regime:
- Does the success of ATI need a body with a specialised remit, or can it be covered by more general bodies?
- Does the success of ATI require a body that can enforce its judgements?
These questions lead to four types of oversight bodies on a 2 x 2 grid : Courts, Ombudsman Institutions13, Mediator Information Commissioners and Regulator Information Commissioners (Table 1). These are broad categories where many examples do not exactly fit. For instance, Regulator Information Commissioners may still spend a lot of their time mediating, and not all ombudsman institutions lack enforcement powers.
Notes ():
13: There's no settled gender neutral pluralisation of an Ombudsman. The original Swedish term (ombudsmän) is gender-neutral, but generally it is not incorrect to apply English rules and assumptions once something has become an English word (ombudsmen, or ombudspersons). Alternative approaches include shorting to ‘Ombuds'. The International Ombudsman Institute sidesteps the problem by referring to multiple ‘ombudsman institutions', which also reflects that the idea is often localised using a term that is more descriptive in the local language (public advocate, etc).
| Mediator/Advisor | Enforcer | |
|---|---|---|
Specific | Mediator Information Commissioner | Regulator Information Commissioner |
General | Ombudsman | Court system |
Table 1 - Oversight body types.
All ATI regimes have some role for the court system, but some countries use it as the main means of appeal. In these situations, Access to Information is just one case among many where a public authority can make an unlawful decision, and the court system can be used to question this and invalidate the original decision. Depending on the country, this might be through a special administrative tribunal or a civil court system. The powers of the court to enforce an action may vary and the effect may be to require the authority to reconsider their original decision to be compatible with the ruling rather than the court directly releasing the information. Access to this appeal system depends on general access to justice issues, such as court fees. This is the default appeal mechanism.
Court systems can differ in how they approach judicial review. In some cases they can issue an order to release the information, in other cases (such as Belgium's Council of State14) they can invalidate the refusal, but leave the possibility that the authority may craft another, valid refusal.
Notes ():
14: Stef Keunen and Steven Van Garsse (2018), Access to Information in Belgium : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
Ombudsman is a general term for an office that helps the public deal with grievances about how public authorities or bureaucracies have behaved. Generally, the ombudsman has a mediation and advice role to help bring a resolution to a dispute, but without the power to compel an outcome. Exactly how this interacts with the court appeal system depends on the country. In the absence of a specialised ombudsman (usually an ‘Information Commissioner'), an ombudsman with responsibility for investigating wider maladministration (actions of a government body that cause harm or injustice) may be able to investigate ATI complaints. Access Info/Centre for Law and Democracy's RTI Rating shows that 79% of European ATI regimes have some right of appeal to an independent oversight body.15
Notes ():
15: Analysis of data downloaded for European countries from rti-rating.org.
There are common features to many Ombudsman institutions, but the generic term hides a large amount of differences between countries. The spread and success of the Ombudsman model is an example of what Jamie Peck and Nik Theodore16 call "fast policy". This is where the search for "ideas that work" can lead to ideas spreading between very different jurisdictions. However, as these new institutions need to take root in very different systems, this process is best described as translation rather than transfer. Generally policy that successfully spreads between different countries both solves a clear problem and is adaptable to different national contexts, and the founder institution can be notably different to the generalised understanding. While the ombudsman concept originates in Sweden, the version that spread around the world is based on the Danish model, which has a more generalised goal of reducing "maladministration", but has reduced powers.17 This "classic" model ombudsman typically only has the ability to investigate and recommend actions. The Swedish Ombudsman by contrast has some powers to bring prosecutions (if not for ATI issues in the modern era). Gabriele Kucsko-Stadlmayer shows that in some new democracies the ombudsman has legal powers that echo back to the Swedish approach approach.18 This report uses Ombudsman in the generic "classic" sense in this report, but individual institutions will vary significantly from this.
Notes ():
16: Peck, J., & Theodore, N. (2015). Fast Policy: Experimental Statecraft at the Thresholds of Neoliberalism. Minneapolis: Univ Of Minnesota Press.
17: Kucsko-Stadlmayer, G. (2009). The Spread of the Ombudsman Idea in Europe. Retrieved from http://www.theioi.org/downloads/34chi/Stockholm Conference_25. Back to the Roots_Gabriele%25Kucsko Stadlmayer.pdf
18: Kucsko-Stadlmayer, G. (2009). The Spread of the Ombudsman Idea in Europe. Retrieved from http://www.theioi.org/downloads/34chi/Stockholm Conference_25. Back to the Roots_Gabriele%25Kucsko Stadlmayer.pdf
An Information Commissioner can sometimes be a specialised ombudsman, or more like a specialised court. This name for the office follows the Canadian version, which was established in 1973. Like 'ombudsman', the term is broad and is applied to bodies with very different powers and functions. In addition to mediator or regulator powers, they may operate as a softer promotional power, creating guidance and advice for authorities and the public, or in promoting legal rights to the public more broadly or producing a systematic report of usage of information rights. The Access Info/Centre for Law and Democracy's RTI Rating shows that 54% of European ATI regimes have an oversight body with investigative powers, 46% can directly order remedies (release of information), 38% can issue binding decisions, and 36% have the ability to impose structural measures to increase compliance.19
Notes ():
19: Analysis of data downloaded for European countries from rti-rating.org.
The most significant difference between the more powerful commissioners (who can make binding decisions) and appeals to the courts is cost, both to make the appeal and in some cases the risk of a costs award. An appeal to an Information Commissioner is free in almost all European Jurisdictions (Ireland being the exception20), and they are highly specialised in information law. This makes the role quasi-judicial and establishes a body of jurisprudence that is subject to review by the courts, but also that shapes information release systems more widely than the initial case.
Notes ():
20: Depending on the nature of the request and the appeal: Fees For A Review | OIC
The range of possible roles of Commissioner makes multiple ways of categorising them valid. While Table 1 divides them into two sets, in a 2016 paper, Victoria Lemieux and Stephanie Trapnell21 divided commissioners into three roles, not two. They separate out mediation, adjudication and binding adjudication as separate roles a Commissioner may play. An argument against making too firm a split into different groups is that even commissioners who can make binding decisions may spend most of their time mediating. All these types of specialised commissions will also manage what Lemieux and Trapnell call monitoring, the "management of nationwide implementation and guidance on the design of policies". Commissioners' roles, while varied, tend to be concerned with the whole lifecycle of the ATI process.
Notes ():
21: Lemieux, V. L., & Trapnell, S. E. (2016). Public Access to Information for Development: A Guide to the Effective Implementation of Right to Information Laws. https://doi.org/10.1596/978-1-4648-0879-1, p. 55
Common features of ATI oversight systems
Access Info Europe and the Centre for Law and Democracy maintain and publish an RTI (Right To Information) Rating that compares Access to Information laws around the world. This rating includes a number of indicators that relate to the appeal process. This information can be used to understand how common different features of an appeal process are across Europe, and to find countries that are similar in general terms, if not in exact institutions.
Limiting the RTI Ratings to 14 indicators that are related to the review or oversight process and 39 European countries in the index, Chart 6 shows how many countries either fully or partially passed an indicator test. This shows that most European countries have some sort of judicial appeal for FOI decisions (82%)22, and many similarly allow an appeal to an oversight body (79%).
Notes ():
22: It is likely that the true number is higher, as several comments for 0 scoring countries indicated a lack of direct mention of judicial appeal in the law, when this may simply be implied by the wider structure.
It is much rarer that the oversight body can order remedies (49%) or that their decisions are binding (38%).
Chart 6 - RTI Ratings indicators for appeal/oversight process in European jurisdictions
| Group | Indicator | Count | % |
|---|---|---|---|
Review process | Judicial appeal | 32 | 82.05 |
Review process | Appeal to oversight body | 31 | 79.49 |
Review process | Non-outcome appeals (late, administrative silence, etc) | 30 | 76.92 |
Review process | External appeal is free | 25 | 64.1 |
Review process | Free internal review | 24 | 61.54 |
Politically independent oversight | Good appointment process and tenure protection | 22 | 56.41 |
Politically independent oversight | Oversight body has good financial independence | 21 | 53.85 |
Oversight body powers | Investigative powers | 21 | 53.85 |
Statistics | Public bodies must publish annual statistics | 19 | 48.72 |
Oversight body powers | Can order remedies | 18 | 46.15 |
Statistics | Report presented to legislature | 18 | 46.15 |
Politically independent oversight | Prohibitions on political connections | 15 | 38.46 |
Oversight body powers | Binding decisions | 15 | 38.46 |
Oversight body powers | Oversight body can impose structural measures | 14 | 35.9 |
Table 2 - Counts and percentages of European countries that pass RTI Index indicator tests.
Common processes
Despite the lack of a standard model for ATI appeal and oversight processes, there are some common patterns that can highlight similarities or contrasts between countries. The following section explores some common features and differences found between European appeal processes. In charts showing general escalation processes, there may be additional legal appeal options beyond these points.
A common pattern in appeal processes is an escalation chain of appeals to successively more senior authorities. This typically starts with an administrative review, before moving to an Information Commissioner or court, before further appeals in the court system.
While systems in UK, Scotland, and Ireland use these steps23, both Scotland and Ireland only allow an appeal of the commissioners decision on a point of law, while in the UK allows an appeal to the Information Tribunal that may introduce new arguments or evidence. The 2016 Burns review recommended aligning the UK with the Scottish/Irish approach).24
Notes ():
23: Courts Service (Ireland) (2020), How to Appeal an FOI Decision, https://www.courts.ie/how-appeal-foi-decision
24: Independent Commission on Freedom of Information. (2016). Independent Commission on Freedom of Information Report. Retrieved from https://www.gov.uk/government/publications/independent-commission-on-freedom-of-information-report
Escalation chains do not need to contain all these elements. Slovenia25 and Croatia26 both allow an appeal to the Information Commissioner without a form of administrative review. Ukraine does not have any specialist infrastructure for ATI, and allows appeals directly to the courts without an administrative review.27
Notes ():
25: Polonca Kovac (2018), Slovenia on the Path to Proactive Transparency : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
26: Anamarija Musa (2018),Croatia: The Transparency Landscape : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
27: Доступ до правди (2022), About/Complaints, https://dostup.pravda.com.ua/help/about#complaint
Some forms of assisted internal review (Germany or Belgium at the federal level) are not escalation chains because they require simultaneous routes of appeal to be taken).In France, the steps proceed in sequence and all steps must be taken, but the case is not escalated to successively more senior authorities. Instead, the Information Commissioner (CADA) is appealed to give evidence to an authority, who can use this information in reconsidering their original decision.28
Notes ():
28: Yseult Marique and Emmanuel Slautsky (2018), Freedom of Information in France: Law and Practice : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
Some ATI appeal processes allow complaints to be made through one of several different routes. In some cases a reason to choose to appeal to the Ombudsman rather than the court might be cost. The Ombudsman offers a free way of getting an expert and independent opinion on the information request. This reduces the information gap between requester and authority, and may provide a quicker and easier way to acquire at least some of the requested information. Depending on the standing of the Ombudsman in that area, this might lead to reconsideration of the appeal, or the authority may stick with the original decision despite the new opinion.
Chart 1 - Different escalation chain approaches in UK, Scotland, Ireland, France
Chart 1 - Different escalation chain approaches in Ukraine, Netherlands, Slovenia and Croatia
Chart 2 - Escalation processes where there is a choice of options
In some contexts different appeal routes seem to exist for different kinds of problems. Both Denmark and Sweden have Ombudsman roles that take an interest in Access to Information matters, but their role fits into the overall system in different ways. In both cases complainants have the option of complaining to the ombudsman or the court system. In Sweden the court is the primary route, while in Denmark it is much more rarely used, and generally only by journalist requesters.29 This is because in Denmark, the Ombudsman is more explicitly involved in the escalation chain and sometimes has opinions that information should be released. Their annual reports highlight several cases over the last few years where they had a different view to the original authority.30 These decisions are non-binding, but are respected and can lead to the original case being reopened and the information released. In Sweden, the Ombudsman tends to contain themselves to issues of process, with decisions highlighted in annual reports criticising authorities for slow responses or data destruction.31 Despite a similar mix of institutions, the way these institutions interact with requesters can be different in different countries.
Notes ():
29: Pernille Boye Koch, Rikke Gottrup, and Michael Gøtze (2018), Special Report: Transparency on a Bumpy Road— Denmark : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
30: See English language reports of the Danish Ombudsman, 2013, 2016 and 2018 and search "was entitled to access": https://en.ombudsmanden.dk/publications/annual_reports/
31: See annual reports of the Swedish Ombudsman https://www.jo.se/en/About-JO/Annual-reports/
Chart 3 - Swedish and Danish escalation routes
In some countries, different devolved or federal structures have their own Access To Information laws. In Germany, 14 of the 16 Länder have their own ATI laws with different oversight offices (a joint position that also deals with data protection).32 Transparenzranking.de has ratings of the different Länder-level regimes, which includes information about the different oversight bodies. None of these differ significantly from the federal Information Commissioner (for instance, none of these have sanctioning power), but do vary in their political independence/funding and reporting requirements.
Notes ():
32: Christoph Emanuel Mueller, Bettina Engewald, and Marius Herr (2018), Freedom of Information in Germany : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
Sometimes the processes at a devolved/regional level are easier for the requester to access than those at the national/federal level. While similar in structure, the ATI law and oversight body in Scotland has a number of differences with the UK-wide ATI law, with clearer approaches to administrative silence and time scales for internal review.33 In Belgium, while the federal system has a variation of an assisted internal review where the requester needs to appeal to the original body and Information Commissioner at once, the Flemish region has a separate, and much simpler system, where the Information Commissioner can issue binding orders to release information.34
Notes ():
33: Rumbul, R. and Parsons, A. (2021). Reforming FOI. [online] mySociety Research. Available at: https://research.mysociety.org/publications/reforming-foi [Accessed 1 Mar 2022].
34: Stef Keunen and Steven Van Garsse (2018), Access to Information in Belgium : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
Chart 4 - Difference in escalation approach at the Federal level (Belgium/Germany) and in the Flemish region in Belgium.
Some countries have different systems of appeal for different kinds of information. For instance, Belgium has a separate Appellate Body of the Access to Environmental Information that handles environmental information requests.35 Several countries (Serbia, Romania, Czechia) have different information regimes available for journalists/mass media which allow expedited Access to Information. This access is generally dealt with through the same court system as ATI requests. Italy, because of the confusing way its three ATI laws interact, has a separate appeal process for requests made under its 1990 law.36 This 1990 law is the only one where there is a specific Information Commissioner available to appeal to, but it also restricts the ability to ask for information to those who have a demonstrable interest in the information. This means the more confusing appeal route (which may still be cheaper) is only available in a narrow set of circumstances.
Notes ():
35: Stef Keunen and Steven Van Garsse (2018), Access to Information in Belgium : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
36: Paola Savona and Anna Simonati (2018), Transparency in Action in Italy: The Triple Right of Access and Its Complicated Life : in The laws of transparency in action : a European perspective. Palgrave Macmillan., background interview B
Chart 5 - Italy ATI's escalation routes
The case for stronger Information Commissioners
A key distinction between different kinds of oversight bodies is the mediator/enforcer division. The enforcer model has the ability to make binding judgements and either compel new decisions, or compel information to be released, if they disagree with the argument of a public authority that it should be withheld. This model can work similarly to a specialist court or arbiter, and effectively generates case law based on previous decisions. The mediator model has no such binding powers, and can either play a role as a neutral party in mediating the dispute, or provide their view of the law to an authority to take into account in making their decision.
A key reason for preferring an enforcer model, as covered in a previous section, is good evidence that the end benefits of an ATI regime are strongly dependent on the strength of a right of appeal.37 Actual sanctions for non-compliance score well in a final goal of reduced corruption.38 Against this argument is the idea that mediator oversight can be effective, and achieves much of the same effect collaboratively. While mediator commissioners often have some option of intervening in court proceedings, they are not required to justify their non-binding decisions on review. This means there are lower legal costs to running a mediator commissioner. The argument that mediator commissioners are a viable approach deserves examination as it would mean there is less of a need for change in many existing regimes.
Notes ():
37: Honig, D., Lall, R., & Parks, B. C. (2022). When Does Transparency Improve Performance? Evidence from 23,000 Projects in 148 Countries. American Journal Of Political Science, 00(0), 1–21. https://doi.org/10.1111/ajps.12698
38: Cordis, A. S., & Warren, P. L. (2014). Sunshine as disinfectant: The effect of state Freedom of Information Act laws on public corruption. Journal of Public Economics, 115(March), 18–36. https://doi.org/10.1016/j.jpubeco.2014.03.010
In a paper titled "More Power to You? A Case Against Binding Decisions as the Ultimate Access to Information Enforcement Tool", Holsen and Pasquier note that the general consensus of ATI advocates is for more powerful commissioners39, but argue that mediator commissioners can have very similar results in jurisdictions where the decisions of the ombudsman are respected.40 For instance, they argue both Switzerland and Scotland have a similar percentage of appeal decisions that result in the disclosure of information, but only Scotland's commissioner has binding powers, and most of its activities do not depend on this binding power. Polonca Kovač similarly argues that "formalised legal remedies" are less relevant in Nordic countries due to an "understanding of transparency deeply rooted in their culture". In this case "the infringements of rights are so rare that an elaborated formal system might even hinder the development of transparency".41 If we accept this argument, then the local transparency culture is an additional important element to consider in what powers are needed.
Notes ():
39: For instance, see Access Info Europe's recommendations for the requirements and powers of an oversight body in Austria, https://www.access-info.org/wp-content/uploads/Legal-Analysis-of-Austrian-FOIA_-Access-Info-Europe.pdf
40: Holsen, S., & Pasquier, M. (2019). More Power to You? A Case Against Binding Decisions as the Ultimate Access to Information Enforcement Tool. Administration and Society, 51(1), 34–62. https://doi.org/10.1177/0095399715611171
41: Kovač P, Legal Remedies in Exercising the Right to Information: A Comparative Overview, p. 640 [in laws of transparency in action]
Even assuming that authorities would rather keep more information secret and there is no culture of openness, a mediator commissioner might still be effective, because they reduce the gap in information law knowledge between authorities and requesters. In all models of ATI, the use of an exemption can introduce the potential for disagreement between requester and authority, and due to the fact that requesters cannot see the contested information, they cannot know whether the exemption has been applied appropriately. Even a non-binding assessment by a third party who has seen the information may be reassured that the public body who took the decision is behaving appropriately in withholding information. In other cases the mediation might change the decision of the authority to release information, as the requester may now have a much clearer sense of whether a legal action would be successful. This means the authority might now judge it a better use of time and effort to release (at least some information) immediately.
A mediator might adjust the balance, but it still leaves considerable space for an obstructive authority to undermine the ATI law. Where it is unlikely that requesters have the time or resources to pursue court action, even the knowledge that such an appeal has a good chance of success might not be enough to make it worth it. This can mean that authorities, even with the knowledge that their decision is likely to be incorrect in law, may consider it worthwhile to test whether the requester will appeal, and whether the commissioner's legal view is correct. At a minimum, public authorities have the ability to delay the release of information by forcing the full legal process. For citizens, who may not have the resources to make a legal challenge, this makes it more likely that information that should be released is denied. For journalists, who may have the institutional resources to contest decisions but need information promptly, any power of authorities to delay the release of information makes the ATI system less effective as a way of gathering information.
While there is such a thing as institutional culture of openness, and there are differences in how strong this culture is between countries, it is important to be careful when examining how this concept is used in practice. Culture is not fixed, and ATI is often meant to be part of how a culture shift in state secrecy is introduced.
Kovač highlights the idea that transparency is deeply rooted in Nordic countries, and points especially to Denmark as an example. The Danish Ombudsman has demonstrable success in having decisions reversed despite holding no powers to enforce this. In the Ombudsman's 2016 report they highlight an instance where they criticised a decision and recommended the Ministry of Taxation reopen a case. The ministry did so and granted the journalist full access to a document.42 However there is a limit to this approach. In 2018 the Ombudsman argued that correspondence between the Crown Prince and the Ministry of Culture had been written in the Prince's personal capacity as a member of the IOC and was therefore not covered under exemptions for Royal correspondence.43 The ministry reopened the case, but made a new decision that relied on an exemption for foreign policy interests; the Ombudsman said they did not have any grounds to criticise this new approach.44 From a system point of view, it may have been correct to withhold the information under the law, but it took several years to work this out because the lack of a binding order meant the original authority has the potential to create new reasoning when old reasoning is dismissed, rather than making a complete case at the start.45 It is difficult to see this as a complete victory for an open culture, and it is an example of how less binding arrangements can play to the advantage of a secretive authority.
Notes ():
42: Danish Parliamentary Ombudsman (2016), Annual report 2018, https://en.ombudsmanden.dk/publications/annual_reports/ar_2016/
43: Danish Parliamentary Ombudsman (2018), Annual report 2018, https://en.ombudsmanden.dk/publications/annual_reports/ar_2018/
44: Danish Parliamentary Ombudsman (2019), Annual report 2018, https://en.ombudsmanden.dk/publications/annual_reports/annual_report_2019/
45: This dynamic is a particular problem in Czechia, where the lack of a power to Czechia leads to a ‘ping pong' loop where the superior authority can continuously ask the original authority to reconsider the decision, meaning the complaint can never be escalated to the courts (which may only happen when both authorities are agreed) - see Stanislav Kadečka, Jan Brož, and Lukáš Rothanz (2018), The Laws of Transparency in Action: Freedom of Information in the Czech Republic : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
Looking at another Nordic country, Sweden has had a right to access public documents for centuries. This makes it the longest running Freedom of Information legislation in the world, and an important part of the story is how this legislation was accompanied by the existence of an ombudsman who did not just criticise but could take legal action to ensure access to information.
The Swedish ombudsman has the power to initiate prosecutions, and while this is rare in its modern duties (cases are far more commonly referred to other bodies), it was an important part of how the Ombudsman functioned when it was first established as an institution.46 The Ombudsman was created by the 1809 constitution and was tied into regulation of freedom of the press and the right to access public documents (ATI).47 During the first century of its operation, the Ombudsman used their prosecution power extensively.48 While the ability to bring prosecutions for ATI was absorbed back into the justice system by the 1949 Freedom of the Press Act, in the intervening century when ATI was becoming established, the Ombudsman could and did bring prosecutions against public officials for denying Access to Information. For instance, in the 1830s the Ombudsman "prosecuted the president of the central board of finance after he and his subordinates had prevented a citizen from seeing part of certain documents".49 The general understanding today of the ombudsman as a soft power institution blunts the sharper edges of how both the Ombudsman and ATI became so established in Sweden in the first place.
Notes ():
46: Lundgren, Fred (2014), Riksdagens ombudsman, En granskning av JO:s tillsynsverksamhet och domstolars självständighet, ur ett konstitutionellt perspektiv, https://www.diva-portal.org/smash/get/diva2:798642/FULLTEXT01.pdf,p. 50
47: Hans-Gunnar Axberger (2014), JO – i riksdagens tjänst, p. 105, Available at: https://www.jo.se/Global/Bok%20JO%20-%20i%20riksdagens%20tj%c3%a4nst.pdf
48: Hans-Gunnar Axberger (2014), JO – i riksdagens tjänst, p.18-19, Available at: https://www.jo.se/Global/Bok%20JO%20-%20i%20riksdagens%20tj%c3%a4nst.pdf
49: Jägerskiöld, S. (1961). The Swedish Ombudsman. University of Pennsylvania Law Review, 109(8), 1077. https://doi.org/10.2307/3310586
The long-running Swedish experiment should not be seen as an argument for the success of gentle persuasion. The historical Swedish Ombudsman was a prosecutor for citizens' rights. If anything, it implies that cultural change might take some decades or centuries before institutions have sufficient buy-in and respect that their power to compel action is less required.
Designing an Information Commissioner is not just a matter of cultural fit, but of designing an agent of cultural change. It can be expected that there will be general inertial resistance from public officials, but also that resistance will be sharpest from those whose behaviour would be exposed by greater transparency. The ATI requests that would have the most public benefit in succeeding (because they expose poor performance or corrupt behaviour) are those it is worth bad public officials working hard to prevent becoming public. In a system where the worst actors are habitually concealing their poor performance, there is little incentive for any other public officials to spend time or effort to be more transparent.
In Holsen and Pasquier's study, the Scottish Information Commissioner makes the point that the infrequent use of their formal power is what makes the softer enforcement work that they spend most of their time on more effective. An office that needs to steer behaviour change across many authorities benefits both from the ability to have a gentle and collaborative conversation, and also the ability to use more forceful measures when an authority is uninterested in this collaboration. A commissioner that constantly used coercive powers would paradoxically be a poor example for the usefulness of those powers. This hyper-combative commissioner would be ineffective in setting the pace of the wider system and would not have the resources to intervene in every case. In reality, enforcer commissioners spend most of their time taking the same actions as mediator commissioners; what is different is the option of switching tactics to solve difficult challenges.
Even if actively obstructive authorities represent a small minority, they can have a corrosive effect on the functioning of the overall systems. For the majority of public officials, a collaborative approach can help authorities adapt to new processes in a way that is least disruptive of existing processes and autonomy. However, this approach will not be enough for those who benefit (to the public detriment) from a culture of secrecy. For the fair treatment of the majority of public officials, the law needs to have real teeth.
An argument for a friendly, but likely to be less effective, collaborative culture can in part work as ideological defence of the current level of transparency. Holsen and Pasquier's interviews with the authors of the Swiss law found that they held a positive attitude towards the collaborative and voluntary relationship between authorities and the commissioner. However, this is not how the Commissioner's office saw it. The interviews with people working inside the Commissioner's office showed that the Commissioner's believed they'd be better able to fulfil their function with stronger powers. If there is not a strong intent to change public culture around secrecy, then it is unlikely a commissioner will be created that can effectively achieve that goal. Holsen and Pasquier's ultimate conclusion is there is not a straightforward relationship between binding decision power and effectiveness, but that "it depends" because even commissioners with binding decision powers may be useless in a culture where they cannot effectively use those powers. The key question is whether there is a willingness for cultural change, and appropriate institutions to support that.
Despite the arguments made in their defence, there are reasons to be sceptical of the merits of commissioners without binding powers, even in countries that are considered to have a strong culture of transparency. Aside from arguments about which is most effective for culture change, there is another argument that mediator commissioners are lower cost, but a cheaper commissioner is a false economy given the evidence than an effective ATI regime both cuts corruption and improves efficiency. It is better to spend the money once, on building an effective watchdog, than a thousand times on the inefficiencies that come from a culture of secrecy in public authorities.
Mediator commissioners are not pointless. There are clear theories about how they might have an impact, and are better than the alternative of no oversight body at all. In a system where the first form of redress is currently a court hearing, free and impartial advice from a mediator commissioner would be a significant improvement. In some cases they have also been transitional institutions that may later be given additional powers. For instance, Croatia's ATI law has been through several phases that strengthened the role of the oversight body.50 Mediator commissioners are a common approach that reflect a desire to create an advocate for the system while being ambivalent about the resources required. But they are not enough to unlock the full benefits of an ATI regime.
Notes ():
50: Anamarija Musa (2018),Croatia: The Transparency Landscape : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
Administrative review can be made more focused and effective
Before appeal to an oversight body, it is common that some kind of administrative review is required. Examining processes across Europe, administrative review can be grouped into three types. Some of these have a clear theory of effectiveness, but others run a risk of slowing down the process and providing a worse system of appeal to requesters.
For instance, assisted internal review, where the authority makes a new decision with the benefit of an opinion from the oversight body, has a clear reason why the authority may choose to make a new decision. There is new information that is suggestive of the success of a formal appeal. Superior review, where a superior administrative body reviews the decision also provides a clear explanation of how it might be effective. This process introduces a new decision maker and an opportunity for consistent practice across a sector. In this case, it is also possible to see why a decision might change. The third case, internal review, presents more of a problem. In this case, the requester complains to the original authority and asks for a review of their request, which is dealt with by their internal administrative policy. While this is an opportunity for a requester to challenge the reason for a refusal, the high success rate of decisions changing at internal review either points to widespread poor decision making on the initial decision, or that internal review has a perverse incentive for authorities to withhold information they might release without this process.
This report recommends that systems of internal review be replaced by a form of commissioner/ombudsman-led review, where the oversight body can track the number and success of internal review processes made throughout the system, and strategically decide when to allow authorities to review their own cases before considering it themselves. This represents a balance between allowing compliant authorities leeway to review and improve their own processes and decisions, minimising the options of non-compliant authorities, and the practical limits on resources of the oversight body.
Internal review is a process whereby an administrative organisation reconsiders its own decisions. For ATI appeals it can be found in Sweden, the UK, Ireland and the EU. This process is not specific to ATI and is a wider administrative practice meant to challenge decisions in a way that "is more accessible, quicker, and more cost-effective than external remedies such as appeals to tribunal and judicial review, and encourages improvement in the quality of initial decision-making".51 For instance in the UK internal review systems can also be found in social security administration and immigration control, but Tom Mullen argues that internal review systems in these cases "have failed to deliver the benefits claimed for them".52 In the case of ATI, there is good evidence that internal review can lead to changes in decisions, but also a danger the process incentivises more restrictive initial opinions that can be revised in the event of an appeal.
Notes ():
51: UK Law Commission, Consultation paper 187, Administrative redress: public bodies and the citizen, https://www.lawcom.gov.uk/app/uploads/2015/03/cp187_Administrative_Redress_Consultation.pdf
52: Mullin, T(2021), Internal review systems and administrative justice. In: Tomlinson, J., Thomas, R., Hertogh, M. and Kirkham, R. (eds.) The Oxford Handbook of Administrative Justice. Oxford University Press: New York. ISBN 9780190903084 (doi:10.1093/oxfordhb/9780190903084.013.4)
The good faith case for internal review is that it helps resolve simple errors and disagreements quickly. A more experienced official may come to a different decision, or correct an error made by a more junior official. For the Sweden system, Per Hagström makes the point that the initial decision may be made by an administrator with subject area knowledge of the document, but not the ATI law. Making a request for an appealable decision (internal review) can escalate the issue to an official with greater knowledge of the law.53 The requester has the opportunity to give the authority a preview of what their argument might be on appeal, and argue the case that an exemption has been misapplied. This provides new information to the authority that might lead them to at least partially reconsider their decision. In this case, internal review provides a quick and cheap way of challenging faulty decisions. From the point of view of the system, it allows obvious misapplication of exemptions to be corrected without putting strain on the resources of the ombudsman or Information Commissioner.
Notes ():
53: Per Hagström (2022), Myndighetens hierarki (Authority hierarchy), https://www.allmanhandling.se/myndighetens-hierarki/
There is also a less positive view of internal review. Mullin raises several potential problems: internal review might delay access to external remedies, discourage citizens from using them altogether, lead to different results in different areas, and more generally fail to correct bad decisions.54 We can add to this list that internal review creates an incentive for authorities to "try their luck" with a deliberately incorrect decision, knowing they have a chance to change their reasoning before an external party can review the decision.
Notes ():
54: Mullin, T(2021), Internal review systems and administrative justice. In: Tomlinson, J., Thomas, R., Hertogh, M. and Kirkham, R. (eds.) The Oxford Handbook of Administrative Justice. Oxford University Press: New York. ISBN 9780190903084 (doi:10.1093/oxfordhb/9780190903084.013.4), p. 5
Internal review is so successful at changing outcomes that it raises questions about the general picture of decisions being made. Scotland is one of the few jurisdictions to have comprehensive statistics on how almost all the country's public authorities handle ATI requests, giving the Commissioner statistics on the success of internal review. This data shows that 16% of internal reviews result in a new decision (not including when the original decision was partially upheld). According to the UK statistics for central government, 25% of internal reviews lead to at least some more information being released.55 At the EU level, internal review was successful in an average of 46% of requests over five years for the Commission and 51% of documents considered by the Council.56 This will mean different things in different cases, and partially upheld internal reviews may only be releasing a little more information. However, this still suggests that a significant proportion of decisions could be changed, simply by asking again. This makes a good case that requesters should ask for an internal review, but is not reassuring about the state of initial decision making.
Notes ():
55: Rumbul, R. and Parsons, A. (2021). Reforming FOI. [online] mySociety Research. Available at: https://research.mysociety.org/publications/reforming-foi .https://research.mysociety.org/html/reforming-foi/l/7.62.cqnd7.zdv8i.6njs-.y6jyl.html [Accessed 1 Mar 2022]
56: Bogdana Neamtu and Dacian C. Dragos (2018), Freedom of Information in the European Union: Legal Challenges and Practices of EU Institutions : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
There are a number of reasons why an internal review might lead to additional information being released. More junior decision-makers may withhold information out of inexperience or lack of familiarity with the law, and on internal review, more senior decision-makers may take a different view. But another potential explanation for why internal review is so effective is that it is a required step before appealing to the ombudsman or commissioner. In Bogdana Neamtu and Dacian Dragos's analysis of the EU ATI regime they suggest that "the willingness of institutions at the confirmatory stage to consider partial access to the documents requested [...] may be related to the very active role of the Ombudsman in this field".57 If internal review always takes place before the input of the commissioner/ombudsman, then there is an opportunity for authorities to reconsider their case before it is reviewed. This may result in a change of the logic of the decision, and possibly some additional information being released, in the hope of either satisfying the requester and avoiding an external review of the decision, or avoiding holes in the original logic leading to the release of more information.
Notes ():
57: Bogdana Neamtu and Dacian C. Dragos (2018), Freedom of Information in the European Union: Legal Challenges and Practices of EU Institutions : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
This implies that, when authorities are trying to withhold or delay information, the first response can withhold information using a weak justification that can be substituted for a more nuanced decision in the event of a request for internal review. There is direct evidence that this sometimes happens. In Scotland, an investigation by the Information Commissioner found an example where:
A special adviser [political appointee] wished to use an exemption to redact information despite advice from the FOI Unit indicating that the reasons would be "flimsy" and if the case appealed to the Commissioner it was doubtful that the exemption would be upheld. The internal correspondence in this case also indicated that information could be withheld in response to the initial request on that basis, but that the position could be reconsidered should a review be requested. The requester did not ask for a review.58
Notes ():
58: Office of the Scottish Information Commissioner (2018), Intervention Report, Scottish Government, para 151, https://www.gov.scot/binaries/content/documents/govscot/publications/foi-eir-release/2018/07/foi-18-01655/documents/foi-18-01655-related-documents-part-2-pdf/foi-18-01655-related-documents-part-2-pdf/govscot%3Adocument/FOI-18-01655%2B-%2Brelated%2Bdocuments%2B%2528part%2B2%2529.pdf
For evidence from the UK government, internal discussion of an ATI request similarly suggested an initial withholding of data where the justification could be changed later:
The HMT [Treasury] team will need to do a lot of consultation with former ministers who I suspect will be very sore about this[.] Much better to do as part of a release with the Inquiry. And I would use [Section] 31 for now. Can always revisit if goes to ICO [Information Commissioner].59
Notes ():
59: OpenDemocracy (2020), Revealed: Governments ‘Orwellian' unit blocked infected blood scandal disclosure, https://www.opendemocracy.net/en/freedom-of-information/revealed-governments-orwellian-unit-blocked-infected-blood-scandal-disclosure/
The existence of internal review becomes part of internal discussions about the political or institutional costs of compliance, allowing more information to be withheld by default without repercussions. When dealing with a sensitive request, an initial refusal with an opinion change at internal review is a cheap way of delaying the release of the information. For the requester who does not apply for an internal review in this event (as in the Scottish example), the requester will not receive information that they may have been legally entitled to receive.
Internal review systems present a challenge. On the one hand, if the first response from an authority tends to be from someone inexpert in the local ATI law, internal review is likely to be a cheap and fast way of correcting obvious mistakes. Ideally there would be fewer mistakes, but this might be a practical compromise. On the other hand, internal review has clear potential to be used by reluctant authorities to attempt to withhold information, or to stretch out the time scale of the request within legal processes. From the point of view of the requester, internal review does not provide independent reassurance that information has been withheld correctly.
The alternative to a system of administrative review is to allow (as with Slovenia and Croatia) appeals directly to the Information Commissioner after the initial decision. There is a logic to this approach: it assumes that the initial decision is generated by a process that produces defendable results. Even if it costs nothing to ask for an internal review, requesters should not have to ask twice for the ‘real' decision. The downside of this approach is that it sends far more complaints to the Information Commissioner. Given problems across jurisdictions with appeals to the regulator taking far longer than they should, managing their resources is a real concern.60 Put this way, internal review is a form of rationing access to the Information Commissioner, which raises the questions about whether this is the fairest or most effective way of doing so.
Notes ():
60: Background interviews C and E
The ideal solution to this problem is one where authorities working in good faith are able to correct mistakes in a way that is fast and cheap to the requester, while limiting the ability of obstructive authorities to delay or avoid the release of what should be public information. There is some low hanging fruit in this area for some jurisdictions; internal review needs well defined time limits for the process and explicit rules to deal with administrative silence (Belgium, Ireland and Scotland have explicit rules for administrative silence, while Denmark and the UK do not). However, this does not resolve the problem that even well-defined internal review can be used to illegitimately obstruct the release of information.
A way of resolving this is to make internal and external review less distinct stages, and more different strategies that can be adopted by the oversight body. Mullin recommends redefining internal review as part of, rather than an alternative to, external review and that "internal review should take place automatically when an appeal is lodged and should not delay the hearing of the appeal".61 For ATI requests, this would mean that appeals should be sent directly to the oversight body, at which point the authority should also reconsider its initial decision. While this runs into existing problems of funding and delays, this would create more awareness in the oversight body of the scale of the internal review problem, and provide more options for intervention.
Notes ():
61: Mullin, T(2021), Internal review systems and administrative justice. In: Tomlinson, J., Thomas, R., Hertogh, M. and Kirkham, R. (eds.) The Oxford Handbook of Administrative Justice. Oxford University Press: New York. ISBN 9780190903084 (doi:10.1093/oxfordhb/9780190903084.013.4), p. 10
From this position of greater knowledge, oversight bodies (with sufficient resources) would have new potential approaches to reacting to problems. Oversight bodies could make greater use of existing powers (where present) to recommend/require structural changes to internal processes to improve the effectiveness of the initial decision making, when the statistical evidence suggested problems.
There is also the prospect of more tailored interventions in how oversight bodies prioritise the wider range of appeals they then have access to. While in general, all types of request should be treated equally, it would be appropriate as a form of sanction for the oversight body to pay more attention to non-compliant authorities, to ensure the authority's review happens at an appropriate timescale and that the oversight body's review is itself prompt.
The goal of this process would be to tether successful, cheap, and quick internal review to external incentives. Rather than overload the oversight body by requiring them to process all reviews, this approach would hopefully lead to a dynamic of supervised decision-making, where more problems are resolved at a lower level. Many appeals are likely to be resolved between the requester and the authority (with light intervention by the oversight body) in advance of the oversight body formally processing the appeal. This would in substance be similar to the existing internal review process for most authorities. However, by making internal review a component of an appeal to the oversight body, rather than an initial step, it gives the oversight body more ability to deal with the rare case when the internal review process is being used as a delaying tactic. Special strategies for highly non-compliant authorities are an important part of the oversight toolkit, in that they reward authorities who are engaged in the system in good faith.
The above process has existing jurisdictions with internal review in mind, but the basic premise of providing the oversight body powers or resources to make strategic enforcement choices applies in other contexts. In jurisdictions where this was practical, the idea that the oversight body is the manager (if not necessarily the reviewer) or every appeal would create a framework where peer review of appeals between similar organisations could be managed. In jurisdictions where escalation proceeds directly to the court, this could be achieved by the oversight body providing targeted legal aid and discounted fees for requesters to certain authorities. The goal is a targeted allocation of resources to lead to the correct behaviour by authorities, rather than relying on requester time, money or persistence to escalate cases past the internal review stage.
Trade-offs between general and specific oversight bodies
A key division in oversight bodies is between specific and general oversight. These can be divided into specific oversight bodies focused on ATI (Information Commissioners) and more general approaches (Ombudsman or court systems). The trade-off in this question is between "specialised knowledge and focus" and "ability to impact change". More generalised bodies might have greater stature and soft influence, but also see ATI as a smaller part of the mission.
In general, this report favours more specialist organisations as having the greatest potential for good ATI oversight. As Linda Reif argues, specialist bodies accumulate expertise leading to "more consistent interpretations and rulings" and have greater focus on the task of ATI.62 That said, there are practical considerations in terms of working best within existing systems that need to be considered, along with the powers and goals of the oversight body.
Notes ():
62: Linda Reif (International Ombudsman Institute), quoted in Holsen, S., & Pasquier, M. (2012). Insight on Oversight: The Role of Information Commissioners in the Implementation of Access to Information Policies. Journal of Information Policy, w, 214–241.
If the intention is that the body has non-binding powers only, there is a question about how to make this soft power more effective. When existing institutions exist that are well respected, more general ombudsman roles might be more effective at promoting Freedom of Information. A new institution that depends on soft power, by contrast, may struggle to establish itself. Where there are not any existing institutions, this dynamic is part of the case for more formal powers as a way of encouraging compliance with the new oversight body.
Related to this question is whether combining data protection and ATI as a single role is a good idea. Both arrangements - single and dual responsibilities - are common. Croatia, Portugal and Ireland have separate institutions, while in Germany, Hungary and the UK they are joined. In a previous report, we've made an argument for splitting the institutions in the UK.63 The original reason the institutions were joined, a concern that they would adopt different understandings of privacy, is less relevant as definitions of private information have become increasingly part of international frameworks like GDPR. This issue also reflects a change over time, as the growing importance of data protection over the past few decades has shifted the relative importance of the two roles, and in practice the head of the organisation is sourced from a international recruiting pool of data protection regulators. Similarly, we heard from several interviewees that the Länder Information Commissioners in Germany tended to be primarily data protection roles.64 Specialised data protection roles benefit from international learning and recruitment, but this is less the case for ATI oversight, where knowledge of a specific state's system of administrative oversight might be more relevant.
Notes ():
63: Rumbul, R. and Parsons, A. (2021). Reforming FOI. [online] mySociety Research. Available at: https://research.mysociety.org/publications/reforming-foi [Accessed 1 Mar 2022].
64: Background interviews A and C.
In general, this question of the merits of joined or separate data protection and ATI roles is under-explored. The most detailed discussion of this question was a report commissioned in Canada in 2005 to review whether their separate information and data protection commissioners should be merged.65 Ultimately recommending against a merger, this report laid out a clear case of the conflicting arguments. The report concluded there was not a strong argument on the basis of resources in either direction. There is likely to be some efficiency in a merger/shared services arrangement, but this was minimal in the context of two relatively small public bodies.
Notes ():
65: La Forest, G. V. (2005). The Offices of the Information and Privacy Commissioners : The merger and related issues. Retrieved from https://www.justice.gc.ca/eng/rp-pr/csj-sjc/atip-aiprp/ip/index.html
The key question is whether the existence of ATI and data protection within the same organisation is to their mutual benefit. Part of this question is whether conflicts between public access to information and privacy are best resolved internally, or via two separate institutions. One argument that the powers belong together is based on the idea that they are fundamentally complementary powers. As the former UK Information Commissioner Elizabeth Denham (who had previous experience of Canada's system) put it, the two areas are "two sides of the same coin" and "FOI and data protection [...] belong together, because they have to adjudicate on freedom of expression as well as privacy."66 This argues that one institution can resolve problems internally and provide a consistent approach.
Notes ():
66: Evidence before UK House of Commons Public Administration and Constitutional Affairs Committee (2021), https://committees.parliament.uk/oralevidence/3069/html/
The argument against this is there is less room for divergence than assumed. Arguing against there being a consistency problem, (and as with our later argument in a UK context), the 2005 Canadian report argues that "the courts have developed an extensive body of jurisprudence on the meaning of ‘personal information,' which to a considerable extent dictates the advice the Information Commissioner renders to institutions". In the rare cases (two as of 2005) where the Canadian Privacy Commissioner and Information Commissioner had been in conflict, the court agreed with the Information Commissioner. In contrast to the benefits of resolving the conflict internally, the report argued that "Canadians are generally well served by having the commissioners presenting contrasting views to the government, the courts, and the general public". This suggests both combined and separate Information Commissioners are likely to come to similar conclusions on privacy within the wider legal framework.
The more interesting argument about the combination of data protection and ATI is around the soft power of commissioners. The Canadian report outlined an argument for a case where the combined oversight body might have greater persuasive power over government, and provide more consistent advice. There is a specifically an argument that the privacy work of a joined commission may be to the benefit of ATI work:
It is also possible, however, that the interactions that single commissioners have with government on the privacy front, which tend to be more consultative and less adversarial than is the case with access, may foster a more cooperative approach on access issues. Government officials who have engaged in productive dialogue with the commissioner on privacy issues are less likely to adopt a hostile and adversarial stance in discussions with the same commissioner on access issues.67
Notes ():
67: La Forest, G. V. (2005). The Offices of the Information and Privacy Commissioners : The merger and related issues. Retrieved from https://www.justice.gc.ca/eng/rp-pr/csj-sjc/atip-aiprp/ip/index.html, p. 36
This is a more specific version of the argument for using an existing ombudsman. This argument says that the soft power of an ATI oversight role can be increased through pairing with data protection, even if that means the organisation is less focused on ATI. The appropriate position in this trade-off depends on the perceived problem in the current system. The opposing arguments to this are the risk that the combined authority would focus more on one portfolio than the other, or that the differences in tasks would overburden the combined oversight body and diminish its capacity to achieve either goal.
While there is a good case for more specific oversight bodies in general, there are clearly trade-offs that mean combination with other functions can have benefits. Our general point is that new institutions need to be set up to succeed, with powers, structure and (especially) funding that is coherent, and matches the goals of the organisation. A concern that a split organisation may have insufficient funding, or a joined organisation give focus to one role at the expense of the other, may in reality be better expressed as a problem in how the oversight bodies are funded or supported in the first place.
Effective ATI oversight needs sufficient budget to deliver on its goals
A problem for ATI oversight is that the role includes overseeing the workings of governments who have levers to obstruct their work. This may be informally, by setting the tone of how officials in government and the wider public sector should treat Access to Information. It may also be active, obstructing the work of oversight, or ensuring that the commissioner or ombudsman is underfunded.
In 2017, a group of European Freedom of Information Commissioners and ombudspersons Institutes made a resolution supporting "strong Information Commissioners and ombudspersons as mediators between the state and citizens". The resolution argues that people in their countries should "have the right to access independent offices" and that these offices "should have appropriate funding as well as human and legal resources corresponding with the importance of their task".68 In many countries ATI is recognised as a separate constitutional right, or otherwise it is understood that part of its function is around prompting government accountability to citizens. As well as having legal powers and influence to enforce FOI law, oversight bodies need to be independent and well funded. As Sarah Holsten argues, political independence is "crucial to the organization's capacity to carry out its mandate".69 Despite this, there remain clear problems in a number of countries in how commissioners and ombudsman institutes are appointed or funded.
Notes ():
68: Resolution of the European Freedom of Information Commissioners and OmbudspersonsBerlin, February 24th, 2017, https://www.bfdi.bund.de/SharedDocs/Downloads/DE/IFG/AGID_IFK/2017_Treffen_EU_Infobeauftragte_EN_Resolution.pdf?__blob=publicationFile&v=4
69: Holsen, S., & Pasquier, M. (2012). Insight on Oversight: The Role of Information Commissioners in the Implementation of Access to Information Policies. Journal of Information Policy, w, 214–241.
Political independence is not the only factor limiting the effectiveness of oversight bodies, but because of the way independence affects resourcing, it is a significant one. Even if there are limits to the powers of mediator Information Commissioners, there are much clearer limits to under-funded mediator Information Commissioners.
An Information Commissioner or Ombudsman in Europe is typically one defined person, occupying an office. Using the RTI rating statistics, 56% of European ATI regimes had a good appointment process for this office. The most significant example of political interference in removing an Information Commissioner was when the role was abolished and recreated in Hungary, having the effect of prematurely ending the term of the Information Commissioner.70 This kind of action is very uncommon, but it is far more common for the Information Commissioner/ombudsman to be a government rather than parliamentary appointment. In general, it is better for constitutional watchdogs to be a parliamentary appointment, reflecting that their role is one of regulation of the government, rather than implementation of the government agenda. However, the more common complaint is not that political independence is compromised by appointment powers, but that political independence (and effectiveness) is undermined by government control of the funding of the organisation.
Notes ():
70: Petra Lea Láncos (2018), Freedom of Information in Hungary: A Shifting Landscape : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
Even if Information Commissionerss are both willing and have good legal powers to enforce ATI law, they can be substantively limited by funding that constricts their ability to manage required duties (such as appeals casework), or extend themselves into promotion or monitoring work that would help improve compliance. Using the RTI Ratings, 54% of European oversight bodies had "good financial independence". However, individual comments made by RTI scorers suggest this figure might be optimistic, as in some cases a good mark was given for financial independence, but the comment indicated political problems with funding. Where funds are assigned by a government department, there is a conflict of interest where underfunding the oversight body decreases the effectiveness of the overall ATI system.
In recent evidence before the UK Parliament, the UK Information Commissioner, explaining why they hadn't used an enforcement notice against a non-compliant government department, pointed out a long-term cut in their budget and that "[f]or us to take enforcement action we have to have the resources—the legal resources—to be ready for litigation. It is not a light decision to take. I think that if we had more resources, we would be able to do more."71 The UK Information Commissioner has a wide range of legal tools available (in its RTI Rating score, the UK has one of the more comprehensive sets of commissioner powers in Europe), but cannot practically make use of them because of limited funding. In Ireland and France, scorers for the RTI rating made comments that the funding by the ministry left the Commissioner under-resourced and politically vulnerable.72
Notes ():
71: Evidence before UK House of Commons Public Administration and Constitutional Affairs Committee (2021), https://committees.parliament.uk/oralevidence/3069/html/
72: Scorer's comments on RTI Rating, France, Ireland http://www.rti-rating.org/country-data/France/, https://www.rti-rating.org/country-data/Ireland/
Funding via the Parliament/national legislature removes a structural conflict of interest, but still requires Parliament to have an active interest in better promotion and compliance with the ATI law. For instance, comments made by the RTI rating scores for Croatia said that the funding was voted by Parliament, but that these funds were not sufficient for full implementation.73 The case still has to be made that, in line with the evidence base, a well working ATI system improves governance and reduces corruption. The case for generous funding of relatively small oversight bodies is that it is part of unlocking good governance throughout the system.
Notes ():
73: Scorer's comments on RTI Rating, Croatia http://www.rti-rating.org/country-data/France/, https://www.rti-rating.org/country-data/Croatia/
Better statistics help make oversight bodies more effective
The production of public, authority-level, statistics about ATI compliance are a vital tool in improving understanding of existing practice and enforcement of good practice.
A recurring question in this report is how oversight bodies can make the best use of a small budget for enforcement and awareness. Understanding the effectiveness of a system on its own terms over time, or comparatively between countries, requires information on how it is functioning. A key tool that most oversight bodies and campaigning groups do not have is a detailed statistical view of how the local ATI system is functioning.
Specialist oversight bodies may publish reports covering their actions, but will rarely have understanding of the wider picture of ATI appeals. General oversight bodies (such as ombudsman covering a range of public services) may collate statistics internally, but ATI processes are such a small part of their role that the area does not get much coverage in their public reports. Systematic statistics of the number of ATI requests being made, or their flow through the process, are rare.
This absence of information is especially important when administrative review is used as a filter before appeals to the oversight body. In these systems, while the oversight body will know how many people have applied for advice or an appeal, they will have no idea if this represents a small or a large fraction of requests where the decision had previously been appealed. They will also have no idea if administrative review is leading to different outcomes, and if so, whether certain authorities reverse/have their decisions reversed more than others.
While it is important for the oversight body to have better access to this information, public access to the statistics would also be useful and allow external groups to understand the effectiveness of the system, and campaign for specific improvements.
Statistics about ATI fall into (roughly) six categories:
- Process - How many ATI requests were received, under what law were they processed74, or were they judged as outside ATI law.
- Timeliness - Compliance with legal timescales for replies, and use of legally allowed time extensions, or the number of days between request and response.
- Decision - Whether information was partially or entirely granted to the requester.
- Appeals - Number of appeals made, and whether they lead to a change in outcome.
- Exemptions - Which legal grounds were used to deny ATI requests.
- Demographics - Either of the overall group of requesters, or disaggregating of statistics by different groups (e.g. are their demographic patterns to decisions granted, etc).
Notes ():
74: For instance, the statistics in Scotland separate into Freedom of Information and Environmental Information Requests. This is unusual to separate, but useful for understanding the different effects of different disclosure laws.
Statistics may exist at several different levels of aggregation:
- Request level - Details for each request made, the outcome, if an appeal was made, and if the appeal was successful.
- Authority level - Information at the authority level to show how many requests were made, how many of those were granted, and overall use of exemptions.
- System level - Aggregate understanding of the number of requests, the proportion granted, or the number of appeals across the entire ATI system.
It is rare for information to exist that is not at the system level and mostly about appeals. This is because the most common form of statistics are produced by oversight bodies on their own activities.
The best source of information an oversight body has is information about their own activities. Where an Information Commissioner has a formal or informal role in mediation or complaint dispute, it is likely that statistics on the overall volume of complaints and outcomes of these complaints will be published in annual reports where these are required (46% of European countries in the RTI Rating). For instance, CADA in France publishes annual statistics that can be used to understand the relative proportion of complaints to different kinds of public authority, as well as changes over time.
When ATI oversight is handled by generalist oversight bodies, the statistics may be hidden by other areas of work. For instance, statistics in the annual reports of the Swedish Ombudsman count the total number of ATI cases, but do not separate out their status or outcome. Instead, this is aggregated by authority type over all complaints to the Ombudsman. Similarly Swedish court statistics summarise ‘secrecy cases' cases into ‘other'.75 As ATI may be a small part of a generic oversight body's role, statistics may be much harder to access, even if they technically exist. These reports can provide details about this level of the oversight process, but they will not have information about internal or administrative review to public authorities.
Notes ():
75: Official statistics of Sweden(2020), Court Statistics, https://www.domstol.se/globalassets/filer/gemensamt-innehall/styrning-och-riktlinjer/statistik/2021/court_statistics_2020.pdf
To close this gap, one approach is to require public authorities to publish statistics of how they have dealt with ATI requests. The OAS's Inter-American Model Law 2.0 on Access to Public Information suggests public bodies should record annual reports covering the number of requests received and granted, those replied to late, the part of law used to deny access and fees charged. They also recommend that these statistics where possible should be disaggregated by gender and ethnicity through an information form to be completed by requesters.76 In practice, very few countries require levels of disclosure approaching this amount of detail.
Notes ():
76: OAS (2020), Model Inter-American Law 2.0 on Access to Public Information, Available at: http://www.oas.org/en/sla/dil/docs/publication_Inter-American_Model_Law_2_0_on_Access_to_Public_Information.pdf
Public authorities in 49% of European countries in the RTI Rating have a requirement to publish statistics on how many requests were received and how they were dealt with. The disclosure may not be consistent, and some may include it in annual reports while others do not.77 In Belgium, there is a requirement for authorities to register all ATI requests. In the Flemish region, there is a central register which is used by the Information Commissioner (Flemish Appellate Body on the Openness of Government) in their annual reports.78 Even when this data is not formally published, partial pictures can sometimes be understood through ATI requests to the authorities, but this requires a large amount of work, and can reveal that different authorities are working to different standards in how much information they collect. For instance, mySociety in 2017 was able to use ATI requests to get information about the volume of ATI requests made by the UK local government, and the different standards of information they collect.79 The data may not cover information that is useful for understanding the effectiveness of appeals such as internal reviews, or non-outcome complaints such as administrative silence or late replies.
Notes ():
77: Background interview D
78: Stef Keunen and Steven Van Garsse (2018), Access to Information in Belgium : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
79: Parsons, A. and Rumbul, R. (2019). Freedom of Information in Local Government. [online] mySociety Research. Available at: https://research.mysociety.org/publications/freedom-information-local-government [Accessed 1 Mar 2022].
The problem with requiring individual disclosure of statistics is that it is harder to get the benefit the more public authorities are covered by the law. For instance, for the EU's three core institutions, comparing statistics from the individual reports is not a large amount of work. However, for systems with hundreds or thousands of authorities, the work required to put the pieces together becomes unmanageable. Requirements to collect and publish information can be made more effective if this data is standardised and republished by a central authority - creating a single dataset of information across the system.
There are several examples of this that cover the core government departments of a country. Germany's Ministry of the Interior (BMI) publishes statistics on bodies covered by the federal level legislation. This includes information on the number of requests received, granted, partially granted and refused, fees charged, the number and status of internal reviews and legal action, but not appeals to the commissioner.80 The UK's Cabinet Office publishes similar information on ATI requests made to UK central government departments. This dataset includes the different kinds of exemptions used by each authority. The risk is that generalising from the experience of the central government (which covers different government functions and may have a different balance of requester types) may not give a good guide as to how the national system of ATI is working.
Notes ():
80: German Federal Ministry of the Interior and Community (2022), Freedom of Information page, https://www.bmi.bund.de/DE/themen/moderne-verwaltung/open-government/informationsfreiheitsgesetz/informationsfreiheitsgesetz-node.html
Solving this problem means scaling up the approach to build a comprehensive picture across the whole system (or at least, most significant authorities). The Irish Information Commissioner's annual report includes a section of statistics on the overall number of ATI requests and appeals, which are supplied to the Information Commissioner either directly by public authorities or by intermediary groups (for instance, for local authorities).81
Notes ():
81: Information Commissioner (Ireland) (2020), Annual report 2020 https://www.oic.ie/publications/annual-reports/OICAR2020FinalWebEN.pdf
The Croatian Information Commissioner's annual report compiles data from reports by all public authorities covered by the ATI law. The information gathered allows common reasons for refusal to be explored. The Croatian Information Commissioner is exploring creating an online register to track these statistics and simplify the process for both authorities and the commissioner.82 The Scottish Information Commissioner's process provides a good example of how a system like this might work. Their process requires public authorities to complete a form and deposit it in an online portal four times a year. It collects 110 different statistics (including counts of different kinds of exclusion used) from 508 public authorities.
Notes ():
82: Information Commissioner (Croatia) (2021), Annual report 2021 https://pristupinfo.hr/wp-content/uploads/2021/05/IC-Annual-Report-2020-Summary.docx?x57830
The Scottish dataset is the only system using internal review that has statistical information about the effectiveness of that mechanism across all major public authorities and releases this on a per-authority level (Ireland does have information about the number of internal reviews but no more than that). This information is available for public download on an authority level, and represents the richest level of statistical information available about a whole jurisdiction.
A common theme between these three jurisdictions is that they are small in scale compared to some European countries. Ireland and Croatia each reported around 32,000 ATI requests in their latest annual report, while Scotland reported around 70,000. For a sense of scale, just the local governments in the UK received around 567,000 FOI requests in one year. The UK may be an outlier in this scale of local use, but in general there is not a good example of data collection in larger countries that can also contain thousands of authorities subject to ATI. This is not to say that this approach cannot be scaled up or adjusted. Even providing coverage of 500 central and local government authorities is likely to build a rich statistical picture that can be extrapolated to represent the majority of Access to Information requests.
ATI request portals provide another source of statistics. When all ATI requests are made through a single portal it presents an opportunity to have a single set of statistics about usage, and to ask additional questions to understand more about the characteristics of requesters. The Spanish Transparency Portal is an example of this. However, despite the information being readily available, the Spanish public statistics only show aggregate analysis and not a comparative picture of success at different authorities. Third party portals, such as Alaveteli websites, provide a version of this single portal picture. These statistics are partial figures (especially if certain kinds of ATI users do not use the site), and are dependent on users of the site accurately updating it with details about the request. However, these sites are also a rare source of request level information, which can be useful in examining the content that is denied under specific exemptions, and can provide a wider range of these requests than are available to any individual authority.
Gathering more statistics about ATI systems requires effort, but is an effective way of identifying where the system is working, and where information is needed. A notable feature of Scotland's high level of detail is how analytical and informed the Information Commissioner's understanding of the health of the system can be. In a 2021 speech, they were able to easily detail the total number of requests, how this number had changed, changes in the amount of requests released on time, and how much information was being released.83 This is a level of detail that is just unavailable for most systems, and has clear uses in articulating the value of ATI and the problems in delivering it. An oversight body that only knows about the problems that are brought to its attention will miss important issues, but also will not be able to talk about the success and impact of a functioning system of access to public information.
Notes ():
83: Information Commissioner (Scotland) (2021), Reflections on FOI in 2021, https://www.itspublicknowledge.info/home/News/20211123.aspx
Prospects for reform
Once created, ATI regimes are not fixed. Through neglect or obstruction, they might decay in usefulness. But when they are shown to be worthwhile, and the inherent problems engaged with, ATI systems can improve. Croatia's oversight process went through several iterations improving the independence and specific focus of the Information Commissioner.
In 2016, Germany's federal Information Commissioner became independent of the government.84 In 2022, a new draft law in Ukraine had provision for a new independent supervisory body for ATI (a process we hope will continue in a free and democratic Ukraine).85 On the other hand, stronger Information Commissioners have been proposed as part of reforms in Italy and the Netherlands, but were removed as part of the legislative process.86
Notes ():
84: German Federal Ministry of the Interior and Community (2015), Federal Commissioner for Data Protection and Freedom of Information will be strengthened, https://www.bmi.bund.de/SharedDocs/pressemitteilungen/DE/2015/12/bfdi-wird-unabhaengig.html
85: Доступ до правди (2022), Experts will discuss the draft law on personal data and the supervisory body in the field of access, https://dostup.pravda.com.ua/news/publications/eksperty-obhovoriat-proekt-zakonu-pro-personalni-dani-i-nahliadovyi-orhan-u-sferi-dostupu>>>; Oleksiy Stolyarenko (2021), Data Privacy Reform in Ukraine: What's New? <<<FLOATING LINK: https://www.connectontech.com/data-privacy-reform-in-ukraine-whats-new/
86: Kars J. de Graaf, Albert T. Marseille, and Hanna D. Tolsma (2018), Transparency and Access to Government Information in the Netherlands : in The laws of transparency in action : a European perspective. Palgrave Macmillan.; Background interview B.
The argument about commissioners is not just an argument about the appropriate way to reach an agreed upon level of openness. The question is part of the debate on whether more openness is a good thing. The most important fact in the argument for better ATI oversight is not the obvious point that better oversight leads to more effective ATI regimes, but that more effective ATI regimes are good for societies. In aggregate terms, they have demonstrably positive effects in reducing corruption and improving outcomes. ATI oversight is important because getting it right is an influential lever that moves the whole system of government towards more effective and transparent processes.
It is important to remember that ATI is a citizen's right, in many cases recognised as a fundamental constitutional right. It helps people exercise greater leverage over the institutions that have power over their lives. To make it a tool that citizens can use, rather than one that is mostly used by journalists and NGOs, it is important that free and effective rights of redress exist, and effective ATI is not priced out of everyday use. If this is not the case, then regardless of how open the law is, only certain kinds of requesters can realistically make use of their ability to appeal, and the system will not reach its full potential.
About mySociety
mySociety is a not-for-profit group, based in the UK but working with partners internationally. We build and share digital technologies that help people be active citizens, across the four practice areas of Democracy, Transparency, Community and Climate. As one of the first civic technology organisations to be established, we are committed to building the Civic Technology community and undertaking rigorous research that tests our actions, assumptions and impacts. Our global research work into digital democracy, civic technology and user-centred design has positioned mySociety as a leading authority in digital civic engagement and participation.
mySociety runs WhatDoTheyKnow.com in the UK, and maintains the Alaveteli software and network which powers Freedom of Information sites around the world. This software allows users to make public requests through a website that are sent to authorities via email. The replies are also published so any information received is accessible to the general public as well as the original requester.
This research was funded as part of a mySociety transparency programme funded by the Adessium Foundation and the Swedish Postcode Foundation.
Copyright and licencing
This work is licensed under a Creative Commons Attribution 4.0 International Licence. (CC BY 4.0)
With the exception of any third-party material listed below, the contents of this report can be copied and adapted into new formats, as long as there is appropriate attribution.
Report cover by Joshua Hibbert on Unsplash
This report, and accompanying material, can be downloaded at:
Related research
Footnotes
1 Dragoș, D. C., Kovač, P., & Marseille, A. T. (Eds.). (2018). The Laws of Transparency in Action : a European perspective. Palgrave Macmillan.
2 Worthy, Ben, Freedom of Information in Europe: Creation, Context and Conflict (June 24, 2019). Contested Trade-Offs: Transparency and Secrecy in European Democracies. Forthcoming , Available at SSRN: https://ssrn.com/abstract=3409066 or http://dx.doi.org/10.2139/ssrn.3409066; Dragoș, D. C., Kovač, P., & Marseille, A. T. (2018). The laws of transparency in action : a European perspective.
3 Chen, C., & Ganapati, S. (2021). Do transparency mechanisms reduce government corruption? A meta-analysis. International Review of Administrative Sciences. https://doi.org/10.1177/00208523211033236
4 Vadlamannati, K. C., & Cooray, A. (2016). Do freedom of information laws improve bureaucratic efficiency? An empirical investigation. Oxford Economic Papers, 68(4), 968–993. https://doi.org/10.1093/oep/gpw008
5 Cordis, A. S., & Warren, P. L. (2014). Sunshine as disinfectant: The effect of state Freedom of Information Act laws on public corruption. Journal of Public Economics, 115(March), 18–36. https://doi.org/10.1016/j.jpubeco.2014.03.010
6 Honig, D., Lall, R., & Parks, B. C. (2022). When Does Transparency Improve Performance? Evidence from 23,000 Projects in 148 Countries. American Journal Of Political Science, 00(0), 1–21. https://doi.org/10.1111/ajps.12698
7 Analysis of data downloaded for European countries from rti-rating.org.
8 Petra Lea Láncos (2018), Freedom of Information in Hungary: A Shifting Landscape : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
9 Stanislav Kadečka, Jan Brož, and Lukáš Rothanz (2018), The Laws of Transparency in Action: Freedom of Information in the Czech Republic : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
10 Pernille Boye Koch, Rikke Gottrup, and Michael Gøtze (2018), Special Report: Transparency on a Bumpy Road— Denmark : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
11 Yseult Marique and Emmanuel Slautsky (2018), Freedom of Information in France: Law and Practice : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
12 Stef Keunen and Steven Van Garsse (2018), Access to Information in Belgium : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
13 There's no settled gender neutral pluralisation of an Ombudsman. The original Swedish term (ombudsmän) is gender-neutral, but generally it is not incorrect to apply English rules and assumptions once something has become an English word (ombudsmen, or ombudspersons). Alternative approaches include shorting to ‘Ombuds'. The International Ombudsman Institute sidesteps the problem by referring to multiple ‘ombudsman institutions', which also reflects that the idea is often localised using a term that is more descriptive in the local language (public advocate, etc).
14 Stef Keunen and Steven Van Garsse (2018), Access to Information in Belgium : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
15 Analysis of data downloaded for European countries from rti-rating.org.
16 Peck, J., & Theodore, N. (2015). Fast Policy: Experimental Statecraft at the Thresholds of Neoliberalism. Minneapolis: Univ Of Minnesota Press.
17 Kucsko-Stadlmayer, G. (2009). The Spread of the Ombudsman Idea in Europe. Retrieved from http://www.theioi.org/downloads/34chi/Stockholm Conference_25. Back to the Roots_Gabriele%25Kucsko Stadlmayer.pdf
18 Kucsko-Stadlmayer, G. (2009). The Spread of the Ombudsman Idea in Europe. Retrieved from http://www.theioi.org/downloads/34chi/Stockholm Conference_25. Back to the Roots_Gabriele%25Kucsko Stadlmayer.pdf
19 Analysis of data downloaded for European countries from rti-rating.org.
20 Depending on the nature of the request and the appeal: Fees For A Review | OIC
21 Lemieux, V. L., & Trapnell, S. E. (2016). Public Access to Information for Development: A Guide to the Effective Implementation of Right to Information Laws. https://doi.org/10.1596/978-1-4648-0879-1, p. 55
22 It is likely that the true number is higher, as several comments for 0 scoring countries indicated a lack of direct mention of judicial appeal in the law, when this may simply be implied by the wider structure.
23 Courts Service (Ireland) (2020), How to Appeal an FOI Decision, https://www.courts.ie/how-appeal-foi-decision
24 Independent Commission on Freedom of Information. (2016). Independent Commission on Freedom of Information Report. Retrieved from https://www.gov.uk/government/publications/independent-commission-on-freedom-of-information-report
25 Polonca Kovac (2018), Slovenia on the Path to Proactive Transparency : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
26 Anamarija Musa (2018),Croatia: The Transparency Landscape : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
27 Доступ до правди (2022), About/Complaints, https://dostup.pravda.com.ua/help/about#complaint
28 Yseult Marique and Emmanuel Slautsky (2018), Freedom of Information in France: Law and Practice : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
29 Pernille Boye Koch, Rikke Gottrup, and Michael Gøtze (2018), Special Report: Transparency on a Bumpy Road— Denmark : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
30 See English language reports of the Danish Ombudsman, 2013, 2016 and 2018 and search "was entitled to access": https://en.ombudsmanden.dk/publications/annual_reports/
31 See annual reports of the Swedish Ombudsman https://www.jo.se/en/About-JO/Annual-reports/
32 Christoph Emanuel Mueller, Bettina Engewald, and Marius Herr (2018), Freedom of Information in Germany : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
33 Rumbul, R. and Parsons, A. (2021). Reforming FOI. [online] mySociety Research. Available at: https://research.mysociety.org/publications/reforming-foi [Accessed 1 Mar 2022].
34 Stef Keunen and Steven Van Garsse (2018), Access to Information in Belgium : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
35 Stef Keunen and Steven Van Garsse (2018), Access to Information in Belgium : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
36 Paola Savona and Anna Simonati (2018), Transparency in Action in Italy: The Triple Right of Access and Its Complicated Life : in The laws of transparency in action : a European perspective. Palgrave Macmillan., background interview B
37 Honig, D., Lall, R., & Parks, B. C. (2022). When Does Transparency Improve Performance? Evidence from 23,000 Projects in 148 Countries. American Journal Of Political Science, 00(0), 1–21. https://doi.org/10.1111/ajps.12698
38 Cordis, A. S., & Warren, P. L. (2014). Sunshine as disinfectant: The effect of state Freedom of Information Act laws on public corruption. Journal of Public Economics, 115(March), 18–36. https://doi.org/10.1016/j.jpubeco.2014.03.010
39 For instance, see Access Info Europe's recommendations for the requirements and powers of an oversight body in Austria, https://www.access-info.org/wp-content/uploads/Legal-Analysis-of-Austrian-FOIA_-Access-Info-Europe.pdf
40 Holsen, S., & Pasquier, M. (2019). More Power to You? A Case Against Binding Decisions as the Ultimate Access to Information Enforcement Tool. Administration and Society, 51(1), 34–62. https://doi.org/10.1177/0095399715611171
41 Kovač P, Legal Remedies in Exercising the Right to Information: A Comparative Overview, p. 640 [in laws of transparency in action]
42 Danish Parliamentary Ombudsman (2016), Annual report 2018, https://en.ombudsmanden.dk/publications/annual_reports/ar_2016/
43 Danish Parliamentary Ombudsman (2018), Annual report 2018, https://en.ombudsmanden.dk/publications/annual_reports/ar_2018/
44 Danish Parliamentary Ombudsman (2019), Annual report 2018, https://en.ombudsmanden.dk/publications/annual_reports/annual_report_2019/
45 This dynamic is a particular problem in Czechia, where the lack of a power to Czechia leads to a ‘ping pong' loop where the superior authority can continuously ask the original authority to reconsider the decision, meaning the complaint can never be escalated to the courts (which may only happen when both authorities are agreed) - see Stanislav Kadečka, Jan Brož, and Lukáš Rothanz (2018), The Laws of Transparency in Action: Freedom of Information in the Czech Republic : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
46 Lundgren, Fred (2014), Riksdagens ombudsman, En granskning av JO:s tillsynsverksamhet och domstolars självständighet, ur ett konstitutionellt perspektiv, https://www.diva-portal.org/smash/get/diva2:798642/FULLTEXT01.pdf,p. 50
47 Hans-Gunnar Axberger (2014), JO – i riksdagens tjänst, p. 105, Available at: https://www.jo.se/Global/Bok%20JO%20-%20i%20riksdagens%20tj%c3%a4nst.pdf
48 Hans-Gunnar Axberger (2014), JO – i riksdagens tjänst, p.18-19, Available at: https://www.jo.se/Global/Bok%20JO%20-%20i%20riksdagens%20tj%c3%a4nst.pdf
49 Jägerskiöld, S. (1961). The Swedish Ombudsman. University of Pennsylvania Law Review, 109(8), 1077. https://doi.org/10.2307/3310586
50 Anamarija Musa (2018),Croatia: The Transparency Landscape : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
51 UK Law Commission, Consultation paper 187, Administrative redress: public bodies and the citizen, https://www.lawcom.gov.uk/app/uploads/2015/03/cp187_Administrative_Redress_Consultation.pdf
52 Mullin, T(2021), Internal review systems and administrative justice. In: Tomlinson, J., Thomas, R., Hertogh, M. and Kirkham, R. (eds.) The Oxford Handbook of Administrative Justice. Oxford University Press: New York. ISBN 9780190903084 (doi:10.1093/oxfordhb/9780190903084.013.4)
53 Per Hagström (2022), Myndighetens hierarki (Authority hierarchy), https://www.allmanhandling.se/myndighetens-hierarki/
54 Mullin, T(2021), Internal review systems and administrative justice. In: Tomlinson, J., Thomas, R., Hertogh, M. and Kirkham, R. (eds.) The Oxford Handbook of Administrative Justice. Oxford University Press: New York. ISBN 9780190903084 (doi:10.1093/oxfordhb/9780190903084.013.4), p. 5
55 Rumbul, R. and Parsons, A. (2021). Reforming FOI. [online] mySociety Research. Available at: https://research.mysociety.org/publications/reforming-foi .https://research.mysociety.org/html/reforming-foi/l/7.62.cqnd7.zdv8i.6njs-.y6jyl.html [Accessed 1 Mar 2022]
56 Bogdana Neamtu and Dacian C. Dragos (2018), Freedom of Information in the European Union: Legal Challenges and Practices of EU Institutions : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
57 Bogdana Neamtu and Dacian C. Dragos (2018), Freedom of Information in the European Union: Legal Challenges and Practices of EU Institutions : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
58 Office of the Scottish Information Commissioner (2018), Intervention Report, Scottish Government, para 151, https://www.gov.scot/binaries/content/documents/govscot/publications/foi-eir-release/2018/07/foi-18-01655/documents/foi-18-01655-related-documents-part-2-pdf/foi-18-01655-related-documents-part-2-pdf/govscot%3Adocument/FOI-18-01655%2B-%2Brelated%2Bdocuments%2B%2528part%2B2%2529.pdf
59 OpenDemocracy (2020), Revealed: Governments ‘Orwellian' unit blocked infected blood scandal disclosure, https://www.opendemocracy.net/en/freedom-of-information/revealed-governments-orwellian-unit-blocked-infected-blood-scandal-disclosure/
60 Background interviews C and E
61 Mullin, T(2021), Internal review systems and administrative justice. In: Tomlinson, J., Thomas, R., Hertogh, M. and Kirkham, R. (eds.) The Oxford Handbook of Administrative Justice. Oxford University Press: New York. ISBN 9780190903084 (doi:10.1093/oxfordhb/9780190903084.013.4), p. 10
62 Linda Reif (International Ombudsman Institute), quoted in Holsen, S., & Pasquier, M. (2012). Insight on Oversight: The Role of Information Commissioners in the Implementation of Access to Information Policies. Journal of Information Policy, w, 214–241.
63 Rumbul, R. and Parsons, A. (2021). Reforming FOI. [online] mySociety Research. Available at: https://research.mysociety.org/publications/reforming-foi [Accessed 1 Mar 2022].
64 Background interviews A and C.
65 La Forest, G. V. (2005). The Offices of the Information and Privacy Commissioners : The merger and related issues. Retrieved from https://www.justice.gc.ca/eng/rp-pr/csj-sjc/atip-aiprp/ip/index.html
66 Evidence before UK House of Commons Public Administration and Constitutional Affairs Committee (2021), https://committees.parliament.uk/oralevidence/3069/html/
67 La Forest, G. V. (2005). The Offices of the Information and Privacy Commissioners : The merger and related issues. Retrieved from https://www.justice.gc.ca/eng/rp-pr/csj-sjc/atip-aiprp/ip/index.html, p. 36
68 Resolution of the European Freedom of Information Commissioners and OmbudspersonsBerlin, February 24th, 2017, https://www.bfdi.bund.de/SharedDocs/Downloads/DE/IFG/AGID_IFK/2017_Treffen_EU_Infobeauftragte_EN_Resolution.pdf?__blob=publicationFile&v=4
69 Holsen, S., & Pasquier, M. (2012). Insight on Oversight: The Role of Information Commissioners in the Implementation of Access to Information Policies. Journal of Information Policy, w, 214–241.
70 Petra Lea Láncos (2018), Freedom of Information in Hungary: A Shifting Landscape : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
71 Evidence before UK House of Commons Public Administration and Constitutional Affairs Committee (2021), https://committees.parliament.uk/oralevidence/3069/html/
72 Scorer's comments on RTI Rating, France, Ireland http://www.rti-rating.org/country-data/France/, https://www.rti-rating.org/country-data/Ireland/
73 Scorer's comments on RTI Rating, Croatia http://www.rti-rating.org/country-data/France/, https://www.rti-rating.org/country-data/Croatia/
74 For instance, the statistics in Scotland separate into Freedom of Information and Environmental Information Requests. This is unusual to separate, but useful for understanding the different effects of different disclosure laws.
75 Official statistics of Sweden(2020), Court Statistics, https://www.domstol.se/globalassets/filer/gemensamt-innehall/styrning-och-riktlinjer/statistik/2021/court_statistics_2020.pdf
76 OAS (2020), Model Inter-American Law 2.0 on Access to Public Information, Available at: http://www.oas.org/en/sla/dil/docs/publication_Inter-American_Model_Law_2_0_on_Access_to_Public_Information.pdf
77 Background interview D
78 Stef Keunen and Steven Van Garsse (2018), Access to Information in Belgium : in The laws of transparency in action : a European perspective. Palgrave Macmillan.
79 Parsons, A. and Rumbul, R. (2019). Freedom of Information in Local Government. [online] mySociety Research. Available at: https://research.mysociety.org/publications/freedom-information-local-government [Accessed 1 Mar 2022].
80 German Federal Ministry of the Interior and Community (2022), Freedom of Information page, https://www.bmi.bund.de/DE/themen/moderne-verwaltung/open-government/informationsfreiheitsgesetz/informationsfreiheitsgesetz-node.html
81 Information Commissioner (Ireland) (2020), Annual report 2020 https://www.oic.ie/publications/annual-reports/OICAR2020FinalWebEN.pdf
82 Information Commissioner (Croatia) (2021), Annual report 2021 https://pristupinfo.hr/wp-content/uploads/2021/05/IC-Annual-Report-2020-Summary.docx?x57830
83 Information Commissioner (Scotland) (2021), Reflections on FOI in 2021, https://www.itspublicknowledge.info/home/News/20211123.aspx
84 German Federal Ministry of the Interior and Community (2015), Federal Commissioner for Data Protection and Freedom of Information will be strengthened, https://www.bmi.bund.de/SharedDocs/pressemitteilungen/DE/2015/12/bfdi-wird-unabhaengig.html
85 Доступ до правди (2022), Experts will discuss the draft law on personal data and the supervisory body in the field of access, https://dostup.pravda.com.ua/news/publications/eksperty-obhovoriat-proekt-zakonu-pro-personalni-dani-i-nahliadovyi-orhan-u-sferi-dostupu>>>; Oleksiy Stolyarenko (2021), Data Privacy Reform in Ukraine: What's New? <<<FLOATING LINK: https://www.connectontech.com/data-privacy-reform-in-ukraine-whats-new/
86 Kars J. de Graaf, Albert T. Marseille, and Hanna D. Tolsma (2018), Transparency and Access to Government Information in the Netherlands : in The laws of transparency in action : a European perspective. Palgrave Macmillan.; Background interview B.