76254 IMPLEMENTING RIGHT TO INFORMATION Lessons from Experience ANUPAMA DOKENIYA IMPLEMENTING RIGHT TO INFORMATION Lessons from Experience Copyright © 2013 The International Bank for Reconstruction and Development / The World Bank Group 1818 H Street, NW Washington, DC 20433, USA All rights reserved. First printing: February 2013 Manufactured in the United States of America. Cover photo credit: Curt Carnemark, World Bank The findings, interpretations, and conclusions expressed in this report are entirely those of the authors and should not be attributed in any manner to the World Bank, or its affiliated organizations, or to members of its board of executive directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. Contents Abbreviations and Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Preface and Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii 1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 Background and Objectives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 1.2 Assessing Effectiveness and Impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 1.3 Explaining Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 1.4 Organization of the Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7 2 Institutions and Capacity for Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.1 Why Implementation Institutions are Important . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9 2.2 Designating and Sustaining Oversight Agencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10 2.3 Organizational Arrangements in Implementing Agencies . . . . . . . . . . . . . . . . . . . . . . . . . . . .12 2.4 Request Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14 2.5 Performance Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 2.6 Records Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17 2.7 Proactive Disclosure and Open Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18 2.8 Political Will for Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18 3 Appeals and Enforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.1 Rationale for Information Commissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19 3.2 Record on Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20 3.3 Ensuring Independence and Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21 3.4 Sanctions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21 4 Enabling Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.1 Signficance of the Enabling Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 4.2 Capacity and Influence of Civil Society and Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24 4.3 Check-and-Balance Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26 4.4 Broader Regulatory Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26 4.5 Implications for Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26 5 Conclusion: Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 iii CONTENTS Annexes 1 Global Developments in Transparency and Openness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33 2 Observations on the Passage of RTI Reforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35 3 Ratings of RTI Laws by the RTI Ratings Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43 4 Scope of RTI Legislation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45 5 Methodological Issues in Assessing RTI Effectiveness and Impact . . . . . . . . . . . . . . . . . . . . . .49 6 Findings on Use and Compliance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55 7 Examples of Use of RTI for Governance and Service Delivery . . . . . . . . . . . . . . . . . . . . . . . . .61 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Boxes 2.1 RTI Technology Platforms in Mexico . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 A.5.1 Evidence from Research and Studies on Impact of Information . . . . . . . . . . . . . . . . . . . . . . . . . .52 Figures 1.1 Chronology of Adoption of ATI Laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 1.2 Implementation Institutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 2.1 Proportion of Requests Received per Monitored Body—United Kingdom and Mexico . . . . . . .13 3.1 Typical Appeals Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19 4.1 Worldwide Governance Indicators—Rule of Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 4.2 Civil Liberties, Political Rights, and Voice and Accountability . . . . . . . . . . . . . . . . . . . . . . . . . .24 A.1.1 Transparency and Access to Information at the World Bank . . . . . . . . . . . . . . . . . . . . . . . . . . . .34 A3.1 RTI Ratings Scores for Case Study Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43 A.5.1 RTI Results Chain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50 Tables 1.1 Case Study Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 1.2 Requests and Responsiveness in Selected Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 2.1 Institutional Arrangements for Oversight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10 iv Abbreviations and Acronyms ATIA Access to Information Act CIC Central Information Commission CSO civil society organization DoPT Department of Personnel and Training EITI Extractive Industries Transparency Initiative FOIA Freedom of Information Act GAC Governance and Anticorruption HURINET-U Human Rights Network Uganda ICO Information Commission IFAI Instituto Federal de Acceso a la Información Pública (Federal Institute for Access to Information) IPP Institutul pentru Politici Publice MDA ministries, departments, and agencies MKSS Mazdoor Kisan Shakti Sangathan NCPRI National Campaign for People’s Right to Information Campaign NGO nongovernmental organization PETS Public Expenditure Tracking Surveys PIO public information officer POT Portal de Transparencia (Transparency Portal) RaaG RTI Assessment and Analytical Study RTI right-to-information SFP Secretary of Public Function v Preface and Acknowledgements The Bank’s 2007 Governance and Anticorruption Yamini Aiyar and Mandakini Devasher (India), Marcos (GAC) strategy highlighted that improving governance Mendiburu and Yemile Mizrahi (Mexico), Sergiu Lipcean in countries was necessary to fight corruption and and Laura Stefan (Moldova), Robert Pereira Chumbe strengthen development, and underlined the importance (Peru), Sorin Ionita and Laura Stefan (Romania), of transparency as a critical dimension of good governance. Anupama Dokeniya (Uganda), and Tom McClean (United It pointed out that citizens and media that have access to Kingdom). Robert Beschel provided invaluable advice information on the operation of state institutions are crucial to developing the final report. Lisa Bhansali, Marcos for fostering accountability. Transparency might extend to Mendiburu Sahr Kpundeh, Vivek Srivastava, Francesca publication of budget and procurement data, access to state Recanatini, Gary Reid, and Graham Teskey provided records and reports, and the state’s active dissemination significant advice and inputs during the development of the of information on its operations and performance. The project. Chris Finch, Verena Fritz, Helene Grandvoinnet, strategy pointed out that building on a growing track record Guenter Heidenhof, and Jana Kunicova reviewed earlier of success in this area, the Bank will scale up its work with drafts and provided very useful comments. Toby Mendel interested governments to strengthen transparency in and Richard Calland provided expert inputs on specific public policy-making and service provision. This project topics. Tammar Berger, Sabina Schnell, and Kinga Krisko is part of the effort to scale up knowledge development provided research assistance. Audrey Liounis edited various on transparency and access to information issues. The sections of the report. Diana Obrero and Laryssa Chiu project focuses specifically on the implementation of Right helped with logistical issues throughout the project. Laura to Information laws—laws that establish citizens’ right to Johnson provided editorial, design, and layout assistance. have maximum access to government documents in all Anupama Dokeniya led the preparation of the report. areas, barring a limited regime of exceptions. She can be reached at: adokeniya@worldbank.org. The The discussion in this report is based on research study was carried out with funds from the Governance undertaken in eight countries. The case studies were Partnership Facility. The case studies are available at: prepared by: Jolanda Trebicka and Gerti Shella (Albania), http://www.worldbank.org/publicsector/gpa/transparency. vii 1 . Overview 1.1. Background and Objectives Of course, the international momentum translated into law when domestic dynamics were favorable, when politi- Over the last two decades, the number of countries with cal elites perceived that it is to their advantage to support right-to-information (RTI) laws—laws that establish citi- the law in order to win political points with domestic con- zens’ “right� to have access to public information or op- stituencies and establish their democratic credentials inter- erationalize such a right found in the constitution—has nationally. In several countries, the transition to democracy exploded (figure 1.1). In 1990, only 13 countries had RTI also provided an opportunity when pro-reform coalitions laws,1 all of them western liberal democracies. By 2012, this of ruling and opposition parties, civil society groups, and number had risen to more than 90.2 And most of the new media obviated sources of opposition or resistance to the adopters are countries in Eastern Europe, Asia, Latin Amer- passage of an RTI law (annex 2). ica, and most recently, Africa and the Middle East—coun- As more countries have adopted the law, the right to tries with diverse political histories, difficult governance information has been cast in fairly ambitious terms by both environments, and persistent development challenges. its international proponents and its domestic champions. It Both international and local dynamics have contrib- has been characterized as fundamental to enhancing citi- uted to this trend. Internationally, the combination of a zen participation in governance, improving citizens’ un- growing global emphasis on transparency as critical to good derstanding of public policy choices and decision-making governance, the recognition of the right in international processes, and enabling them to assert claims on service conventions,3 its championship by international policy entitlements, scrutinize public officials and public expen- networks and epistemic communities,4 and pressure from ditures,6 and exercise a more direct form of social account- development aid agencies and intergovernmental bodies ability.7 The Indian RTI law passed in 2005, for instance, created a global momentum for reform (annexes 1 and 2). was characterized as “a great and revolutionary law,�8 one The influence of international networks that links global with the potential of “fundamentally altering the balance experts and nongovernmental organizations (NGOs)5 with of power between the government and citizens.�9 local policy makers has also been an important element in But because RTI laws are relatively new in most coun- the convergence of legislative standards across countries tries and because their adoption has been a difficult and (annex 3 and 4). contested process, much of the research in this area has focused on analyzing the passage of legislation and on the Figure 1.1. Chronology of Adoption of ATI Laws comparison of the provisions in pieces of legislation against 90 global good practices.10 There is little empirical research on 80 how these provisions have worked in practice, especially number of countries 70 in developing country contexts, whether or not they have 60 50 been adequately enforced, and if they have been effective 40 in fulfilling their stated goals of improving transparency, 30 accountability, good governance, and service delivery. The 20 research project on which this report is based attempts to 10 0 address this gap. 1950 1960 1970 1980 1990 2000 2010 Source: Based on data from Vleugels 2009. 1 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE It looks at two broad sets of questions on the implemen- different regions, and represent a range of income levels tation of RTI. and varying political and administrative traditions (table 1.1). In each country, local researchers were commissioned • The first set of questions relates to the effectiveness to collect data on the use of RTI law by citizens as a tool for and impact of RTI laws. What happens after the law requesting information and the responsiveness of officials is passed? Does it actually enable citizens to gain ac- to such requests. In some countries, this data is monitored cess to information? What types of information? What by government agencies; and civil society groups have also evidence exists of the broader impact of RTI laws (in launched surveys and studies to assess the use of and com- strengthening accountability of public officials, anticor- pliance with RTI. Additionally, researchers surveyed the ruption, and so on)? secondary literature on the implementation of RTI in the • The second set of questions relates to the factors that case study countries, reviewed official documents, and con- might explain how effectively the law works. In coun- ducted in-depth interviews with a range of stakeholders, tries where there is evidence of the law working well, including officials, civil society, and media groups. Details what factors might explain this? In countries where it is on the methodology for addressing the key questions are in not being used, what constraints and challenges can be annex 5. identified? This report provides a concise summary of the key The project looked at these issues in eight coun- lessons vis-à-vis implementation emerging from the case tries—Albania, India, Mexico, Moldova, Peru, Romania, studies. It is intended to draw lessons from the experience Uganda, and the United Kingdom. These countries have of these countries as a guide to policymakers and officials all had RTI laws in place for five years or more, span across charged with implementing RTI in countries that have Table 1.1. Case Study Countries Country Income classification Name of RTI law Year Context adopted Albania Lower-middle Law 8503—Right to 1999 Post-communist transition; income Information about influence of European institutions Official Documents India Lower-middle Right to Information Act 2005 Congress Party came to power on income an electoral platform promising RTI Mexico Upper-middle Transparency and Access 2002 Transformational regime income to Public Information change after 71 years of Partido Revolucionario Institucional rule Moldova Lower-middle Law 982 on Access to 2000 Post-communist transition; income Information attempt by new political forces to be seen as progressive and break from the past Peru Upper-middle Law on Public Access to 2001 Political transition post-Fujimori income Information Romania Upper-middle Law 544 on Access to 2001 Post-communist transition; income Public Information pressure of EU accession Uganda Low income Access to Information 2005 Attempt by ruling NRM to Act maintain a progressive image for international donors, and domestically faced with elections United High income: OECD Freedom of Information 2000 Labor Party came to power after Kingdom Act several years Source: Compiled by author based on case studies. 2 1. OVERVIEW recently adopted legislation or that are working toward it. Although as a percentage of the population, the total Although the number of countries under review is small, number of requests is very small (0.03–0.4),15 the number their experience offers useful lessons on the potential chal- is not significantly higher even in developed countries lenges and constraints in setting in place the institutional with several years of experience on RTI and ranges from changes, capacity enhancements, and normative shifts nec- 0.3–0.4 percent,16 leading to RTI being characterized as an essary for RTI laws to function effectively. Policy makers “extraordinary tool by definition.�17 In the countries where need to be cognizant of, and make concerted and explicit data was available, it also showed that average compliance efforts to address, these challenges in the course of imple- rates—the percentage of RTI requests that received a re- mentation if the right to information is to become a reality. sponse—ranged from 75–90 percent, suggesting that RTI had proved a useful tool for many. The next section briefly outlines the key findings on the relative effectiveness and impact of RTI in the case study Although requests and responsiveness statistics provide countries. The subsequent section and the following chap- a useful snapshot of the overall functioning of the RTI re- ters explore the key dimensions that have a bearing on the gime, they also have several limitations that are discussed relative effectiveness of the law. in annex 5. For instance, they do not provide sufficient dis- aggregation on the kinds of information that are refused. Requests for information on the government’s function- 1.2. Assessing Effectiveness and ing, decision-making, expenditures, and contracts—in- Impact formation particularly critical to holding the government accountable—might be precisely the requests that are re- The performance of an RTI regime is usually measured by fused. Anecdotal data about specific instances of informa- assessing the extent to which citizens exercised the right tion requests and responsiveness were therefore collected to information (through the number of annual requests), to assess if RTI was successful in accessing this more “sensi- and whether the existence of a legally-mandated right tive� information. propelled officials to disclose information (through statis- tics of fulfillment versus denial of requests).11 Although In India,18 Mexico, Romania, and the United King- implementing and oversight agencies are usually required dom,19 media and NGOs also had several successes in using to track and record these statistics, in three of the coun- RTI to elicit information on issues related to mismanage- tries studied—Albania, Moldova, and Uganda—no official ment of public funds; nonperformance of service-delivery data could be accessed. But data from interviews and civil agencies; and instances of fraud, corruption, or nepotism. society organization (CSO) surveys suggests that both the More details on instances in which the use of RTI led to use of RTI and the responsiveness by officials is very low. pushing the envelope on transparency in these countries In Uganda, the 2005 Access to Information Act (ATIA) can be found in annex 7. is not being used by the majority of citizens, and requests The project also looked at anecdotal data to assess if by CSOs, were mostly without success.12 In Albania and there is evidence that RTI could have an impact on im- Moldova, interviews suggested that although the respective proving broader governance outcomes, such as officials RTI laws are being used, both interviews and civil society being sanctioned when corruption was exposed, corrective surveys suggest that information requests are often ignored measures being taken when poor performance was brought or responded to with considerable delay or in incomplete to light, better safeguards being established to improve gov- formats.13 ernance and prevent corruption, firms being disqualified In the other case study countries, more systematic data when collusion was exposed, and service providers being was available. In these countries, the significant number held accountable for the efficient delivery of services. A of popular requests (table 1.2) as well as anecdotal data few instances of RTI disclosures resulting in such correc- (annex 7) suggests that RTI has been very useful to citizens tive actions emerged, but there were also several instances gaining access to information. 14 in which the exposure of information did not lead to any 3 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE Table 1.2. Requests and Responsiveness in Selected Countries Percentage Percentage Number of requests per yer (rounded of receiving Country to nearest 1,000) population response Level of government Albania Data not available n/a n/a n/a India 250,000 (2007–08) .03 81 Central (increased to 625,000 in 2011–12) Mexico Increase from 24,000 in 2003 to more .04 96.9 Central than 120,000 in 2012 Moldova n/a n/a n/a n/a Peru Increase from 56,000 in 2004 to n/a 90 Central 68,000 in 2010 Romania 600,000 (2002–09) 3.2 95–98 Central and regional Uganda Data not available n/a n/a n/a United Kingdom 16,000 in 2006 to 44,000 in 2010 .04 75–90 Central Notes: n/a = not available. It is important to note that a negative response—that is, a denial of the information request—also counts in these statistics as a response. Sources: IFAI 2009; Central Information Commission 2006; Department of Government Services, Romania; Constitution Unit 2011; PCM, Peru; Popula- tion Reference Bureau 2011. Population data figures used are from mid-2011. follow-up action—neither the punishment of officials nor 1.3. Explaining Outcomes any systemic improvements. In the short term, access to information might also lead In-depth interviews with a range of stakeholders and an to more mistrust in government and higher perceptions of extensive survey of the existing research and empirical data corruption as information about corruption that was previ- enabled the identification of two broad dimensions that ously hidden is put into the public space. This might ex- have a bearing on the effectiveness of an RTI law both to plain the conclusions from a study that found a negative elicit information of a personal interest and to leverage RTI correlation between the existence of RTI laws and percep- for accountability outcomes. tions of corruption.20 In India, for instance, high-profile • First, setting up the formal institutional architecture for corruption scandals, several exposed through RTI requests implementation was important to build capacity in the and followed by aggressive media, have created a wide- public sector to respond to RTI requests, aligning in- spread perception of massive corruption. But these exposés centives and creating a culture of openness. have also galvanized a popular movement directed against corruption and for setting in place stronger accountability • Second, features of the underlying governance and po- mechanisms. Although this is an area that still needs much litical economy environment—the willingness of offi- research, the Indian example might point to ways in which cials to follow the rule of law, the responsiveness of the RTI can have a broader systemic impact, possibly through state to civil society demands for information, the role the mobilization of voices against corruption to create of complementary check-and-balance institutions to more systemic reforms in the accountability environment. demand follow-up action and maintain vigilance—all had a bearing on the effectiveness of RTI. 4 1. OVERVIEW Establishing formal institutions was critical to important to provide redress against noncompliance, on signaling political commitment, creating aware- both more routine cases, and on specific more high profile ness and support for the law, and building cases. capacity in the public sector to comply with the On the other hand, where these formal institutions mandate for disclosure. for implementation were not established, or where they The key institutions, actors, and relationships relevant were not provided with adequate resources or political sup- to the implementation of RTI are depicted in Figure 1.1. port, capacity-building efforts were incomplete and frag- Where specific agencies for oversight and promotion of the mented and there was little evidence that RTI had made law were designated and provided adequate resources and any difference to the environment for transparency in the political support, they were important in driving implemen- government. However, the case studies also showed that tation, overseeing and supporting efforts by implementing sustainability can be challenging even in countries that es- agencies, framing guidelines and regulations, ensuring that tablish robust implementation and oversight mechanisms, adequate arrangements for responding to requests were once the initial momentum behind the law subsides. In in place, promoting awareness about the law within the Romania, for instance, budget constraints led to the pro- government and among the public, and supporting tech- gressive marginalization of the nodal agency charged with nological and human resource capacity. Both independent overseeing implementation. oversight agencies (such as information commissions) and nodal agencies within the government each played an important part in promotional, capacity-building, and Undoubtedly, in many countries, budget limitations will oversight efforts. Some countries implemented technology influence feasible options for implementation. Although systems for registering requests and monitoring compliance exact costs of implementation have been very hard to as- that also provided a useful interface for lodging requests to sess, additional budget is required for oversight agencies, for citizens. The establishment of independent commissions or additional personnel in implementing agencies, training of tribunals, with adequate capacity, autonomy, resources, was information officers, awareness-raising, monitoring, and Figure 1.2. Implementation Institutions Oversight and Implementing Appeals and Capacity-building Promotion Agencies Enforcement Agencies Oversight Ministries Rulings, Agencies Nodal Agency Capacity Sanctions Information Commission Departments Archives Agencies Ombudman Training Institutes Other Institutions Judiciary Information Commissions Covered by Law Disclosure Proactive Responses Requests, Appeals Citizens, Civil Society Source: Author. 5 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE technologies for information management, and for more The underlying political economy relationships structural improvements such as records management sys- and governance environment influenced the tems. An incremental approach to capacity building across effectiveness of RTI—for instance, determining agencies could make implementation more feasible and if officials followed the rule of law, how blunt the criticisms of RTI being a drain on resources. For responsive the state was to civil society pres- instance, in the short term, it might be useful to focus ef- sures, and if other check-and-balance institu- forts on building the capacity for information management tions maintained vigilance over the process. in agencies that have high demand. For cost-intensive The influence of underlying political economy relation- areas, such as records management, actions that might be ships and governance characteristics on RTI outcomes has have the largest payoffs and address the most challenging rarely been assessed. Some commentators23 have pointed constraints could be prioritized, rather than expecting a to the importance of this dimension. For instance, Kriemer full-fledged rollout of advanced records management sys- refers to the “ecology of transparency�24 comprised of dedi- tems in the short term.21 cated oversight and monitoring groups that routinely mon- Both the failure to put in place implementation and itor the efficacy of RTI, accountability and media groups oversight mechanisms, and challenges to their sustain- that use RTI to uncover information on corruption or non- ability indicates the erosion of political will during the performance and demand corrective actions, and media implementation phase. The passage of the law is often a and press laws that encourage free expression and free as- high-profile effort by its political champions, undertaken sociation, as essential to the effective functioning of RTI. to establish democratic credentials and gain support and The experience of the case study countries suggest that legitimacy both domestically and with international part- implementing RTI in countries with a more difficult record ners (annex 2). The implementation phase rarely embodies on key governance dimensions is challenging. Countries the same kind of enthusiasm and political dividends, so as that ranked higher in terms of certain indices of gover- political will dissipates, so do the incentives of administra- nance—such as rule of law, voice and accountability, and tive officials to implement the mandate of the law. As a civil liberties, were the countries where the record on im- result an “implementation gap� has characterized de jure plementation was better. In these countries, more robust reforms in several areas, not only RTI.22 Moreover, as politi- data on the use and compliance with RTI was available, cal momentum behind the legislation erodes, both political where civil society and media advocates were able to use actors and public officials have also employed a variety of RTI to extract information about government functions, strategies to blunt the effectiveness of the law. Resistance public expenditures, procurement, performance, and so to implementation might come in the form of attempts to on—areas critical to ensuring accountability. create amendments to the scope of the law, refusal to help requesters, or even outright harassment. Sustaining politi- The limited anecdotal data that could be collected also cal will for implementation might, therefore, be the most points to the importance of the characteristics of the under- momentous challenge of RTI, and one that has to be kept lying political economy and governance environment in on the forefront of implementation efforts. determining the ability of RTI to translate into “account- ability outcomes� and broader governance improvements, such as if corruption and nonperformance were punished and if safeguards against corruption or oversight systems for performance were put in place. Some of the factors identi- fied include the capacity of the legislature, judiciary, and other check-and-balance institutions such as supreme audit 6 1. OVERVIEW institutions to follow through on disclosures, the extent to 1.4. Organization of the Report which the rule of law is followed, and the extent to which the media keeps pressure on follow-up action once infor- The following chapters discuss the key challenges and ob- mation is disclosed. stacles that emerged in the case study countries vis-à-vis Why civil society efforts had different results in dif- implementation. Chapters 2 and 3 discuss the lessons from ferent countries, why they succeeded or failed in different the case study countries on setting up formal institutions contexts—in other words, the political economy of state- for implementation and capacity-enhancing measures. society relations and the factors that propel state actors to Chapter 2 focuses on implementation and capacity within respond to civil society pressure—are important areas for the public sector. Chapter 3 focuses on appeals and en- further analysis. Although the scope of the study is limited, forcement mechanisms set up to provide redress for denial the experience of the case study countries does highlight of information. Chapter 4 assesses the role of the broader that factoring in the constraints and limitations of the un- political economy environment. Chapter 5 summarizes the derlying political economy relationships and governance key findings. The note only highlights some of the key is- environment is critical, and RTI reforms need to be part sues that emerged from the case study countries. More de- of a more comprehensive, consolidated set of reforms that tails on how these issues were addressed in these countries address these underlying dynamics. and on their more broad implementation experience can be found in the detailed case studies available electronically.25 7 2 . Institutions and Capacity for Implementation 2.1. Why Implementation commit to protecting state secrets. RTI laws presume that Institutions are Important the information held by the state belongs to the people who have an intrinsic right to access it. An RTI law gives Addressing the supply-side of transparency—both capac- citizens26 the right to access government records without ity as well as administrative culture and incentives within being obliged to demonstrate a legal interest or standing— the public sector—requires concerted efforts. Unlike other a fundamental shift in the principle guiding access to pub- transparency measures, the scope of RTI extends across lic information from “need to know� to “right to know.� multiple sectors and various levels of the administration. Global standards on RTI have also converged on the prin- Capacity for responding to RTI requests, proactively releas- ciple of “maximum disclosure�—that is, the right of access ing information, and monitoring statistics of requests and should extend to a broad range of government documents responsiveness needs to be strengthened in MDAs across unless specifically included in a narrowly-defined list of ex- the administration as well as in other entities or arms of emptions. Under an RTI law, therefore, a presumption of government subject to the law, which might include legis- disclosure and transparency supersedes the presumption of latures, judicial bodies, state-owned enterprises, and other secrecy that has traditionally characterized the functioning government bodies. This requires considerable effort and of government.27 This is a fundamental paradigm shift, and investment to boost both technical and human resource effective implementation requires fostering a new culture capacity. Information systems and archival practices must of openness. be upgraded and records digitized so that information can Building support and capacity for RTI implementation be retrieved within mandated timeframes. Staff capacity is also a long-term exercise rather than a short-term activity. and training is required to improve information manage- During political and administrative transitions, especially ment and dissemination skills. Organizational units must in settings where a large number of public officials might be be set up to respond to requests. Formal institutions over- political appointees, political change at the top might lead seeing capacity-building are important for these efforts. to changes in other parts of the administration, necessitat- The mismatch between the demands of openness and ing training new cohorts of officials, rebuilding a commit- the inherent incentives and bureaucratic culture in the ment to RTI, and renewing organizational structures will public sector also need to be addressed. Control of infor- require sustained attention, support, and resources. mation translates into rents for public officials when, for The case studies showed that dedicated institutions fo- instance, bribes have to be paid to get information or when cused on strengthening the supply side were very important corruption can continue unchecked in the absence of in promoting and monitoring compliance with the law and transparency. Relinquishing control could lower discretion, supporting implementation measures, framing guidelines lessen the avenues for corruption, and cut into rents by re- and regulations for implementation, promoting awareness ducing opportunities for bribery, patronage, and kickbacks, about the law within the government and among the pub- creating more open competition in procurement contracts, lic, and supporting technological and human resource ca- and greater civil society monitoring of income and assets. pacity across government. Where some of the key elements Public sectors have traditionally been characterized by of the institutional architecture were not set up, there was principles of secrecy and control—principles often codified inadequate attention to capacity development and promo- in a legacy of secrecy laws that require public servants to tion of the law. The next sections briefly discuss the key 9 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE dimensions of the institutional architecture and the lessons might not specify an oversight agency within the govern- from the experience of the case study countries. ment, India, Mexico, the United Kingdom, Uganda, and Romania designated particular government departments as focal points for implementation. In India, Mexico, and 2.2. Designating and Sustaining the United Kingdom, these agencies complemented strong Oversight Agencies independent oversight institutions. This combination of an independent oversight institution and a nodal agency In countries where oversight agencies with sufficient ca- within the government proved most effective because of pacity and political support were set up, they provided an the specific strengths of each body vis-à-vis implementa- important focal point for capacity-building efforts. Inde- tion, as discussed below. On the other hand, the absence pendent oversight agencies— independent information of an independent agency in Uganda and Moldova meant commissions or ombudsmen—often designated by the RTI that there was no independent champion of the law. And law itself and charged with both implementation oversight the absence of clear focal points for implementation within and to serve as an agency for appeals against noncompli- the governments in Moldova and Albania led to fragmen- ance are considered best practice. However, implementa- tation of efforts and poor incentives for compliance. tion experience showed that designating an entity within In the countries that adopted a mixed model, indepen- the government for oversight and promotion and coordi- dent oversight agencies were important as champions of nating change management is also important, both dur- implementation. At the same time, a nodal agency within ing the initial years of implementation as well as over the the government was essential for systemic change manage- long term to ensure the sustainability of implementation ment. In Mexico, for instance, the role of the Secretary of measures. Public Function (SFP), the agency responsible for public The various models adopted by the countries under administration issues, was particularly critical during the review are outlined in table 2.1. Even though the law initial stages of the law’s implementation, when liaison Table 2.1. Institutional Arrangements for Oversight Country Executive oversight body? Independent oversight body? Albania No Ombudsman; People’s Advocate—charged by RTI law with implementation; functions determined by complementary law India Department of Personnel and Training (DoPT) (de Central Information Commission (CIC); State facto) Information Commissions Mexico Secretary for Public Function (SFP) Federal Institute for Access to Information (IFAI) Moldova No No, although Article 21 says that “person can turn to Ombudsman for the protection of his or her legal rights,�But in practice, no oversight agency identified Peru Office of the President of the Ministerial Cabinet Ombudsman Romania Ministries of Public Information, Communications No & IT, and Ministry of Public Finance tasked with developing IT for public information (Article 24) Oversight function now under General Secretariat of Government Uganda Directorate of Information and National Guidance No (de facto) United Department of Justice Data Protection Commission rechristened Kingdom Information Commission (ICO), and its mandate expanded to include oversight and appeals under 2000 FOIA Source: Author compiled from case studies 10 2. INSTITUTIONS AND CAPACITY FOR IMPLEMENTATION units and information committees had to be set up across senior public officials, and the consequent weakening of approximately 250 federal agencies, and thousands of fed- administrative capacity and training systems. This eroded eral public officials needed to be trained. At the same time, capacity and skills and has created risk aversion among of- the Federal Institute for Access to Information (IFAI), re- ficials. Although the Albanian ombudsman, the People’s garded as the gold standard for information commissions, Advocate, charged with handling appeals under Law 8503, aggressively promoted access by disseminating information served an important function in being an advocate and about the law at the state level, persuading governors to champion of stronger implementation, pushing for several pass similar laws, encouraging citizens to make information implementation measures, attempting to get regulations requests, and disseminating information about the Mexi- passed, and drafting amendments to make disciplinary mea- can experience internationally. sures under the law stronger, it was not successful in these efforts because of a lack of resources and political support. In India, the Department of Personnel and Training (DoPT) in the Ministry of Personnel, Public Grievances, In Uganda, the Directorate for Information and Na- and Pensions, the agency with overall responsibility for the tional Guidance, housed in the Office of the Prime Min- civil service, was charged with overseeing implementation ister—purportedly to signal its significance—and staffed in the central government. DoPT took several measures to with progressive officials, launched training, monitored the push for compliance with the law, issuing detailed notifica- appointment of information officers, and developed an im- tions, supporting training, and issuing orders to ministries plementation plan. But the absence of any significant bud- to appoint public information officers, proactively disclose get being allocated to it reflected the absence of political information, and improve their records-management prac- commitment. The absence of implementation regulations, tices.28 On the other hand, information commissions have in particular, proved to be a significant stumbling block to- pushed for compliance29 even when DoPT demonstrated ward implementation for several years because it resulted ambivalence to the law.30 in the absence of a clear, detailed set of guidelines and was used as a pretext for noncompliance. The absence of the In the United Kingdom, the Information Commis- regulations was also used as an excuse for not implementing sion (ICO) has worked with authorities to improve their the law in practice.33 internal handling processes; maintained comprehensive documentation, guidance, and model publication schemes; However, the experience of Romania showed that sus- supported research; and conducted public information cam- taining the oversight function over a period of time can also paigns. It was also empowered to undertake “own motion� be challenging. As RTI is institutionalized and as proactive reviews of implementation and report to the parliament on disclosure and open data activities make more information issues of concern—such as surveillance and privacy—on available in the public space, oversight functions might an ad hoc basis. The nodal authority within the govern- become lighter. This has happened, for instance, with the ment31 has coordinated the effort among public authorities information clearinghouse in the United Kingdom, whose to put in place appropriate administrative procedures and role has shrunk as MDAs became more experienced and records-management processes, assisted public authorities the processes for handing requests, and complaints under in preparing for the law, and publicized good practices.32 the 2000 FOIA became embedded in normal operating The Access to Information Central Clearinghouse has procedures. However, it has continued to provide expertise been charged with ensuring consistent application of the and coordination for sensitive requests and for handling Data Protection Act, FOIA, and Environmental Informa- cases referred to the commissioner, tribunals, and courts. tion Regulations. In Romania, on the other hand, bureaucratic restructur- In Albania, on the other hand, the absence of an agency ing and declining budgets led to the marginalization of the within the government to oversee RTI led to fragmented departments charged with RTI oversight. Initially, the min- efforts and prevented the rebuilding of capacity in the face istry of public information charged with implementation of successive changes in government, the reorganization mobilized resources and ensured that information officers or abolition of ministries, the redeployment or removal of were appointed. These initial efforts played an important 11 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE part in instituting some of the basic capacity for implemen- them. On the other hand, departments like Public Works, tation. When the ministry of public information was dis- which are heavily engaged in the implementation of pro- solved, another agency took over its functions, and since grams and receive many requests, set up more formalized 2009, a modestly-staffed and poorly-resourced Department systems, a higher number of information officers, and dedi- of Government Strategies in the government’s Secretariat cated and well-staffed RTI cells. General oversees RTI. The consequences of this erosion in MDAs and other agencies, of course, have information capacity are obvious in the decline in the quality of perfor- departments that already perform the functions of provid- mance data from implementing agencies.34 ing information to the public, and RTI functions can be added to their terms of reference. In fact, it is impractical to expect all agencies to set up new response units, as they are rarely given additional resources to fulfill the RTI mandates 2.3. Organizational Arrangements and are expected to find resources from existing budgets or in Implementing Agencies by charging fees for providing information. Setting up organizational structures for responding to re- The MDAs with the bulk of the popular requests typi- quests and accelerating proactive disclosure efforts is the cally tend to be those that cater to information related to centerpiece of RTI implementation. But this was challeng- personal entitlements and records. In Mexico, for instance, ing in some of the case study countries. In Uganda, for in- the largest number of requests is directed to the social secu- stance, only a few ministries responded with details about rity institute, and in the United Kingdom to the health and the RTI officials they had designated in response to a survey safety executive (figure 2.1).35 These agencies might need by the Directorate of Information and Guidance in 2009. to expand their information management and dissemina- Subsequently, however, there has not been any follow-up, tion cells and hire additional staff focused on implementing and discussions with individual ministries showed that RTI RTI. Other departments where RTI does not create signifi- response structures were haphazard and ad hoc—only a few cant additional demands could co-locate the handling of ministries had designated information officers for RTI re- RTI requests with other functions, such as complaints-han- quests, there was no systematic training, and most officials dling. The key immediate measure in such cases could be were not even aware of the law. the retraining of existing officials charged with information management and dissemination functions, especially to In other countries, specific functions of the ministry, improve understanding of their obligations under the law. department, or agency; the extent of the demand from the public; and the limitation of resources influenced the size of If existing liaison units are given additional responsi- the information management cell and the organization of bilities, it is critical that they be trained in the specifics of RTI. Information officers have typically tended to be public the RTI function. In the United Kingdom, where agencies relations officers for the government; they do not see them- were given considerable autonomy in determining their ar- selves as obligated to release information at the request of rangements, smaller departments have tended to delegate the public.36 RTI to their legal teams or corporate services divisions, while larger departments and those receiving a high vol- Further, although RTI laws are primarily intended to ume of requests set up dedicated teams. In India, depart- democratize access, in several countries, access to informa- ments focused more on formulating policies at the central tion continued to be dependent on participation in privi- level—such as the Department of Rural Development and leged networks—information requests were often handled of the Department of School Education and Literacy—set informally and through personal contacts. In Uganda, for up leaner implementation structures with fewer informa- example, several service-delivery NGOs that participate tion officers, primarily because most of their programs are in policy committees within the administration have been implemented at the state- and local-government levels, able to access information easily while NGOs seen as more and requests for information are most often transferred to oppositional and focused on governance and accountability 12 2. INSTITUTIONS AND CAPACITY FOR IMPLEMENTATION Figure 2.1. Proportion of Requests Received per Monitored Body—United Kingdom and Mexico a. United Kingdom b. Mexico Health and Safety Executive Mexican Social Security Institute Ministry of Justice Secretary of Education Dept. for Work and Pensions Home Office Secretary of FInance National Archives Secretary of Health Ministry of Defence Dept. for Transport Secretary of the Environment HM Treasury Institute for Social Security Dept. of Health Secretary of Communications HM Revenue and Customs Cabinet Office Secretary of the Public Function Dept. for Business and Innovation Secretary of the Interior Dept. for Education All other monitored bodies Office of the Attorney General 0 2 4 6 8 10 12 14 16 0 20 40 60 80 100 120 percentage (thousands) Sources: United Kingdom—Ministry of Justice 2011; Mexico—IFAI. were often denied it completely.37 Signaling by the leader- • In Albania, a 2003 survey by a local NGO found that ship on the importance of the law and modeling openness 87 percent of public employees were not aware of the can help address these attitudes. lawm and very few institutions had appointed informa- tion officers.41 Leadership commitment to RTI can be evaluated by the profile that liaison units are given within the agency, the • In Mexico, a 2007 study conducted seven years after the passage of a high-profile law and the creation of an reporting lines of authority inside the agency, and the pro- oversight agency with a strong promotional program file of officials in the liaison units and information commit- found that 60 percent of public servants still believe tees. In Mexico, for example, the Secretary of Education that they own the information they generate.42 appointed one of his closest collaborators and a high-rank- ing official to the information committee,38 and devoted • In Romania, many publically-funded entities, such as resources to train the department’s staff. public schools and hospitals, were still not aware of their proactive disclosure obligations under the law, ten Training and awareness-raising efforts to create under- years after its passage.43 standing about obligations under the law are also critical to Amendments to a law or to internal policies and pro- begin to shift the mindsets with which officials approach cedures as well as modifications to implementation plans information sharing, but these efforts were inadequate in also require continuity in the training of information of- many countries: ficers and other officials. In the United Kingdom, for ex- • In Moldova, an NGO report looking at 95 public bodies ample, significant training efforts occurred in preparation demonstrated a lack of awareness among public officials for the incremental introduction of the FOIA, including about the law’s provisions.39 several roadshows by ministers of departments and agen- cies to emphasize their support for the FOIA. The shift • In India, 30 percent of rural public information officers to a “big bang� approach resulted in a loss of expertise be- polled in a survey did not know the provisions of the cause officials were redeployed to other positions, requiring RTI law. 40 retraining. 13 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE There was little evidence that efforts to harmonize for responsiveness such as rewards and recognition and en- broader systems and processes with the RTI mandate were suring that additional responsibilities are accompanied by being undertaken, inevitably leading to delays in responses. commensurate remuneration could be useful, but there was Training efforts tended to be narrowly focused on infor- little evidence that this was being done. mation officers, although fulfillment of RTI requests en- In some instances, robust internal appeals systems tails the involvement of officials across the implementing helped improve compliance. Initial nonresponsiveness agency. Information officers typically act as dispatchers of might arise from an inadequate understanding of the obli- requests to the relevant units within the department that gations under the law or from a lack of confidence among then has to make the information available. Therefore, in- the lower-level officials to release information. Decisions formation management systems need to be improved across emanating from more senior officials could carry more the units of all MDAs. Legal capacity needs to be shored weight and credibility with citizens. In Romania, there is up to enable the accurate interpretation of the nuances evidence that the heads of institution receiving adminis- of the law and to handle litigation on RTI. In the United trative complaints against the denial of information have Kingdom, for example, although the FOIA states that the supported plaintiffs in favor of the release of information.45 decision of whether or not to release information lies with the authority holding it rather than the authority that pro- But higher-level officials and committees could also refuse vided it in the first place, in practice, it is routine for all appeals if the MDA is keen to guard the information or if affected authorities to be consulted. Systems for communi- it interprets the law in a way that excludes the information cations and information-sharing, consulting and clarifying from disclosure. requests, and coordinating between different departments, must therefore be strengthened. 2.4. Request Procedures Lower-level officials who are typically the point of con- tact for information requests faced conflicting pressures and Good practice RTI laws—usually set out details on the pro- several disincentives for compliance. On one hand, they cedures for requesting information. However, implement- feared reprisals for releasing information that could be dam- ing regulations are important to provide further details and aging to senior officials or politicians; on the other hand, create clarity in the procedures and processes for access, they risked individual penalties for noncompliance. This addressing such issues as establishing the schedule of fees led to a risk-averse culture, where lower-level officials can for access, requirements for application forms, and inter- be over-cautious about releasing information, especially nal procedures for processing requests. When such imple- information on issues that might be seen as “sensitive,� mentation regulations were not adopted—as was the case pertaining to information about the agency’s performance in Albania and Moldova—and regulations were left to to or budget or to the behavior of senior officials—requests individual agencies, only a few adopted them.46 that often emanate from journalists or NGOs. Creating Some countries adopted technology solutions that have the right incentives through empowerment and rewards to considerably eased the request process, at least for those encourage information officers to release information can users who are technologically literate and have access to unblock these constraints to implementation. the Internet (box 2.1). However, while these technolo- Further, RTI laws entailed additional responsibilities gies provide useful prototypes for potential application in without necessarily providing additional resources or re- other countries where conditions could be conducive, bud- balancing of other responsibilities. In India, more than 30 get limitations and contextual realities also dictate what percent of rural information officers polled in one survey technology solutions are practical. For those sectors with a admitted that they did not want RTI to be part of their large number of personal requests, investments in technol- responsibilities, and more than 10 percent of the public in- ogy platforms like Zoom might be useful, especially if the formation officers cited a lack of financial and other incen- majority of the requester base is from the urban population tives as reason for their reluctance.44 Instituting incentives with access to technology. Other departments might need 14 2. INSTITUTIONS AND CAPACITY FOR IMPLEMENTATION to put more effort into communications, creating incen- difficulties faced by people in making RTI requests results tives, and signaling from the department’s leadership to in requests that are often incomplete or unclear, and this create a more open culture. They might make more moder- could be a major impediment to the fulfillment of requests. ate incremental investments in technology platforms and For instance, in a 2007 survey of Mexican officials, 77 per- develop a longer-term vision for full migration. State-of- cent of public officials complained that information re- the-art, advanced technology solutions might also be too quests were not well formulated.48 expensive to implement in poorer countries and in smaller entities who might instead focus on creating a culture of re- 2.5. Performance Monitoring sponsiveness through training and incentives, in the short term. Adequate systems for monitoring RTI performance enable Further, an electronic request system lacks human inter- an assessment of how well RTI is being implemented and action and may make it harder for the requester to explain offer an accountability mechanism for compliance. For in- what he or she actually wants and for the official to pro- stance, in Mexico, IFAI set up a system to systematically vide the relevant information. Requests should therefore track statistics on requests and responsiveness from federal be accepted in a range of formats, including oral requests, agencies, cross-check data from RTI monitoring platforms electronic requests, in writing, or in person. Particularly against data on other technology platforms, and monitor in countries with low literacy, admission of oral requests the outcomes of its appellate decisions. Since it has ac- is very important so that poorer and less-educated citizens tively publicized its findings, this has served as an incentive without access to technology have access to information. for federal agencies to comply with the transparency law. Similarly, in the United Kingdom, the nodal agency has In several countries, laws enjoin public bodies to pro- published quarterly and annual reports online that cover vide assistance to requesters. In practice, however, citizens activities in the bodies under the purview of the FOIA.49 encountered difficulties in lodging requests and gaining ac- cess that ranged from refusal to provide assistance to harass- But in the other countries, monitoring RTI implemen- ment. In India, respondents to a survey pointed out that tation has been a challenge; several countries did not put they often had to make multiple trips to the government in place good monitoring systems and in those that did, office in order to get information released, and that they sustainability has been challenging. In Romania, in the were very rarely able to get assistance from officials in lodg- early years of implementation, all entities covered by the ing requests; in fact, they reported harassment of RTI appli- law published yearly newsletters, statistics, and activity re- cants by officials, particularly in rural areas.47 The technical ports, and data was regularly compiled and published. But Box 2.1. RTI Technology Platforms in Mexico Mexico developed an electronic information request and response platform—the Sistema Informatizado de So- licitudes de Informacion or SISI—that enabled requests to be made anonymously online, responses to be fulfilled and monitored electronically, and appeals to be processed online. A search engine called “Zoom� allows users to search resolutions by topic, agency, or date. Users can also make requests in writing or by mail. When requests are presented in writing and in person, those request are transferred into the e-platform by a staff member so that they can be registered by the system. According to IFAI, 97 percent of all requests were received electronically.) Users not satisfied with the official response can—through INFOMEX—turn to IFAI for an appeal of their request. To com- ply with proactive disclosure provisions and make the formats for posting information uniform, the IFAI designed a central Internet portal called Portal de Transparencia (Transparency Portal or “POT�). Each government agency is required to set a link to the POT on their websites. POT has significantly helped in the verification of compliance with proactive disclosure provisions. 15 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE as bureaucratic restructuring led to a decline in budgets for Performance metrics have also tended to narrowly focus oversight and marginalization of the department charged on the numbers of requests and responsiveness, neglecting with these functions, the overall quality of monitoring data methods that assess or measure the quality, completeness, has also declined. For example, the Ministry of Interior in- and ease of understanding of information or the ease of ac- cluded in its reporting all the applications for personal IDs, cess to it. Agencies are responsible for categorizing their the traffic fine administrative appeals, and the like, push- responses as positive or negative; official data does not ing its number of requests into the tens of thousands per address the quality of the information and responses. RTI year; other ministries made similar uncoordinated calcu- laws usually require that information be disclosed in the lations.50 The completeness and accuracy of data was also format in which it exists. Therefore, although information suspect—data was sometimes only available for some years, was released in response to requests, it was often difficult and agencies over-reported to make themselves look good. to interpret, published in a format that could not be re- used, or contained pages of jargon that are difficult to un- Monitoring and independent surveys on implementa- derstand for an average citizen—or even for CSOs without tion were also carried out by civil society groups, such as adequate training. In Uganda, for example, although de- the RTI Assessment and Analytical Study (RaaG), carried tails of transfers to primary education was posted on notice out by the National Campaign for People’s Right to Infor- boards, communities have often been unable to decipher mation in India.51 Civil society groups in other countries the technical figures.52 also conducted surveys and attempted to assess compliance by floating test requests to different agencies. But these ad Budget reports that were released were at a high level of hoc initiatives often depended on funding from aid agen- aggregation, and disaggregated data on expenditures, nec- cies; provided incomplete data, and cannot be a substitute essary for understanding the minutiae of budget allocations for rigorous and sustainable official monitoring systems. and expenditures (the data most critical for understand- ing where the funding pipe gets leaky) was seldom avail- The development of clear and consistent metrics for able. This could be the result of a deliberate manipulation monitoring has also been challenging in the area of proac- to mask the information or of weaknesses in the underly- tive disclosure. Proactive disclosure can be a cost-cutting ing systems. For example, incomplete data on the flow of measure in the medium term as easing access to the vast funds could come from the public expenditure manage- majority of frequently-sought information eliminates the ment system because not all funds are channeled through need to process requests and reduces the burden on public the government budget or reported. Lack of consistency in agencies. But proactive disclosure efforts revealed several gathering, documenting, and disseminating vital informa- weaknesses that are not adequately captured in official tion in areas like the health sector can create further gaps in statistics. In India, the RaaG study was able to determine the information released. These challenges have emerged some of the challenges of implementing proactive disclo- in several countries, but basic requests and responsiveness sure systems, including the percentage of agencies that indicators neglect these complexities. had published information in various categories and how current it was, but there has been no sustained evaluation Aggregate indicators have also made it difficult to assess process to determine whether or not websites are being the usefulness of RTI for accountability. The requests that maintained or if the information is now outdated. Similar did not receive responses might be precisely the ones that problems emerged in Romania. As countries launch ambi- pertain to information crucial for monitoring government tious open data initiatives, it is particularly important that expenditures and performance and eliciting accountabil- they learn from these experiences and focus on the sustain- ity. Data suggest that individual citizens tend to find ac- ability of these efforts. Appropriate metrics and systems of cess laws useful for eliciting information about a number measurement will be key to this. of areas of personal importance (such as basic entitlements and social security benefits) and for redressing grievances related to the nonprovision of basic services. Procedural 16 2. INSTITUTIONS AND CAPACITY FOR IMPLEMENTATION compliance to these requests is conditioned on the effi- fairly high rates of procedural compliance, even though ciency of the system. But requests that are denied or that records-management systems continue to have challenges. receive no response—even if they constitute a relatively In the short term, it might be useful to focus on some small percentage of total requests—could involve more of the more pressing imperatives that set the foundation for “sensitive� requests, such as those that have to do with effective records management. performance, expenditures, and procurement, among oth- ers. Developing appropriate metrics to measure impact is • Addressing contradictions between regulations and therefore essential. setting in place clear roles, responsibilities, and coor- dination mechanisms is an immediate and basic prior- ity. The relationship between the RTI and archival law 2.6. Records Management was unclear in many countries, or the archives specifi- cally excluded from the scope of the RTI law. This was As RTI laws are set in place, the importance of records problematic because it created two parallel and often management increases because the absence of good records inconsistent regimes for access to information, with management systems53 is often a constraint to the fulfill- civil servants not being sure which to apply. Archival ment of requests.54 In a survey in India, for instance, 25 laws were also inconsistent with modern recordkeeping percent of urban information officers cited poor records systems (particularly in relation to electronic records) management as a key constraint to the swift processing or conflict with access-to-information laws. These dif- of RTI applications.55 Recordkeeping has traditionally ferent regulations have to be harmonized. been regarded as an ancillary function, and archiving and recordkeeping perceived as an administrative or clerical • It would also be useful to set in place clear lines of re- function and as an overhead. sponsibilities and effective coordination mechanisms as the foundation on which to build records-manage- Digitization has been a challenge in several countries. ment systems. In Uganda, for instance, there has been Romanian officials, for instance, reported that officials usu- a perception of fragmentation of institutional respon- ally access information from physical archives in response sibilities on management, storage, retrieval, and dis- to requests because records have not yet been transferred semination of information. The Department of Records to electronic platforms. Further, because of the widespread Information Management in the Ministry of Public misconception by civil servants and elected officials that Service was mandated to oversee records management, the documents they generate belong to them, when leav- provide support and advisory services to registries in ing their post or retiring, they were reported to have been MDAs and local government, and train staff of regis- taking the files home with them and they are forever lost to tries in records management, while the Ministry of In- the archiving system. formation and National Guidance was put in charge of Given the relatively massive effort that upgrading re- retrieval and dissemination. In Mexico and the United cords management systems and digitizing them is likely to Kingdom, the various organizations seem to have been require, it can be quite daunting. It is, therefore, useful to better coordinated. In Mexico, the national archive, in take a more pragmatic approach that prioritizes actions that collaboration with IFAI, defines criteria for cataloging, may have the largest payoffs and that addresses the most classifying, and preserving administrative documents,� challenging constraints. It might not be useful to attempt and the two institutions issue guidelines for the orga- to put in place a full records-management infrastructure nization of archives in all federal government agencies. before implementing the law; much could be accomplished In the United Kingdom, where Section 46 in the FOIA with a sustained incremental approach. In fact, many of requires the government to develop a code of practice the review countries—India, Mexico, and even the United on records management, the national archives and the Kingdom—have implemented the law and demonstrated information commissioner have cooperated to develop this code. 17 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE 2.7. Proactive Disclosure that open data initiatives will need to factor this in going and Open Data forward. Further, open data initiatives should not become a sub- The case studies also highlighted an important emerging stitute for establishing the right to information on an im- issue. As transparency initiatives have proliferated over the portant principle of accountability. last decade, several governments have launched open data initiatives intended to put government data in the public space in formats that can be reused. Various other initia- 2.8. Political Will for tives to create transparency in budgets, contracts, and in Implementation specific sectors have been launched. But these initiatives have rarely been developed in an integrated manner or in Overall, the case studies showed that the implementation conjunction with an RTI law, potentially leading to both a gap resulted from the erosion of political will once the law waste of resources and limited effectiveness. was instituted. The implementation phase rarely embodied the same kind of political momentum as the passage phase. Current international and national standards defining As potential political dividends that can be gained from the scope of RTI typically focus on documents rather than a high-profile reform measure dissipated, implementation on allowing access to full databases, raw datasets, or infor- became a low priority, leading to staggered efforts to put mation in electronic or nonpropriety formats.56 Integrating in place capacity. In Uganda and Romania, the executive access to these types of data within RTI frameworks could oversight agencies faced budget constraints, and in India, address restrictions to the right of the public to reuse gov- the information commission complained of not being given ernment data. For instance, in the United Kingdom, the adequate resources. In Albania, even though the govern- Protection of Freedoms Bill, which was passed in late 2011, ment included favorable pronouncements on freedom of amended the Freedom of Information Act with the inclu- information and transparency in its programs after the sion of an explicit right to many kinds of public datasets. 2005 and 2009 elections, it also issued regulations requiring Coordination between the different departments over- permission from the head of the institution, a minister, or seeing different aspects of information management is also an agency director before releasing any information, dem- critical. In the United Kingdom, for instance, the Access onstrating overt ambivalence to the transparency agenda. to Information Central Clearing House which ensures con- Moreover, as political momentum behind the legisla- sistent application of three main access regimes—the 2000 tion eroded, both political actors and public officials also FOIA, the Data Protection Act, and the Environmental employed a variety of strategies to blunt the effectiveness Information Regulations—has been a central source of ad- of the law. Resistance to implementation came in the form vice for many authorities. In Moldova, the designation of of attempts to create amendments to the scope of the law, an e-government center with committed professionals has refusal to help requesters, or even outright harassment. resulted in significant momentum to the open data initia- In Moldova and India, for instance, the governments at- tives but has not been integrated into RTI implementation. tempted to bring amendments to restrict the scope of the The momentum behind open data initiatives in several RTI law. In both countries, civil society efforts led to the countries could potentially provide a boost to proactive withdrawal of these amendments.57 In India, information disclosure initiatives. However, the experience with imple- requesters reported facing harassment and threats,58 and a mentation also merits a note of caution. All the countries number of RTI activists were even allegedly murdered. In- in the case study complement had initiatives for strength- stances of efforts to block compliance have been evident ening and improving their information-sharing systems in several developed countries as well.59 Clearly, sustaining and some countries, such as Moldova, launched ambitious political will for implementation is the most momentous open data programs. But overall, proactive disclosure sys- challenge of RTI and one that has to be kept on the fore- tems suffered from problems of sustainability, a caution front of implementation efforts. 18 3 . Appeals and Enforcement 3.1. Rationale for Information of information, primarily because general judicial mecha- Commissions nisms are thought to have several limitations as adjudica- tion mechanisms for RTI.61 RTI commissions can signal the Information commissions are part of a multistage appeals importance that the political regime places on strengthen- process that might take on different models in different ing information access. Further, RTI-specific information countries (figure 3.1). The first stage is an internal appeal commissions enable commissioners to become specialists to a committee or senior official within the agency if the on RTI, develop the competence to appropriately interpret official charged with responding to information requests the law, and deal with complaints expeditiously. It enables does not respond or refuses information. If the internal ap- speedy resolution of cases and ease of access, especially in peal does not result in the release of information or the regard to more procedural cases on points of law or pro- requester is not satisfied with the response, he or she can ap- cess.62 Courts, on the other hand, can be inaccessible to peal the agency’s decision to an external body—this could common citizens; can be overburdened, leading to long delays; and entail high litigation costs—even in higher- be direct appeals to the judiciary, information commissions income countries. The weaknesses of the judicial system constituted specifically for this purpose, or an administra- might actually encourage officials to ignore the law or arbi- tive tribunal.60 Information Commissions are usually more trarily deny requests. formal agencies and may have powers to issue rulings on the release of information that are binding on government Information commissions are also considered stronger agencies. Alternatively, the function could be entrusted to than the other common model of redress: the ombudsman. mediation agencies such as an ombudsman with nonbind- An ombudsman is a more informal agency with the man- ing recommendation and advisory powers. date only to issue recommendations and advice rather than to execute binding rulings—a characteristic that can be a Finally, in most countries, both the government and distinct disadvantage in countries where executive agen- the requester can appeal to the courts against the decision cies do not show much deference to the agency. Informa- of the independent agency or the lower court. In Mexico, tion commissions, by contrast, are more formal agencies while individuals can appeal the independent commission’s with powers to issue rulings requiring the release of infor- rulings to the judiciary, MDAs cannot. mation or to undertake other measures that are binding on An independent, specialized appeals tribunal is consid- government agencies. Additionally, an ombudsman, en- ered critical for adjudicating appeals against the denial trusted with multiple responsibilities, might be unable to Figure 3.1. Typical Appeals Process Appeal to the courts Appeal to judicidary or indpendent Appeal to commssion/ombudsman committee or official Request to within the agency information officer upon denial Source: Author. 19 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE dedicate resources to develop a specialization in RTI. Also and prevention of HIV-AIDS to pro-life organizations and an independent information commission with the power to the embezzlement of resources of large government trust issue binding rulings is more costly to administer because it funds—also signaled its willingness and ability to take on is accompanied by a host of additional legal requirements politically-challenging cases and established the serious- and due process mandates. Its independence can be difficult ness of the new transparency reform.65 The British ICO was to guarantee. similarly responsible for improving the compliance of sev- eral MDAs, such as the Department of Health, the Depart- The effectiveness of such a commission depends very ment of Defense, and the cabinet office. Aggregate data on much on the local tradition and administrative culture, the India CIC’s performance is not available, but it has also that is, how much real weight its recommendations carry ruled in favor of transparency in several high-profile cases. and the relative efficiency of the system of checks and balances. Initially, in the United Kingdom, the decision In Uganda, where the law provides for direct appeals to to vest binding authority in the commission was seen as the judiciary—first to the chief magistrate and subsequently undermining the principle of ministerial accountability to to the high court—lengthy judicial processes, absence of the parliament; the FOIA incorporated the provision for a independence from political influence, unaffordable access ministerial veto as a compromise. In practice, however, the by citizens, low staff capacity, financial constraints, and executive has refrained from undermining the authority of current case backlog levels within the courts have been a the ICO and has rarely used the veto, even during poten- deterrent to effective enforcement of ATIA.66 The record tially controversial rulings by the commissioner and the of judicial appeals on one high profile case—that of Tullow tribunal on issues like cabinet papers and the protection of Oil—and more broadly on accountability issues, shows that names of the private staff of members of parliament because the judiciary does not have the specialist technical capacity it would have been politically costly and could have been to address ATIA issues. In response to an appeal on non- perceived as attempts to resist being held accountable.63 disclosure of oil exploration contracts with Tullow Oil, the judge ruled that because the plaintiff—the CSOs—had not Three of the case study countries—United Kingdom, demonstrated the public interest in the disclosure of the Mexico, and India—set up independent information com- contracts, the government’s decision to withhold was valid. missions. In the United Kingdom—the ICO, in Mexico— This was a considerable misinterpretation of the law.67 Lack the Federal Institute for Access to Information, and in of specialist knowledge might be exacerbated by legacies of India—the Central Information Commission (CIC) at the inefficiencies in the judicial system that stretch cases out central level, and state information commissions at the over several years, as in Uganda, where cases have been state level. Two countries—Uganda and Romania—ad- pending for years, including several high profile cases in- opted the model of direct appeals to the judiciary. In Peru volving the freedom of expression. and Albania, an ombudsman was charged with the RTI In Romania, the overall record of the judiciary in re- function. In Albania, Law 8503, the Access to Information sponding to RTI appeals is mixed. In the years following the law, mandated an ombudsman—the People’s Advocate— adoption of the law, the progressive approach of the courts with oversight over the law. on the issues of disclosure and classification strengthened a culture of transparency and clarified the interpretation 3.2. Record on Performance of provisions of the law, setting legal precedents in favor of disclosure. Several high-profile initiatives led to a deter- The three countries that adopted RTI-specific independent rence effect, with public institutions becoming more forth- tribunals have a better record of independence. In Mexico, coming. But in recent years, judges have become swamped between 2003 and 2010, IFAI confirmed the decision of the with heavy caseloads, and the effect is being felt on the federal agency in only 17.6 percent of the 30,833 appeals overall record on adjudication.68 An analysis of judicial de- it received.64 Its rulings on several high profile cases—such cisions on 851 FOIA cases between 2005 and 201169 shows as the deviation of resources allocated to the treatment that only about 34 percent of FOIA cases were adjudicated 20 3. APPEALS AND ENFORCEMENT in favor of information requesters by the first instance budgets and staff as a key constraint to independence.72 Un- court, and of the information requesters who appealed the like IFAI, for example, CIC has faced a staff shortfall, with initial court decision, only 29 percent obtained a favorable several of its sanctioned posts remaining vacant. There is decision at the courts of appeal. also a tendency toward bureaucratization; although the law provides for prominent individuals and experts from differ- ent fields to be appointed, commissioners at both the cen- 3.3. Ensuring Independence and Capacity tral and state levels have tended to be former bureaucrats who bring strong skills and experience on administrative IFAI and ICO both had a high degree of political support, matters but who have come to be viewed as perpetuating a significant allocation of resources, politically-neutral ap- bureaucratic culture and compromising objectivity. pointment procedure, and protections against arbitrary dis- In Albania, the experience of the People’s Advocate missal. A politically-neutral appointment procedure for the demonstrated that a disabling political environment and agency leadership and other personnel, security of tenure, resource constraints can render such a body ineffective. and protections against arbitrary dismissal were important In Peru, the capacity of the ombudsman has been limited in ensuring their independence. The requirement that the because it has several areas of responsibility and activities; appointed commissioners fulfill criteria for technical ex- the maintenance, expansion, or strengthening of RTI work pertise and political neutrality was also critical in ensuring depends on institutional priorities, the will of its directors, competence and for safeguarding against politically-moti- or both. vated appointments. IFAI cannot directly negotiate its budget with the con- 3.4. Sanctions gress; its budget is presented to the ministry of the treasury, which then presents its budget to the congress. But it was The sanctions system for refusal to provide information in politically empowered and given significant resources— the case study countries has displayed several weaknesses, US$18 million in the first year70—enabling it to recruit a creating disincentives for compliance. Sanctions for non- highly competent and well-trained technical staff, which compliance are important to create a credible system of provided it with institutional continuity and stability. In enforcement, but have been weak de facto. Although the 2010, after the passage of the Federal Law of Personal Data laws themselves provide for some mechanisms for sanctions Protection Held by Private Entities, the functions of the or reparations in the event of noncompliance, evidence of IFAI were further expanded to include the responsibilities sanctions being imposed when information is not released of guaranteeing and overseeing compliance with the new law, providing guidelines for the protection of personal despite the rulings of a court or commissioner are rare.73 In data held by private entities, and resolving disputes about most of the countries surveyed, there was little evidence of personal data protection in the entire national territory.71 sanctions being imposed for noncompliance.74 When its mandate increased, it was granted additional re- When the mandate of the appeals and enforcement sources. The Mexican transparency law was passed, and agencies on sanctions is limited, the effective functioning IFAI was established in the wake of what was seen as a of the sanctions system might also require the support of transformational regime change: the Fox administration the executive agency; this can be challenging. For example, came to power after 71 years of Partido Revolucionario in Mexico, although IFAI’s decisions are binding and defin- Institucional rule, championing the transparency law as a itive for all 250 federal entities covered under the law, only flagship program. the SFP can impose sanctions.75 And since 2003, of the 68 In India, on the other hand, where the CIC is funded officials accused of noncompliance with IFAI resolutions, from the central budget as part of the allocation to the only one has been sanctioned by the SFP. IFAI had only ministry, it has cited dependence on the government for sent sanctions recommendations to the SFP as a last resort, 21 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE preferring instead to negotiate with government authori- not releasing this information might make them liable for ties and persuade them to comply. penalties. One way of addressing this is by designating a senior-level person to be responsible for noncompliance Another important issue that was cited referred to on and designating a more junior officer for the day-to-day whom the sanctions are imposed. When junior-level in- functioning of the law. formation officials can be liable for the failure to provide information, they have felt this to be an unfair penalty on More systematic monitoring of the imposition of sanc- them. They might decide not to release what the depart- tions and follow-up actions that result from disclosure deci- ment might consider “sensitive� information because it sion by courts or information commissions can also create could lead to them being reprimanded or receiving other a foundation for the development of stronger sanctions punitive action from senior officials; at the same time, systems. 22 4 . Enabling Environment 4.1. Signficance of the Enabling rights, and civil liberties (figures 4.1 and 4.2—data on U.K. Environment is removed from these charts because it ranks significantly above other countries). This enables better comparison In writing about the experience of the United States in im- among the rest of the countries. In countries that rank plementing its freedom of information law, Kreimer (2008) lower on these indices, there was little evidence of the ef- suggests that transparency succeeds when it is embedded in fectiveness of RTI in accessing information, especially on “a broader web of legislation and regulation,� what he calls governance issues. the “ecology of transparency.� He suggests that the United For instance, countries with stronger rule of law, as as- States FOIA has been effective because the federal gov- sessed through indicators such as Worldwide Governance ernment is surrounded by NGOs and media outlets with Indicatory (WGI) Rule of Law Indicator that measures, the resources to aggressively use the right to information. inter alia, the quality of contract enforcement, property Roberts (2006) also points out that the countries that ini- rights, the police, and the courts (figure 4.1) are likely to tially adopted disclosure laws were “… politically stable de- perform better in terms of following the letter of the law, mocracies with a long tradition of respecting citizen rights complying with information requests in response to the and the rule of law, a lively popular press, and healthy and law’s mandate, and undertaking punitive and corrective independent nongovernmental organizations.�76 The expe- measures when corruption or nonperformance is exposed. rience of the countries included in this analysis shows that In the United Kingdom, for example, the operation of the those that provided evidence of the effectiveness of RTI in FOIA has been facilitated by a dominant norm among pub- accessing information—both of personal interest for citi- lic officials to fulfill their duties under the law. The exposés zens and for holding government accountable—were ones of malfeasance in the parliamentary expenses scandal has with better rankings on a number of governance indices, led to several members of parliament being censured or including the rule of law, voice and accountability, political sanctioned (annex 7). Figure 4.1. Worldwide Governance Indicators— On the other hand, an administrative culture of nonre- Rule of Law sponsiveness and poor service orientation can drive lower- level officials to simply ignore information requests or can 2010 India 2005 lead to an absence of corrective actions, even when infor- 2000 mation obtained through an RTI request exposes misgover- Romania nance. Stronger rule of law might also have an impact on Moldova the broader governance environment through a deterrence effect—if officials anticipate being held to account, they Mexico may alter their behavior accordingly, but this is difficult to Uganda measure. Other research has found that transparency poli- cies do not have an impact on reducing corruption if other Peru conditions for accountability—such as media circulation Albania and free and fair elections—are weak.77 The case study approach in this study—as well as the -1.5 -1.0 -0.5 0 0.5 1.0 1.5 2.0 limitations of the scope and the data that was collected— Higher values correspond to better governance does not enable firm conclusions about the relationship Source: http://info.worldbank.org/governance/wgi/sc_country.asp. 23 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE Figure 4.2. Civil Liberties, Political Rights, and between various indicators of governance and outcomes on Voice and Accountability RTI. 78 Further, attempts to limit the scope and effective- ness of RTI are by no means absent in the more advanced a. Civil Liberties democracies, as several studies have shown.79 The empiri- Romania 2010 cal knowledge in this area is clearly rudimentary. Under- 2005 Mexico 2000 standing the impact of the underlying political economy relationships and institutional forms on RTI outcomes will India require devising appropriate methodologies and indicators Peru for assessing outcomes. The present study attempts to draw Albania out some of the dimensions that emerged as critical to the Moldova functioning of RTI. Uganda 0 1 2 3 4 5 6 1 = highest freedom; 7 = lowest freedom 7 4.2. Capacity and Influence of Civil Society and Media b. Political Rights The supply–demand link in RTI is particularly significant 2010 because it gives citizens a tool to request and gain access to Romania 2005 information that is not part of the exclusions under the law. 2000 Mexico This request-driven aspect of RTI as a tool for transparency India makes the demand side particularly important, providing Peru access to information that otherwise might not be disclosed Albania in the absence of a legal mandate because strong incentives might be attached to keeping it secret. It also implies that Moldova civil society and media groups are particularly critical to Uganda the functioning of RTI—not only as users or intermediaries 0 1 2 3 4 5 6 7 1 = highest freedom; 7 = lowest freedom for information, but also in using RTI to make information public. A large number of requests for information under an RTI law, of course, come from private citizens interested c. Voice and Accountability in information of personal interest about, for example, a health benefit, an adverse decision on school admission or immigration status, or entitlements to services (annex 2010 India 2005 6). But civil society, advocacy groups, media organizations, 2000 and oversight institutions are particularly important for Romania the potential for RTI to serve as a tool for scrutinizing the Moldova functioning of government, forcing the disclosure of infor- mation on government expenditures, decision-making pro- Mexico cesses, performance, and contracts. Uganda The evidence from the case study countries demon- Peru strated the important role that civil society groups played during implementation phases, maintaining pressure for Albania implementation of the laws, launching monitoring ini- -1.5 -1.0 -0.5 0 0.5 1.0 1.5 2.0 tiatives, mounting campaigns against efforts to bring Higher values correspond to better governance amendments to the laws, and using the laws to seek out information on service delivery, corruption, and develop- Sources: Freedom House. http://info.worldbank.org/governance/wgi/ ment issues. The capacity and diversity of civil society and sc_country.asp. 24 4. ENABLING ENVIRONMENT media actors, their awareness of the law, and their financial and policies, from the regulation of state advertising in capacity and technical and legal expertise to seek out infor- the media to new procurement legislation.86 In Moldova, mation and carry out legal battles against noncompliance Uganda, and Albania, on the other hand, civil society were critical. But civil society groups had varying degrees of groups face significantly higher capacity constraints, some- success in these efforts. In Moldova and India, for instance, times even lacking basic resources.87 CSOs were successful in resisting efforts to amend the law Further, in India and Mexico, RTI has been used not to restrict its scope.80 In Albania, on the other hand, civil only by larger, metropolitan NGOs, but also by grassroots society efforts to resist amendments to the law that narrow groups and NGOs in the service-delivery sectors. In India, the scope of disclosure failed.81 following the genesis of RTI was in a grassroots move- In India, Mexico, Romania, and the United Kingdom, ment,88 the nationwide coalition, the National Campaign media and NGOs gained several successes in using RTI to for People’s Right to Information Campaign (NCPRI), has elicit information on several issues related to mismanage- brought together NGOs from across regions and sectors ment of public funds, nonperformance of service-delivery in the country as a powerful galvanizing force for RTI. In agencies, instances of fraud, corruption or nepotism, in Mexico, similarly, there are a number of cases of the trans- both urban and rural areas.82 In Romania, for example, parency law being used by people in poorer and rural com- human rights, governance, and anticorruption NGOs pur- munities as a tool to access basic services (annex 7). In the sued a “legal activism� or litigation strategy, pursuing cases United Kingdom, while groups with more cooperative rela- on a number of topics, including sensitive issues like public tionships with the state, and those attempting to influence procurement contracts and conflict-of-interest cases, with policy development or implementation, are more inclined considerable success.83 Similarly, in Mexico, NGOs84 used to get information through personal networks than through the transparency law to access information about the oper- FOIA, there are a number of campaign and advocacy or- ation and financial management of government programs. ganizations,89 animal welfare groups, and public interest In other countries, CSO efforts to leverage RTI for disclo- groups working against agency capture by industry groups sure have been less successful. In Uganda, NGOs85 have that have actively used FOIA to elicit information. actively tested the ATIA, floating requests for information In Uganda, on the other hand, engagement on this under the law, but their efforts have been consistently met issue seems to have been largely restricted to more promi- with refusals or nonresponsiveness. In Moldova and Al- nent NGOs in the capital. Smaller NGOs, working in a bania, while CSOs regularly used and monitored RTI, the number of service delivery areas, have not had the capac- evidence on success in eliciting information, particularly ity to meaningfully participate in technical areas. In addi- on governance, was very sparse. tion, increasing political competition has actually further Why did the pressure of civil society groups succeed in polarized state-society relations.90 Many NGOs are service some countries and not in others? The relative effective- providers, and have served as instruments for implement- ness of civil society groups in improving governance and ing government programs or filling gaps in service deliv- policy outcomes in different contexts is likely a function of ery; they were able to gain access to information through several characteristics in the broader governance environ- personal networks.91 But for other NGOs, information has ment. Two sets of factors might be highlighted here. The not been easily accessible in the absence of strong personal first is the capacity and diversity of civil society itself, which networks with public officials. The government sees NGOs influences the level of independence under which they are that are working on governance and accountability issues able to operate. India, Mexico, and the United Kingdom, as being strongly antagonistic to the regime rather than as for instance, have strong, mature civil society groups with being legitimate and important actors in the governance diversified sources of income and a historical track record space. The RTI law is seen as a political tool rather than a of engaging the government on tough governance issues. developmental one, as an instrument that can be misused Also, over the last two post-Communist decades in Roma- by the opposition and by antagonistic media and civil so- nia, civil society groups have become mature and diverse, ciety groups.92 actively and effectively engaging the state on several laws 25 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE The second set of factors relates to the responsiveness of enabling environment with strong enforcement institu- the state to civil society pressures. The overall accountabil- tions. For instance, the role of the judiciary is important, ity environment, as measured through indicators such as not only to push for compliance with the law—particularly civil liberties, political rights, and voice and accountabil- in high profile cases, as discussed in chapter 3—but also to ity—measuring, inter alia, freedom of expression and belief, ensure that there is follow-up action. associational and organizational rights, personal autonomy To emphasize again, the anecdotal data collected as part and individual rights, perceptions of the extent to which a of this project is limited. But several of the anecdotes in country’s citizens are able to participate in selecting their annex 7 show that broader check-and-balance institutions government, electoral process, political pluralism and par- in the country are critical for pushing on accountability ticipation, and a free media (figure 4.2)—might be used as outcomes in response to disclosure. a proxy for the state’s responsiveness and accountability to civil society. Countries that ranked higher on these dimen- sions are the countries where there was evidence of better 4.4. Broader Regulatory performance of RTI, both with regard to responsiveness to Environment routine requests and evidence of RTI having been lever- aged to gain information about governance. It is also important to ensure that RTI is harmonized with transparency regulations in other areas. The most signifi- cant of these are secrecy and classification laws that might 4.3. Check-and-Balance sometimes conflict with RTI laws, creating confusion and Institutions contradictory incentives for officials. The capacity and independence of broader check-and-bal- In some countries, asset disclosure laws clearly state that ance institutions—particularly parliamentary committees, this information either shall or shall not be made public. In judicial bodies, and audit and anticorruption institutions— the latter case, this may create a potential conflict with the is particularly relevant to the extent to which exposés of RTI law. Data protection laws now in place in a substantial corruption, nonperformance, and other forms of misgov- numbers of countries establish particular regimes governing ernance, elicited through RTI requests, leads to corrective the collection, retention, handling, and sharing of personal and punitive actions, such as when government officials or private data. Privacy protections under these laws place are made to resign when information exposes corruption constraints on the disclosure of personal data. Efforts need or nepotism, officials are fired when nonperformance is to be made to harmonize RTI with these laws as well as exposed, firms are disqualified when collusion is exposed, with transparency clauses of other laws. and, more broadly, measures are put in place that boost openness. 4.5. Implications for Policy In expanding on the relationship between transparency and accountability, Fox suggests that accountability is a Where the institutional environment is challenging, with composite of answerability (the obligation of public offi- politicization of both the bureaucracy and check-and- cials to inform about and to explain what they are doing) balance institutions, and the political environment leaves and enforcement (the capacity of accounting agencies little space for civil society and media organizations, it is to impose sanctions on power holders who have violated important to help build strong internal constituencies for their public duties).93 Transparency enables answerability reform. Efforts to strengthen civil society capacity to use because the actions of officials is in the public space. But the laws are critical. RTI needs to be part of a comple- it will only make them accountable, leading to safeguards ment of reforms that both strengthens the political and against corruption, better governance, and performance regulatory environment for civil society and helps improve improvements, if it impacts their incentives and is backed capacity. The broader political economy environment—es- up by a credible threat of sanctions.94 This requires an pecially the extent to which CSOs have the capacity and 26 4. ENABLING ENVIRONMENT the political space to exercise voice and influence, and the environment and candid discussions about how likely it way in which laws and regulations impact how effectively this that the RTI is being used if other fundamental po- civil society groups are able to function—needs to be ad- litical and civil rights are not respected. A discussion of dressed as an integral, rather than residual, element in the these aspects will enable a broader set of reforms that both implementation process. When advocating for RTI and strengthens the political and regulatory environment for when designing a law, it might be useful to have a diagnos- civil society and helps improve capacity. tic of the strengths and weaknesses of the larger regulatory 27 5 . Conclusion: Future Directions This report has attempted to sketch out the key areas that • Sequencing and prioritizing implementation mea- are important to address during the implementation of RTI. sures—for instance, an incremental rollout of In particular, it has highlighted both the imperative of set- capacity-building measures—can make RTI im- ting up a number of institutions to build capacity and sup- plementation feasible in lower-capacity environ- port for RTI in the public sector as well as the challenges ments. Resource constraints mean that not all countries in sustaining implementation institutions as political will or agencies will be able to implement advanced state- erodes. Further, it has highlighted the critical importance of-the-art technology interfaces for requests or roll out of the underlying political economy and governance en- full-fledged records-management systems, and that re- vironment, an area often ignored in devising reforms but training for RTI rather than establishing new depart- that is key to whether RTI ultimately works effectively or ments might be more feasible. has any impact. • The “softer-side� of implementation—changes Some of the key lessons emerging from the implementa- in mindset—is at least as important as putting in tion experience of the case study countries bears repetition: place the hardware and technology. Implementa- tion requires a shift in administrative culture. For ex- • First, concerted attention to the establishment ample, junior-level information officers need to be and maintenance of formal institutions of imple- empowered to release information, and training pro- mentation to build capacity and support for RTI grams and rewards need to be instituted. Implementa- within the public sector should be an integral tion should be seen as a systemic change management component of reform efforts. Both a viable over- exercise, requiring awareness-raising for all officials, not sight entity within the government and an independent only specifically-designated information officers. oversight agency have important roles, and sustainable funding for these entities is critical for them to continue • Performance systems, including performance to function effectively. metrics, are a particularly weak area of RTI im- plementation and need to be strengthened. In- The need to maintain the continuity of implementa- vestment in performance systems is critical to monitor tion also suggests that locating the oversight function in implementation and provide incentives for compliance. a mainstream department with cross-cutting responsi- At the same time, appropriate metrics need to be de- bilities instead of creating specialized bodies that could vised for assessing implementation as well as impact. be both ad hoc and too close to the specific political group in power might be useful. A high-profile transpar- • The constraints of the political economy and ency initiative championed by the political leadership the broader governance environment cannot be can lead to the oversight agency directly reporting to ignored. RTI should be part of an integrated comple- the chief executive. In the immediate term, this might ment of reforms that also addresses the broader enabling be a useful measure to increase political support and environment. Diagnoses of the enabling conditions, acquiescence by the other ministries and provide the support for training and capacity-building for civil soci- necessary boost of additional resources, but the imple- ety groups, strengthening other check-and-balance in- mentation agency could be too closely tied to the fate stitutions, and harmonizing RTI with other regulations of the political administration and could atrophy with a should also be integral to implementation efforts. change in regime. 29 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE The report—an attempt to bridge the gap between the base as an argument for dedicating resources and sup- discourse on RTI and empirical analysis of how it actually port for its implementation. Devising appropriate indi- works—also highlights several important areas that warrant cators and methodology to assess the performance and further empirical investigation. These potential areas of fu- impact of RTI is an important first step in this direction. ture research would both help further the knowledge base This report has suggested some basic building blocks in this area and provide guidance to policy makers on how of such a methodology (annex 5), but clearly there is RTI might be effectively implemented and how it might be much to do in this area. leveraged for broader governance and development gains. • Third, understanding the influence of political econ- • First, one key challenge is understanding the chal- omy relations and the broader governance environment lenge of implementation of RTI in poorer environ- on the effectiveness of RTI is critical for both a prag- ments with budget constraints and poor administrative matic assessment of its potential in different country capacity. This is a challenge, not only for RTI, but for contexts as well as a basis for understanding how these transparency measures more broadly, such as open data issues might be addressed during implementation. The initiatives. There is clearly a global momentum toward challenge of much political economy analysis is that it transparency reflected, for instance, in the growing points to the constraints and limitations of a certain membership of the Open Government Partnership,96 context without necessarily being able to suggest how as well as a number of other transparency initiatives these might be addressed through policy mechanisms or (annex 1). How will resource and capacity limita- consultative interventions. Making this analysis action- tions influence what is feasible and how implementa- able—drawing recommendations for action—should be tion choices should be made? Undertaking cost-benefit a necessary part of such analysis to enable RTI to suc- analyses, for instance, would be an important element ceed in different environments. of these efforts. A more extensive and comprehensive assessment of • Second, impact analysis for RTI reforms is a particularly the implementation experience of other countries with weak area of research. Although there is some research RTI laws will also further strengthen understanding of the on the impact of transparency measures more broadly conditions under which transparency created through RTI (annex 5), on RTI specifically, there is very little empir- leads to better governance and development gains, and, ical knowledge. This is critical to understand, not only therefore, what policies and actions are necessary to create as an academic question, but also to create an evidence these conditions. 30 ANNEXES Annex 1 . Global Developments in Transparency and Openness Over the last two decades or so, transparency has emerged transparency. Formal laws establishing the right to informa- as a powerful idea in discourses on governance and devel- tion in several developing countries are part of this trend opment. Undoubtedly, rapid developments in information towards transparency. and communications technologies have played an impor- The emphasis on transparency is also reflected increas- tant role in this. The technology to process and share infor- ingly in the work of the World Bank Group (figure A1.1). mation at unprecedented speeds, has massively increased Within the institution, the role of information as impor- information flows, fundamentally changed cultures around tant for governance gained prominence with the results information, and heightening citizen expectations of what of the Public Expenditure Tracking Surveys in Uganda. they are entitled to know about the functioning of the While subsequent research, showed that there are several government. other factors that determine outcomes, the publication of A fairly rapid expansion of global civil society groups has the PETS research pushed the role of information to the been another driver of this trend. Several prominent civil center of the discussions on governance at the Bank.96 Sub- society groups, focused specifically on monitoring transpar- sequently, the World Development Report of 2004, focused ency in development contexts emerged and have become on service delivery, highlighted information as critical to prominent pressure groups on the design and implementa- making service providers more accountable for perfor- tion of development programs: Transparency International, mance and the delivery of services—in strengthening both a preeminent and pioneering global anticorruption NGO; the “short� and “long routes�97 of accountability. In 2002, Open Society Initiative, funded by the Soros Foundation; the Bank also became a partner in EITI. Publish What You Pay, focused on enhancing transpar- ency in the extractive industries; Publish What You Fund, The adoption, in 2007, of the Governance and Anti- focused on enhancing transparency of development aid corruption (GAC) strategy,98 was another milestone, and agencies; and the International Budget Partnership, which provided a strong institutional mandate to accelerate the publishes estimates of budget transparency. Bank’s work in this area. The Strategy highlighted the role of information—on budgets, on procurement, and various This heightened focus on transparency is reflected in other state records, and particularly access and use of this a number of initiatives: the Extractive Industries Trans- information by the multitude of actors that make up the parency Initiative (EITI), a pioneering multi-stakeholder governance ecosystem—civil society groups, media, advo- coalition, focusing on publication of revenues from extrac- cacy groups, and citizens—as indispensable for improving tive industries, and monitoring these revenues by third accountability. The Strategy was updated in March 2012, parties; initiatives such as the Construction Transparency further stressing the importance of transparency in the Initiative that look to EITI as a model; the Global Trans- wake of the Arab Spring and the global fiscal crisis. parency Initiative to press for IFIs to adopt transparency rules akin to national access to information laws; and the The Bank also amended its own Access to Information International Aid Transparency Initiative, a multi-stake- Policy, which was, interestingly, modeled on some of the holder initiative to develop a voluntary standard obliging more advanced RTI legislations from developing coun- participating donor organizations to publish accessible and tries and expand the scope of disclosure of the organiza- free information about their functioning. The latest expres- tion’s documents. The opening of the Bank’s vast reserves sion of this move towards transparency is the Open Gov- of knowledge—more than 7,000 data sets—for public use ernment Partnership, a multilateral initiative, with more in the form of the Open Data Initiative provided a further than 50 governments drawing up action plans to boost boost to the “Open Agenda.�99 33 ANNEXES On promoting RTI specifically, the Bank’s work has laws. Support for the passage of RTI reforms has been part been somewhat more limited, but increasingly impor- of the policy dialogue in several countries. More recently, tant. The World Bank Institute, the learning and capacity the Bank has provided technical assistance for implemen- building arm of the Bank, has a well-established program tation in a number of countries—Tunisia, Bangladesh, that has supported the sharing of best practices and South- Nepal, Chile. South exchanges with countries seeking to establish RTI Figure A.1.1. Transparency and Access to Information at the World Bank ch te ar da se icy Up Re l Po DR TS C GA I AT PE W 01 04 10 12 20 20 20 20 Ti e C iv GA EI at 02 iti 07 In 20 20 ta Da n pe O 11 20 34 Annex 2 . Observations on the Passage of RTI Reforms Right to Information is not a politically neutral reform and instruments like in Article 19 of the Universal Declaration in many countries, even when the right was recognized by of Human Rights as an intrinsic civic and political right the constitution, several years lapsed before it was opera- of citizenship in democracies, a necessary complement to tionalized with a formal law, either because of overt resis- freedom of expression and media rights. Clause 2 of Article tance or lack of attention. In Uganda, for instance, there 19 states that the “right to freedom of expression� includes was a ten-year gap between the adoption of the constitu- “freedom to seek, receive and impart information and ideas tional guarantee in 1995100 and the passage of the RTI law of all kinds.� The International Covenant on Civil and Po- in 2005. In India, the gap between the Supreme Court’s litical Rights integrated the same language. recognition of the right to information as being contained Regional human rights conventions also mirrored this. in the constitution and the passage of the law was more Article 9 of the 1963 African Charter on Human and Peo- than 20 years. In countries across continents—from Nige- ple’s Rights103 declares that “every individual shall have the ria to Indonesia, from Bangladesh to Chile, civil society ad- right to receive information� and that “every individual vocacy movements for the passage of the law lasted several shall have the right to express and disseminate his opin- decades before laws were adopted in recent years. In other ions within the law.� Article 10 of the European Conven- countries, laws have still not been passed despite years of tion for the Protection of Human Rights and Fundamental civil society advocacy. 101 Freedoms and Article 13 of the American Convention on In several countries, despite several years—even de- Human Rights protect the freedom of expression with pro- cades in some instances—of struggles by civil society visions substantially similar to those contained in the Uni- groups, pressures from development agencies and the in- versal Declaration of Human Rights and the International ternational community, and repeated promises by political Covenant on Civil and Political Rights. The European parties and governments, RTI bills have been blocked, and Convention, however, states that this right may be limited governments have repeatedly stalled on taking actions to “in the interests of national security … for the prevention convert such bills into laws. In Indonesia and Nigeria, for of disorder or crime … for the protection of the reputation or rights of others, for preventing the disclosure of infor- instance, the struggle for RTI legislation spanned a couple mation received in confidence, or for maintaining the au- of decades before the laws were finally adopted in 2010 thority and impartiality of the judiciary.�104 The American and 2011, respectively. In several other countries—Ghana, Convention also contains the caveat that the exercise of Sierra Leone, Kenya, and Zambia—bills are still pending the right to freedom of expression may be limited as neces- despite years of civil society advocacy around the issue and sary for “the protection of national security, public order, despite the involvement of donor partners. or public health or morals.�105 The 1998 Aarhus Conven- tion on Access to Information, Public Participation in International Influences Decision-making and Access to Justice in Environmental Matters links environmental rights and human rights. In In its very first session in 1946, the United Nation’s Gen- June 2003, the Organization of American States General eral Assembly adopted Resolution 59(I), stating that Assembly adopted in Resolution 1932, “Access to Public “freedom of information is a fundamental human right and Information: Strengthening Democracy,� which reaffirms … the touchstone of all the freedoms to which the United that everyone has the freedom to seek, receive, access, and Nations is consecrated.�102 The right was subsequently impart information, and that information is necessary for embedded in international and regional human rights the strengthening of democracy. 35 ANNEXES At the country level, the principle of a legally-man- 19, Open Society Institute, Transparency International, dated “right� to public information was first established and the Carter Center created an additional pressure point. in Sweden as far back as 1776,106 but it is only after the Global civil society—comprised of groups such as 1950s as part of post-war democratic consolidation that Transparency International, the Sunlight Foundation, and countries began to enshrine this right in law. Finland ad- Global Integrity, among others—has helped foster a global opted the law in 1951, the United States in 1966, Denmark acceptance of increased transparency and the creation of and Norway in 1970, France and the Netherlands in 1978, widely-used transparency benchmarks, both of which cre- Australia and New Zealand in 1982, and Canada in 1983. ate an environment in which states may see added benefits The recognition of the right in international conventions to adopting RTI as well as negative consequences stem- meant that, in principle, signatory governments were le- ming from the absence of RTI legislation (for example, the gally bound to give effect to internationally recognized loss of investor confidence). Moreover, international trans- rights, but in practice, internal political dynamics—such as parency networks now link activists in each country, en- the need to check expanding bureaucracies and concerns abling them to share information about best practices and about the erosion of accountability in the face of significant approaches in supporting the passage or implementation of growth in the public sector107—rather than international conventions, drove the enactment of RTI laws. Although a RTI. A global environment that both favors and rewards majority of countries are signatories to these conventions, transparency—and RTI as a tangible measure of this—may up until 1990, only 13 countries had adopted RTI laws, all create a climate conducive to RTI passage in states where of them western liberal democracies. strong popular support and some political support for trans- parency already exists. The example of the western liberal democracies un- doubtedly created precedents and led to the incorporation While international considerations played less of a role of RTI in the constitutions and legislative systems - in East- in India and Mexico, the design of legislation was still in- ern European countries over the past two decades in the fluenced by international engagement. In several of the wake of the post-Soviet democratic transition process, as countries, South–South exchange, and the formation of part of democratization movements;108 in South Africa in networks of civil society groups has been very important the wake of the end of apartheid; in Mexico at the end in supporting local movements advocating RTI reform. of Partido Revolucionario Institucional rule; and more re- Where possible, supporting continued engagement of the cently, in the post-revolutionary Arab countries. The latest countries still attempting to pass the law with international wave in the adoption of legislation has come in countries actors can be useful. with significant governance and economic challenges in In Romania, Moldova, and Albania, the influence of Latin America, Asia, and most recently in Africa. European organizations like the Council of Europe and The increasing international momentum toward trans- the Organization for Security and Cooperation in Europe parency and the importance of access to information have proved influential both in precipitating reforms as well as undoubtedly been influential in propelling more than 75 influencing the process that the reform adoption process countries to adopt RTI laws between 1990 and 2012. The took. In the case of Romania, which adopted the law in pressure from international sources, including development 2001, the pressure of European Union accession was very aid agencies, the European Union, and participation in in- significant. Albania and Moldova are further behind in the ternational and regional conventions, has been important European Union accession process, but European institu- as an informal pressure point,109 a form of “norm transmis- tions were closely involved with the discussions around the sion.� The role of international policy networks and epis- passage of their RTI laws. In Uganda, in 2005, ahead of the temic communities was important,110 and in less-developed elections, the pressure from aid agencies likely exerted sig- countries especially, pressure from development aid agen- nificant influence.112 A number of governance assessments cies and international organizations was important.111 A of the country, such as the 2005 Good Governance Assess- global advocacy movement through groups such as Article ment by USAID—criticized the government for limiting 36 ANNEX 2. OBSERVATIONS ON THE PASSAGE OF RTI REFORMS the space for political participation.113 Access to interna- incentives for political elites to espouse progressive reforms tional resources was also premised on improving gover- provided the opportunity for pro-reform coalitions to be nance. Expanding pluralism, impending elections, and the forged between civil society groups and political elites. need for a reformist face within the country as much as In India and the United Kingdom, the coming into with external partners was also significant. The threat of power of political parties committed to RTI laws as part political competition led the ruling NRM Party to seek out of their political platforms, and the deep public expecta- mechanisms to win legitimacy, both among the people and tions that had been built up as a result, precipitated the with aid agencies by passing a series of progressive reforms. passage of right-to-information legislation. In the United In Uganda, for instance, NRM, took the winds out of the Kingdom, the Labour Party came into power after 17 years opposition’s sails in presenting the law to the parliament of the sway of the Conservatives, who were firmly opposed itself. to RTI legislation. Even though the political leadership wa- Promoting RTI legislation and transparency norms vered—even in the Labour Party—freedom of information more broadly through international forums can acceler- had been a long-standing promise of the Labour Party, and ate the pressure for reform when countries see membership popular expectations were too high to ignore. in these forums as important. Aid agencies, including the In India, a grassroots movement had been ongoing for World Bank, can also exert pressure for reform, but it is im- close to a decade before the RTI law was actually passed, portant that reforms not be driven purely by international and discussions on its passage for even longer—since the actors and that there are strong domestic constituencies 1970s. A particular moment in political evolution when and national ownership.114 In Uganda, as well as in Alba- the leaders of the grassroots movement were able to align nia, because internal constituencies for reform were weak, with the political elite created a unique situation for re- implementation has lacked champions. The 2009 Global form. The Congress Party-led coalition came to power and Integrity Report found that, of all the countries studied, formed the National Advisory Committee, which included Uganda—along with Bosnia and Herzegovinahad the big- some members who had championed the grassroots move- gest “implementation gap� between laws and implementa- ment. In India, the passage of the law was seen as one of tion.115 The report points out that these two countries are the biggest achievements of the congress-led government also among the largest recipients of international donor in subsequent elections. assistance, supporting the argument that aid-dependent countries establish laws and institutions to meet donor re- These groups are not homogenous. In India, for in- quirements but do not necessarily implement. stance, several champions emerged from within the bu- reaucracy that allied with civil society groups working on this issue. Officials might choose to support the passage of Domestic Political Transitions RTI reforms because it helps them gain access to informa- tion. In the United Kingdom, there were also several differ- RTI reforms have often been passed when political elites ences within the Labour Party. perceive that it is to their advantage to support the law in order to win political points with domestic constituencies Transitional periods provide a particularly important and establish their democratic credentials. The passage of opportunity for reform. In the eastern European countries, right-to-information reforms followed soon after new po- the 1990s were characterized by breakdown of the Soviet litical parties came to power, in some instances, electoral Union, post-Communist transition, and over the course of shifts in the party in power and in other instances, a much the decade, the establishment of democratic institutions. deeper political transformation. Ruling parties and coali- The establishment of right-to-information reforms was tions that championed the law were driven by the quest for part of this process of decade-long institutional reform. legitimacy or popularity and the need to establish demo- It was critical for governments to establish credentials as cratic credentials—either to appease external partners or to democratizing agents in the wake of the transition from gain domestic popularity. Historical junctures that created communism. In the wake of the transition, RTI laws were 37 ANNEXES seen as part of instituting new structures to enhance state government draft. While the National Advisory Council legitimacy. draft excluded key security and intelligence organizations, it provided that the exclusion did not apply to information Several other recent examples show that political related to allegations of corruption and violations of human change provides a fairly unique opportunity for passage of rights. When this was removed from the government draft, reforms. In Liberia, after years of conflict, in 2005, Ellen civil society lobbied for this provision to be added back in. Johnson Sirleaf was elected president and quickly took steps to broaden accountability within government. The passage Popularization of the law and advocacy through the of the RTI law in 2010 occurred within the context of a media have been critical in taking forward the passage of broader set of accountability reforms. In Nigeria, where ad- the law. In some instances, where other restrictions on press vocacy groups such as Media Rights Agenda, Civil Liberties freedom are pervasive, the media might see the law as im- Organization, and Nigeria Union of Journalists began press- portant. Although the media was in some instances—such ing for access-to-information legislation in 1993, the Free- as in Romania—somewhat ambivalent, mostly out of fear dom of Information Act was finally signed into law in May that the adoption of such legislation would either obstruct 2011 by the newly-elected President Goodluck Jonathan current systems for accessing information or undermine its who had previously stated support for the bill and who had role as social leaders in purveying information of public campaigned on a reform agenda. In Indonesia, the fall of the interest, compromise its exclusive access to information long-standing authoritarian Suharto government in 1998 sources, and create longer timelines for responses, it was precipitated a democratic transition and a host of accompa- largely supportive and played an important role in main- nying measures to increase decentralization, accountability, taining support for the legislation. The next few sections and transparency. Advocacy around freedom of information briefly discuss the passage of the law in case study countries. began at that time, but it took eight years and four succes- sive administrations for the law to finally come into effect ALBANIA in 2008. Thailand’s passage of the Official Information Act in 1997 coincided with a period of increased support for In Albania, a key driver for the adoption of an ATI law was democratic accountability and transparency under the ad- international pressure and support coupled with a general ministration of Prime Minister Chuan Leekpai, who pushed acceptance of the idea of by local decision makers. Alba- for a number of domestic political reforms, including reduc- nia was a member of the Council of Europe and had also ing corruption and leakage of public funds. ratified the International Covenant on Civil and Political Rights. Many provisions from these international human Where there has been a broad-based consultative pro- rights treaties were included verbatim in the new consti- cess, it helped ensure ownership. In India, Mexico, and the tution. In 1995, Albania was accepted into the Council United Kingdom, the discussions and debate around the of Europe, requested membership in NATO (obtained in passage of RTI legislation provided the space for extensive 2009), and is a potential candidate country for accession to participation by civil society. Civil society pressure helped the European Union. As part of the effort to reconsolidate sustain political will in the face of opposition from or am- democratic gains in 1998 following the meltdown of state bivalence on the part of the governmentin the UK and in institutions after the bankruptcy of the pyramid schemes India, as examples. These are also the countries in which in 1997,116 Albania approved its constitution through a civil society groups have sustained momentum for the use popular referendum. The new constitution, which replaced of the law during the implementation process. Civil society a package of laws introduced after the collapse of commu- groups had an important influence in narrowing the excep- nism, provides for the separation of powers, rule of law, and tions regime in some of the countries. In India, civil society the independence of the judiciary; it also established a par- efforts, including its role in National Advisory Council, liamentary democracy. helped ensure that important provisions, such as the dis- closure of information about corruption and human rights As with other political and democratic reforms of violations by intelligence agenciesand the extension of the that period, the initiative to adopt an ATI law came from law to cover state governments, were reinstated in a diluted the political elite, and in the environment of reform, the 38 ANNEX 2. OBSERVATIONS ON THE PASSAGE OF RTI REFORMS bureaucracy did not resist democratic reforms. Although Campaign for the People’s Right to Information was formed there was support from the media and fledgling CSOs, the in 1996. drafting was mostly a closed process with little public input. The National Campaign for the People’s Right to In- Civil society during the early 1990s was still at an early formation managed to elicit considerable support from the stage of development, and the few recently established top echelons of the Congress-led United Progressive Al- NGOs did not yet have the capacity to play an influential liance government,119 which in its manifesto during the role. With the intervention of the Organization for Secu- 2004 election campaign promised a new, stronger, right-to- rity and Cooperation in Europe, one civil society represen- information law.120 The creation of the National Advisory tative—a legal expert from the Institute for Political and Council, which included key RTI advocates, and strong Legal Studies was included in the drafting committee. The lobbying by civil society were crucial in bringing about pro- constitutional commission tasked with drafting the 1998 gressive changes to the government’s draft RTI bill. Passed constitution maintained an inclusive, open approach to in- by both houses of parliament in May 2005, it formally came ternational advice. Civil society groups participated in the into force on October 13, 2005.121 drafting of the 1998 constitution and strongly advocated for the inclusion of the right to information in it.117 In July MEXICO 1999, nine months after the 1998 constitution came into force, the People’s Assembly adopted an ATI law. The adoption of the ATI law represents a distinctive mo- ment in Mexican politics. The 2000 federal elections rep- INDIA resented a turning point in Mexican politics. President Vicente Fox and his Partido Acción Nacional (PAN), came The 2005 RTI Act of India was unique because of its deep to power replacing the Partido Revolucionario Institucio- grassroots support; and the ability of the RTI movement nal, which had controlled the presidency uninterruptedly to forge a consensus with the top political elite was criti- since 1929. The new government faced several challenges: cal. The Indian constitution does not explicitly recognize it did not control the congress, and the president faced op- a right to information but, in a series of progressive judg- position—even within his own party. There was, therefore, ments in the 1970s and 1980s, the supreme court recog- considerable pressure to demonstrate quick wins and a com- nized the right as an aspect of the fundamental right to mitment to democracy and accountability. The passage of freedom of speech and expression under the constitution.118 progressive RTI legislation was part of this effort. The espousal of the right to access information by a small grassroots organization in rural Rajasthan in the 1990s— A new generation of public officials in the federal gov- the Mazdoor Kisan Shakti Sangathan (MKSS), Organiza- ernment, many of them from civil society; a consensus for change; a newly assertive, independent media, and open tion for the Empowerment of Workers and Peasants)—as support from the academic community created an unprec- part of its broader struggle to secure living wages for rural edentedly favorable political environment for reforms. For peasants, gave information access a distinctly developmen- the first time in decades, different—even competing—po- tal orientation. The mass base of farmers and villagers that litical and social groups who had been unable to agree on MKSS mobilized propelled the adoption of a right to infor- virtually any major public policy reform in the past came mation law in Rajasthan in 2000. Organizations working together on this agenda. The media also attempted to vin- on the environment and human rights, such as the Save dicate and distance itself from its reputation of subjection the Narmada Movement, the National Alliance of Peoples and lack of independence from government controls. No Movements, the Rural Workers Campaign, and Dalit Sang- political party could afford to oppose this legislation in the harsh Samiti—had all also challenged official secrecy and new environment. For opposition parties, the law also of- demanded the release of documents related to their causes. fered an important mechanism for keeping the new govern- These groups were able to draw on the support of opin- ment in check. ion makers and prominent individuals, including retired bureaucrats, lawyers, senior journalists, and academics. A In 2001, a leak of the government’s draft bill to the nationwide coalition of civil society groups—the National press provoked the formation of a loose coalition of media 39 ANNEXES executives, journalists, and academics, which became by these debates, during which 16 amendments proposed known as “Grupo Oaxaca.�122 The group drafted an alterna- by communist members of parliament were accepted.125 tive transparency bill, lobbied members of congress, negoti- The law was adopted on May 11, 2000. The Communist ated with government officials, and organized conferences Party, though initially reluctant to support this law, voted and public forums to disseminate information about the law for it in the end. and generate stronger bases of support, putting the topic in the public agenda. It even drafted progressive provisions PERU that were included in the draft bill that went to congress. In Peru, the right of access to public information was first Grupo Oaxaca members were even invited to participate recognized in the 1993 constitution, but not written into in congressional discussions on the draft—unprecedented a law till 2002—the Law on Transparency and Access to in Mexico where civil society is not formally entitled to Public Information, which went into effect in January participate in the law-making process (although they can 2003.126 Although RTI was included in the 1993 consti- be consulted). The bill that the executive finally presented tution, through the 1990s, the constitutional right was to congress included many of the changes proposed by the never operationalized into law. Toward the end of 2000, group. In April 2002, the law was unanimously approved however, the move towards a formal RTI law, as well as by congress and a few months later, Grupo Oaxaca decided other progressive legislation, was precipitated by an anti- to dissolve. corruption movement that was in turn triggered by exposes of corruption in the Fujimori government. Popular outrage MOLDOVA triggered the passage of several legislative measures to im- Declaring its independence in 1991, Moldova held its first prove transparency. The Peruvian Press Council and the free and fair popular elections in 1994; since then, it has ombudsman’s office had already been quite actively propa- held several more direct elections to parliament and—until gating the passage of the law, including with support from 2000—for the presidency. In 1997, Petru Lucinschi, the for- the British Council and the international NGO Article 19, mer first secretary of the Moldavian Communist Party, be- and the participation of national and international experts, came the country’s second president (1997–2001). In 2000, journalists, media directors, government opposition lead- the constitution was amended, transforming Moldova into ers, and public servants. The corruption scandal resulted a parliamentary republic, with the president being chosen in legislators from across parties throwing their support be- through indirect election. hind the new reform measures. A key result of these meet- ings was the Lima Principles document of November 2000, The Moldovan ATI law was adopted in May 2000 by which listed ten principles of transparency and access to the ruling grand coalition made up of center-right and cen- public information to guide legislation and government ter-left parties. Media and civil society groups created the policies on these issues. necessary pressure for the adoption of an ATI law and con- Most of the draft legislation was submitted by repre- tributed significantly to its drafting.123 The campaign for an sentatives of the Perú Posible Party, which had made a ATI law attracted little resistance from either politicians campaign promise to fight corruption. It had the support or the bureaucracy in the heat of the electoral campaign of other Fujimori opposition groups as well as from a range in 2000. of societal actors, such as the Peruvian Press Council, the The first parliamentary debates began with a draft that Press and Society Institute, and the ombudsman’s office, all was a combination of a media law and an ATI law. Follow- of whom had provided several contributions to the draft. ing suggestions by international experts on Article 19, the The working group encouraged broad-based participation adoption of two different laws—one on access to informa- of these organizations and incorporated their views in the tion and the other on freedom of expression and the media. draft legislation. After it was approved by the Congres- The unanimous adoption of the law in the Moldovan par- sional Committee on the Constitution, Regulations, and liament in 2000 was preceded by heated debates and a se- Constitutional Accusations, the bill was passed by the con- ries of 44 amendments.124 The vote on the law was delayed gress, becoming Law 27806, the Law on Transparency and 40 ANNEX 2. OBSERVATIONS ON THE PASSAGE OF RTI REFORMS Access to Public Information, on August 3, 2002, and went Information Act was passed later128 and operationalized the into effect in January 2003. system of exceptions created by the FOIA. ROMANIA UNITED KINGDOM Pressures for European Union accession played a signifi- The United Kingdom Freedom of Information Act (FOIA) cant role in the adoption of the legislation in Romania. was a long time in the making, with the adoption of the Although there is no specific European Union-level regula- law featuring in Labour Party platforms since 1974. A lim- tion or directive dealing with a general law on access to in- ited system for access had already been put in place prior to formation, a “soft acquis� emerged dealing with rule of law, 2000 in the form of the 1993 regulatory Code of Practice the anticorruption framework, civil service reform, and ac- on Access to Government Information. When the Labour cess to information, which were regarded as prerequisites to Party came to power in 1997, the discussion was reopened. the official start of negotiations on technical chapters. This But divisions remained even within the Labour Party. Re- also created political pressure on accession governments, sponsibility for the ATI law was given to the Home Office, formally and informally, to do something about improving which oversees—among other things—the police, the se- transparency and integrity. curity services, and immigration, sectors that have tradi- tionally been less enthusiastic about openness. After the fall of the Ceausescu regime, the Social However, election commitments and the white paper Democratic Party governed Romania from 1990 until 1996 had produced so much public momentum that abandon- through several coalitions. In 2000, the Social Democrats ing the law had become too politically risky. The govern- returned to power and championed the proposal for an ac- ment’s consultation draft bill was significantly weaker than cess-to-information law. Civil society was especially active the proposals in the white paper, containing much broader during these years, influenced the adoption of several pieces exemptions, less effective enforcement and appeal mecha- of key legislation, and organized the campaign to push for nisms, and a more restrictive public-interest test. Civil so- the law’s introduction. The mass media also saw the law as ciety groups fought back hard against the weaker proposals a good instrument, although they were concerned about a contained in the draft law, which was also severely criti- formalization and bureaucratization of interaction with in- cized by select committees in both the House of Lords and stitutions, with information being provided exclusively in the House of Commonsthis resulted in a number of pro- response to written requests, and responses postponed until gressive improvements. The FOIA was passed in 2000. The the legal deadline for several weeks. main operative provisions relating to requests came into A group of conservative members of parliament (mostly effect, in a single so-called “big bang,� on January 1, 2005. ex-communists and members of the extreme right) were It is widely assumed that the electoral concerns of senior pushing a draft of the Classified Information Act127 in the political figures were an important contributing factor. Ad- chamber at the same time, which both the government ditionally, some observers have noted that the timing of and civil society perceived as a threat. If the lassified In- the announcement—in November 2001, shortly after the formation Act was passed first and FOIA later, then non- attacks of September 11 in the United States—might have disclosure would be taken as the norm and disclosure as been significant. the exception. This resulted in a common agreement by all parties to a draft FOIA—a rare consensus in the Roma- UGANDA nian parliament. In 2001, the parliament approved the Law Uganda predates its African counterparts in the passage on Access to Public Information (544) under considerable of access-to-information legislation by several years. The pressure from domestic and international NGOs. The final 1995 constitution incorporated a guarantee on access to version of the law represented a merger between the drafts information to all citizens, and Article 41(2) required the of the ministry and of the opposition as well as various parliament to make laws for obtaining access to informa- provisions and articles proposed by NGOs. The Classified tion, reflecting the concerns of the progressive Uganda 41 ANNEXES Constitutional Commission regarding the human rights treaties that advance the right to access information. How- abuses of the preceding decades. In 2001, the NRM party ever, the ambivalence of the regime was evident by the fact won the single-party election with nearly 70 percent of the that a number of restrictive pieces of legislation were also vote, but challenges to the NRM’s political dominance passed during this period, including the Anti-Terrorism were already emerging amid concerns over the lack of prog- Law 2002, in the wake of the terrorist attack in the United ress on democratic reform and the introduction of a consti- States, which provided the pretext for it); the Uganda Peo- tutional amendment to eliminate presidential term limits. ple’s Defense Forces Act 2002; and the Police Act 2005. There was also considerable pressure from aid agencies Between 2002 and 2005, a number of prominent civil to make governance reforms and open the political space society groups, including Advocates Coalition for Devel- concerns about the worsening political and governance opment and Environment, Foundation for Human Rights environment. Initiative, and the Anti Corruption Coalition Uganda Between 2001 and 2006, key pieces of legislation were were engaged in trying to promote the right to access infor- passed, probably as a result of the imperative faced by the mation, actively advocating for ATI as a tool for develop- NRM to seem progressive and reformist in the wake of the ment and for fighting corruption. In 2004, a group of CSOs 2006 elections—especially as a perception about a decline moved toward introducing a “private members bill.� How- in the political and governance environment and pres- ever, the government requested the withdrawal of that bill sures by aid agencies to open up the political space both and subsequently, the minister of information introduced increased. The 2005 Access to Information Act was part the government’s bill. It passed in July 2005, and came into of this complement of reforms at the international level, effect on April 20, 2006. Uganda signed and ratified several international and re- gional conventions and several international and regional 42 Annex 3 . Ratings of RTI Laws by the RTI Ratings Index Global standards in RTI legislation have been driven pri- innovations in legislative provisions, as well as the influ- marily by advocacy groups, and championed particularly by ence of the international policy and epistemic networks, a London-based global advocacy group, Article 19.129 Some lower income countries are adopting fairly progressive leg- more recent efforts by the Organization of American States islation. The RTI Ratings Index, developed by these orga- (OAS) has also been directed towards developing these nizations is a detailed assessment of global RTI laws against standards. An active NGO community links global experts 61 indicators in seven categories: Right of Access, Scope, on legislation with local contexts—among them Carter Requesting Procedures, Exceptions and Refusals, Ap- Center, Open Society Initiative, Commonwealth Human peals, Sanctions and Protections, and Promotional Mea- Rights Initiative, and the World Bank, among others, and sures, drawn from a range of international standards and have been important in the quality of legislation, leading to a comparative study of RTI laws from around the world. convergence on legislative standards across countries. Figures A3.1–A3.8 provide a snapshot of the rankings of the case study countries in the seven categories. The x-axis In fact, as a recent assessment by two authoritative orga- represents the total score possible for each country. While nizations in the field, Access Info Europe and the Centre several of the laws followed global standards, specific pro- for Law and Democracy, points out, newer laws might be visions might reflect negotiated compromises between more advanced than the earlier Western democratic adopt- different stakeholder groups, local power relations, and ers, because they have the advantage of benefiting from specificities of the national context. Figure A3.1. RTI Ratings Scores for Case Study Countries a. Quality of RTI Laws b. Right of Access maximum = 150 maximum = 6 India India Mexico Mexico Uganda Uganda Moldova Moldova Peru Peru United Kingdom United Kingdom Romania Romania Albania Albania 0 30 60 90 120 150 0 1 2 3 4 5 6 Source: http://www.rti/rating.org. Source: http://www.rti/rating.org. 43 IMPLEMENTING RIGHT TO INFORMATION: LESSONS FROM EXPERIENCE c. Scope of Legislation d. Exceptions and Refusals maximum = 30 maximum = 30 India India Mexico Mexico Uganda Uganda Moldova Moldova Peru Peru United Kingdom United Kingdom Romania Romania Albania Albania 0 10 20 30 0 10 20 30 Source: http://www.rti/rating.org. Source: http://www.rti/rating.org. e. Procedures for Access f. Promotional Measures maximum = 30 maximum = 16 India India Mexico Mexico Uganda Uganda Moldova Moldova Peru Peru United Kingdom United Kingdom Romania Romania Albania Albania 0 10 20 30 0 4 8 12 16 Source: http://www.rti/rating.org. Source: http://www.rti/rating.org. g. Appeals h. Sanctions and Protections maximum = 30 maximum = 8 India India Mexico Mexico Uganda Uganda Moldova Moldova Peru Peru United Kingdom United Kingdom Romania Romania Albania Albania 0 10 20 30 0 2 4 6 8 Source: http://www.rti/rating.org. Source: http://www.rti/rating.org. 44 Annex 4 . Scope of RTI Legislation Since transparency is only useful if it exposes relevant infor- A good practice RTI law is undergirded by the principle mation, the scope of disclosure—the relative scope of what of maximum disclosure, which has a few key corollaries: information is mandated to be disclosed and which is not, • All information held by public bodies should be dis- and the entities that it covers—matters for how effectively closed, unless explicitly forbidden by a limited regime the law can be used to elicit transparency, and account- of exceptions to protect overriding public and private ability in governance. Laws that mandate the disclosure of interests. When the law specifies a positive list—that is, only a limited amount of information will not be an effec- it enumerates the categories of information that should tive instrument for accountability. On the other hand, laws be subject to disclosure, it implicitly limits the scope might incorporate a very strong disclosure regime, but the of disclosure—only those categories of information capacity among public officials to understand and interpret that are in the list are mandated to be disclosed. On the exceptions, among the judiciary to interpret the scope the other hand, when it includes a negative list, that is, of disclosure, and the independence in the judiciary to enumerates the categories of information that are ex- apply the exceptions and the letter of the law with fairness empted from disclosure, it implicitly expands the scope might be limited. Further, which agencies come under the of the law to all other categories of information. purview of the law is important. The right to information applies to public agencies, but the definition of what con- • Related to this is the principle that exceptions should stitutes a public agency, and which arms of government fall be clear and narrow. Good practice standards suggest under the law, is not the same across jurisdictions. On the some legitimate exceptions, typically extending to: na- other hand, laws might incorporate a very strong disclosure tional security; public health and safety; prevention, regime, but the capacity among public officials to under- investigation and prosecution of legal wrongs; privacy; stand and interpret the exceptions, among the judiciary to legitimate commercial and other economic interests; interpret the scope of disclosure, and the independence in fair administration of justice and legal advice privilege; the judiciary to apply the exceptions and the letter of the and legitimate policy making and other operations of law with fairness might be limited. public authorities.130 While this report has focused on implementation mea- • Further, exceptions are usually subjected to certain cri- sures, it is useful to highlight the good practice standards teria, with the application of a harms test—covering vis-a-vis the scope of coverage of the law, dynamics that only information, the disclosure of which would cause influenced the relative scope of the law in different coun- harm to the protected interest. tries, and assess significant challenges that have come up • Another key principle is that of a public interest over- during implementation. The experiences of these coun- ride, meaning that information should still be released if tries in designing and applying the law demonstrates that the overall benefit of disclosure outweighs the harm to broadbased consultative processes expand the scope of the a protected interest. “Hard� overrides apply absolutely, legislation, particularly providing a check against efforts to for example for information about human rights, cor- restrict the scope of the law. It also demonstrates that there ruption or crimes against humanity. are challenges in the course of implementation, ranging from implementation of the law to attempts to restrict in scope ex-post. 45 ANNEXES • Specificity on exemptions is also useful. The UK FOIA, punishments for breach, and are backed up by the Criminal for instance, is considered weak on exemptions because Code. Contradictions with other legislation—for instance, it has a long set of exemptions compared to most other contrary legislation on secrecy, as well as provisions for laws, and something that CSO advocates have been secrecy and openness in other pieces of legislation—can critical of—23 sections exempting various kinds of in- create lack of clarity and contradictory incentives, weaken- formation.131 Yet, because it specifies the exceptions in ing the overall environment for accountability. Disclosure detail, in contrast to many other laws, which rely on a provisions and restrictions in other laws—such as asset dis- smaller number of more general exemption statements, closure laws, general administrative or civil service laws, some of the impact of this is mitigated as it limits of- archival laws, whistleblower laws, budget laws, local gov- ficials’ ability to exploit the law’s ambiguities. ernment laws—can create lack of clarity and contradictory incentives and lack of clarity. For instance, the Albanian • It is also considered good practice to have an overall Cold War Studies Center (ACWSC) and the Cold War time limit on exceptions, particularly in relation to pro- International History Project have been hampered by con- tected public interests. tradictions between the access to information, state secrets While most of these principles are reflected in the de- (Law 8457), and declassification laws in accessing records sign of the laws in the case study countries, civil society in the Central, State, and Foreign Ministry Archives132 groups had an important influence on narrowing the excep- Data protection laws, in place in many countries, place tions regime during drafting and consultations on the law. constraints on the disclosure of personal data. These might In India, civil society efforts, and its role in the National be complementary to the privacy exceptions of RTI laws, Advisory Council (NAC) helped ensure that important but in some cases, there might be a conflict with how the provisions, such as the disclosure of information about two laws define personal data. corruption and human rights violations by intelligence The inclusion of “deliberative information� is another agencies, and the extension of the law to cover state gov- controversial area. For instance, in the UK, the FOIA bill ernments, were reinstated in a diluted government draft. included an exemption on policy advice from civil servants Section 23 of the Indian law explicitly overrides anteced- in response to ministerial desire to preserve confidential ent secrecy laws, such as Official Secrets Act, 1923. Further, advice from civil servants and the neutrality of the civil notwithstanding the regime of exceptions, any information service.133 Debates focused on distinguishing between fac- which can be provided to Parliament or a state legislature tual material on which policy recommendations are based, must also be provided to citizens. In the UK, information and the specific advice that ministers receive from the civil on investigations into health and safety (for example, prod- service. As a result, instead of the government’s proposed uct safety reports, pollution investigations and documents absolute exemption for policy advice, a qualified exemp- concerning workplace accidents), were initially exempted tion was included—although information can still be with- from the consultation bill, but this exemption was removed held while a policy is being developed, once a decision has in Parliament under pressure from civil society groups. been made, the grounds for withholding information are They also played an important part in narrowing of exemp- far fewer. That this was not an anticipated effect is sug- tions covering commercial interests and information pro- gested by Tony Blair’s admission, in 2010, that he consid- vided to governments in confidence (that is, by regulated ers the FOIA to be one of his greatest mistakes because it industries to regulators). However, the UK has a number of impinges on the ability of ministers and civil servants to exceptions that are not harms-tested, and the public inter- discuss policy proposals frankly.134 In India, the government est override also does not apply to nine of its exceptions. has been reluctant to disclose file notings—the opinions To the largest extent possible, the RTI law must be and notes of civil servants on government files that sum up harmonized with other regulations so that the legal envi- the decisions taken on a particular matter. In 2006, soon ronment as a whole does not provide contradictory signals. after the law was enacted, the government issued a notice For instance, pre-existing legal institutions such as Secrecy stating that file notings were not to be disclosed under the laws provide for broad categories of secrets, and severe RTI Act. Subsequently, it prepared draft amendments to 46 ANNEX 4. SCOPE OF RTI LEGISLATION exclude file notings from the purview of the act.135 Civil so- important because most information impacting the welfare ciety groups and activists challenged the notice before the of citizens and delivery of services lies at these levels. In central and state information commissions, who supported India, the government draft of the Central 2005 law ini- the view that file notings could be accessed under the law. tially applied only to the Central government, but as a result They also launched a major campaign with the support of of strong civil society reaction, coupled with support from the media, successfully stalling the government from push- the NAC, the original, broad provisions, extending the law ing through the amendment.136 to cover sub-national entities, particularly important for the poor, was reinstated. Although most states have separate A key challenge with exceptions is to ensure that they RTI laws and state-specific implementation institutions, are interpreted in an appropriately narrow manner. This is such as the state information commissions, the central law something of a challenge, since it is almost inevitable that overrides the state-level laws. In most countries, however, the RTI law describes exceptions in rather general terms. there are separate laws—federal laws governing access to It can be difficult for decision-makers to assess whether information held by federal public bodies and state or pro- disclosure of information will cause harm to the protected vincial laws governing access to information held by state interest, and applying the public interest override is even bodies. In most countries—for instance, Mexico, Canada, more challenging. Training and expertise to interpret the Germany and the United States—the national law only ap- law is important to prevent the scope of the law from be- plies to federal public bodies and the federal government coming too restrictive during implementation. Attempts to has no power under the Constitution to regulate openness strengthen secrecy might be made even after the passage at the sub-national level. The transparency standards of of the RTI law. In Albania, for instance, in 2006, the gov- sub-national bodies might vary and minimum standards of ernment attempted to strengthen secrecy, creating a new good practice might be difficult to apply uniformly. For in- level of classification. While the Parliament approved the stance, in Mexico, concern has been expressed that the RTI amendments, the Albanian Helsinki Committee success- laws which have now been adopted by all 32 sub-national fully intervened, sending a letter to the Speaker of the Par- entities do not meet minimum transparency standards.139 liament requesting her to stop approval of the amendments. In 2005, the Moldovan authorities drafted legislation on The law might apply to only the executive in some state and service secrets—media and NGO groups, sup- countries, to the exclusion of the legislature, the judiciary, ported also by the international community, vehemently or oversight institutions, but leaving out non-executive in- opposed the draft, which was subsequently withdrawn.137 In stitutions from the purview of the Law can be problematic, 2006, Peru adopted separate legislation138 establishing lon- for instance, when parliamentary or judicial discussions ger time limits for the declassification of information, even and debates cannot be made public. In many Common though this could still have been done under the RTI law. Law countries, however, the courts are excluded (Uganda and the United Kingdom), and in Uganda, the legislature Which agencies come under the purview of the law is is also excluded. In Uganda, the law also includes an excep- also important. The right to information applies to pub- tion of Cabinet records (which are only accessible after a lic agencies, but the definition of what constitutes a public minimum of seven years), which civil society sees as being agency, and which arms of government fall under the law, a key weakness.140 In Mexico, because of constitutional di- is not the same across jurisdictions. Most countries extend vision of powers, the oversight body, IFAI, only has juris- coverage of the law to all public agencies, with some excep- diction over bodies that form part of the executive. The tions, usually specialized security and intelligence bodies. law prescribes very precise rules for the executive, but calls In India, while the NAC draft excluded key security and on congress, the judiciary and the constitutional autono- intelligence organizations, it provided that the exclusion mous agencies to put in place comparable arrangements.141 did not apply to information related to allegations of cor- In India, the Supreme Court has resisted being under the ruption and human rights violations. RTI Act, but the CIC has ruled that in a conflict between The extent to which national RTI laws cover sub- the RTI Act, and the internal rules of a Public Authority, national entities, provincial and local governments is including the Supreme Court, the RTI Act must prevail.142 47 ANNEXES Moreover, as several public services are increasingly de- the government exerts influence over governance (e.g., livered by private contractors under various kinds of pub- through appointments governing boards), or through regu- lic private partnership arrangements, the extent to which lation, especially when operating in conditions of a State- these entities should be subject to disclosure laws also be- protected monopoly (postal service, forest management, comes relevant. This is of growing importance in many state export-promotion bank, and so on). For instance, countries given the increasing privatization of government when the prominent NGO, APADOR-CH requested from and the increasingly prominent role private bodies play in Bucharest Municipality information on public procure- the delivery of to be public services. In the United King- ment contracts for road construction and repairs during dom, for instance, quasi-autonomous government organiza- 2000–05, the Bucharest Municipality only communicated tions (so-called “QANGOs�) were often criticized for their the list of contracts concluded, but refused to send copies of opacity and lack of accountability. Their inclusion in the the contracts, invoking confidentiality clauses. The courts FOIA was a victory for openness proponents. In Romania, found that how public money is being spent—including the precedents created by courts lean towards a liberal, through procurement—is information in the public inter- pro-transparency notion of the term “public institution,� est, and that confidentiality clauses in public procurement suggesting that the law should cover not only bodies fi- contracts do not supersede the public’s right to know. nanced entirely from the public purse, but also those where 48 Annex 5 . Methodological Issues in Assessing RTI Effectiveness and Impact Theoretical Links Between RTI laws are based on the premise that because infor- Governance and Development mation asymmetries enable officials to pursue policies and actions that are in their own narrow interests rather than Economists have long recognized the importance of infor- ones that are for the broader public welfare, regulatory in- mation for the functioning of markets. 143 Better informa- tervention is necessary to reduce information asymmetry; tion flow helps private agents make more rational decisions this would lead to more accountability, less corruption, about asset prices, consumption, and investment, leading and, eventually, better development outcomes. to more efficient markets. Lack of transparency, on the Highlighting the importance of information as critical other hand, increases the risk of investment, which trans- to accountability in service delivery, the 2004 World De- lates into higher risk-premiums.144 Transparency is seen as velopment Report (WDR)146 contended that information creating a more favorable environment for investment, enables citizens both to hold policymakers accountable— reducing market uncertainty about policymakers’ prefer- who, in turn, hold providers accountable (i.e.g, the long ences, which results in more predictable monetary policy route of accountability)—and exercise direct influence over and efficient financial markets. service providers (i.e., the short route of accountability): In the political space, there is also intrinsic informa- “Perhaps the most powerful means of increasing the tion asymmetry between officials and citizens. While politi- voice of poor citizens in policy-making is better infor- cal leaders and public officials are supposed to know more mation…. Better information—that makes citizens about a range of public issues, including the defense and more aware of the money allocated to their services, the economy, they lack incentives to work in the public the actual conditions of services, and the behavior of interest because citizens have little control over decision- policymakers and providers—can be a powerful force making and policies—except during elections. Highlight- ing the implications of information asymmetries between in overcoming clientelist politics….� 147 officials and citizens, Stiglitz points out: Access to information can be cast in intrinsic terms: as a “There are asymmetries of information between democratic right of citizens to access information about the those governing and those governed, and just as state’s functioning and the democratic responsibility of the markets strive to overcome asymmetries of informa- state to make this information public. But the rapid spread tion, we need to look for ways by which the scope of RTI legislation in developing countries has led to a focus for asymmetries of information in political processes on RTI in more instrumental terms, addressing some of the can be limited and their consequences mitigated … most difficult governance, welfare, and development chal- Just as asymmetries (of information in the market) lenges by opening up the decision-making and functioning give managers the discretion to pursue policies that of the state to public scrutiny.148 Access to information, in are more in their interests, than in the interests of this instrumental formulation, is important because it is the shareholders, so they allow government officials a precondition for enabling citizens to hold governments the discretion to pursue policies that are more in accountable; itenables them to exercise voice and partici- their interests than in the interests of the citizenry. pation; and it controls corruption.149 This would, in turn, Improvements in information and the rules govern- lead to positive development outcomes, such as improved ing its dissemination can reduce the scope of these allocation and use of public resources and better outcomes abuses in both markets and in political processes.�145 with regard to service delivery and human development. 49 ANNEXES Seen through a rights-based lens, RTI can be seen as RTI Results Chain essential to the exercise and realization of “second genera- tion rights,� that is social and economic rights, including The project looked at the effectiveness and impact of RTI health, education, and livelihoods.150 Calland (2003) notes in terms of a results chain, linking the passage of an RTI that since the right of access to information empowers citi- law to eventual development outcomes through intermedi- zens to demand information from the state, it changes the ate impacts on use and compliance, follow-up actions, and balance of power between them: citizens can hold the state broader governance and anticorruption outcomes. The re- to account, not only for information, but also for how it is sults chain is depicted in figure A.5.1, and the methodolog- delivering on its other obligations, including their social ical issues related to different outcomes along the results and economic rights. chain are explained in the following sections. Jenkins and Goetz (1999) credit the Indian grassroots Assessing use and compliance. The effectiveness of NGO, Mazdoor Kisan Shakti Sangathan (MKSS)151 with an RTI law in making information available—the first link a prominent role in shifting the discourse on RTI. The on the results chain—is usually measured by assessing sta- movement, anchored in the right for wages and livelihood tistics on the extent to which citizens exercised the right at the local level, located the right to information within to information (through number of requests made annu- the Indian Constitution’s provisions guaranteeing the right ally) and whether the existence of a legally-mandated right to life and livelihood, characterizing it as necessary for the propelled officials to disclose information (through statis- exercise of socioeconomic rights for better livelihoods and tics fulfillment versus denial of requests).152 Additional in- access to better services from the state. It exemplified it formation on the profile of users, the kinds of information in such slogans as “the right to know is the right to live.� requested, and the purposes for which it was requested pro- Subsequently, in India and elsewhere, the right to informa- vides insights into how RTI regimes are working in differ- tion became characterized as critical for citizens to under- ent countries. But in most countries, monitoring systems stand public policy decisions regarding basic services, such are extremely weak. Data on use and responsiveness are as health, education, sanitation, water, and infrastructure, available in some countries where oversight agencies track and to exercise a more direct form of accountability for the statistics on different agencies. provision of these services. The next section looks at the results chain between RTI and development outcomes. Figure A.5.1. RTI Results Chain Use and Follow-up Actions Governance Development Complliance Outcomes Outcomes Was corruption What types of punished? Less corruption Better services Right to information in procurement Information Safeguards LIvelihoods are requested? and budgets put in place? More efficient What types of Performance Better policies infrastructure information improvements adopted? development are disclosed? Source: Author. 50 ANNEX 5. METHODOLOGICAL ISSUES IN ASSESSING RTI EFFECTIVENESS AND IMPACT Even when such data are available, they might only Does disclosure make a difference? The second step be available only for a few years, and is sometimes incom- in the results chain refers to whether any actions are taken plete. Agencies might occasionally over-report in order to as a result of exposés through RTI—such as when officials make themselves look good. The absence of baseline data are sanctioned when corruption is exposed, when sanctions about the responsiveness to information requests before are imposed or corrective measures are taken when poor RTI makes assessing improvements difficult. The absence performance is brought to light, when better safeguards are of baseline data also makes it difficult to assess whether or established to improve governance and prevent corruption, not the existence of an RTI law has led to an increase in or when service providers are held accountable for the ef- the available information or in making accessible informa- ficient delivery of services. Specific instances and anecdotal tion that would not have otherwise been released. With data also enable the examination of whether or not these these caveats in mind, the case studies collected data on actions were undertaken. requests and responsiveness to the extent they were avail- Has there been a systematic improvement in gov- able. Official data was available the United Kingdom, Ro- ernance and development outcomes? The third and mania, Mexico, Peru, and India. The study drew from these fourth links in the results chain relate to whether disclo- to understand levels of use and responsiveness. sure through RTI and subsequent follow-up action lead to Official data were supplemented with data from studies broader improvements in governance and, subsequently, implemented by civil society groups, which provide a rich the strengthening of development outcomes, that is, better dataset on use, responsiveness, profile of users, kinds of in- allocation and use of public resources and better outcomes formation requests, procedural and governance issues, the on service delivery and human development indicators. visibility of the law, and public awareness.153 These stud- ies use a triangulation of methods, including floating test requests to a sampling of government agencies and draw- Empirical Evidence on Transparency, ing conclusions about responsiveness. The largest country- RTI, and Development level study of this kind was a 2010 exercise carried out by the Research and Analysis group (RaaG) in India, using The empirical evidence of the impact information access test requests, triangulated with surveys, interviews, and of- and transparency have on accountability, better gover- ficial data sources.154 These studies are valuable for their nance, and better service delivery is fairly limited. One breadth, scope, and content, but are costly enterprises—a method is macro-level, cross-country analyses that inves- drain on CSO resources; funding for such exercises occurs tigate links between different indices of transparency and on an ad hoc basis.155 Absent the high profile and signifi- various governance indicators.156 These studies show that cant civil society momentum, such exercises are difficult to transparency is associated with better socioeconomic and implement in most countries. They can be useful sources human development indicators as well as with higher com- of information, but they are not a systematic substitute for petitiveness and lower corruption.157 There is a positive robust monitoring mechanisms. correlation between RTI legislation, the length of time it has been into existence, and a series of good governance Anecdotal data about specific instances of information indicators, like voice and accountability, regulatory bur- requests, especially when the information is related to ac- den, government effectiveness, graft and corruption, and countability or corruption, and about responsiveness to bureaucratic quality.158 These are mostly studies of correla- these requests is helpful in understanding if information re- tion, not of causation. lated to government functionsis accessible through the use of RTI; this type of information can be particularly useful Other micro-level quantitative research looks more for accountability and might expose corruption and non- specifically at the link between an increase in citizen ac- performance. Extensive research from secondary sources, cess to information on specific issues and improvements in including media articles, court cases, and research papers, service delivery or reduction in corruption. For instance, in addition to in-depth interviews with several stakehold- research showed that a newspaper campaign aimed at dis- ers, enabled access to anecdotal data. tributing information about school grants helped to reduce 51 ANNEXES capture of these funds at the local level in Uganda.159 Sub- examination of a community-based information campaign sequently, it has been argued that other factors accounted on school performance in India found that information in- for the decrease in corruption in this case and that the im- tervention had a positive impact, particularly on teacher pact of information dissemination was overstated.160 Besley presence and effort. Another study of the impact of infor- and Burgess (2001, 2002) find that regions in India where mation on the ability of communities in India to engage in the media are more active are also the regions least likely service delivery-related accountability mechanisms dem- to suffer from famines during droughts because voters are onstrated that providing information had little impact on more informed about political choices and are able to cast accountability.161 In comparing two cases from India and votes accordingly; this makes political leaders more ac- Uganda, two different studies of community engagement countable. But other examples show mixed results. One with information came to two different conclusions.162 Box A.5.1. Evidence from Research and Studies on Impact of Information • Pandey et al. (2009).a Disseminating information to communities in three Indian states about their mandated roles and responsibilities in school management contributed to a number of improvements in education, including read- ing, writing, and mathematics scores in some of the grades that were tested; teacher attendance and effort; the deliv- ery of certain benefits to which students were entitled, such as a stipend, uniform, and midday meal; and increasing community participation. • Nguyen and Lassibille (2008).b In Madagascar, bottom-up interventions involving parents in school monitoring and action-planning were effective in improving teacher behavior, student attendance rates, and test scores, while top- down control interventions seemed to have minimal effects. • Bjorkman and Svensson (forthcoming).c A sustained effort to inform and involve citizens and communities in moni- toring health care providers in Uganda resulted in improvements in both the quality and quantity of health care provision due to increased effort by the staff to serve the community. However, other randomized control trials do not come to similarly encouraging results. • Banerjee et. al. (2008).d In a randomized evaluation of an education program in India, the study found that providing more information to communities had no impact on community involvement in public schools and no impact on teacher effort or learning outcomes in those schools. • Olken (2007).e It was found that increasing grassroots participation in the monitoring of Indonesian village road projects had little average impact, reducing missing expenditures only in situations with limited free-rider problems and limited elite capture. • Ferraz and Finan (2008).f Disclosure of corruption in local governments in Brazil reduced the likelihood of the incum- bent reelection, and this effect was more pronounced in municipalities with radio stations. • Wantchekon (2009).g Town hall meetings in Benin, during which candidates present specific and informed policy proposals, have a positive effect on turnout and electoral support for the candidates—as opposed to a strategy based on campaign rallies with targeted or clientelistic electoral promises. • Malesky et. al (2011).h Delegates whose activities in the national legislature were publicized extensively via Viet- nam’s highest profile online newspaper showed less participation and higher conformist behavior, and they were significantly less likely to be reelected than the control group. a. Pandey, Goyal, and Sundararaman 2009; b. Nguyen and Lassibille 2008; c. Bjorkman and Svensson 2009; d. Banerjee, Banerji, Duflo, Glennerster, and Khemani 2008; e. Olken 2007; f. Ferraz, Claudio, and Finan 2008; g. Wantchekon 2009; h. Malesky, Schuler, and Tran 2011. 52 ANNEX 5. METHODOLOGICAL ISSUES IN ASSESSING RTI EFFECTIVENESS AND IMPACT One of the comparatively better-researched aspects of technologies and a broader attitudinal change in govern- the connection between more information and better ac- ment. Journalists felt that while government had opened countability is the impact of community information and up to an extent, opacity persisted. monitoring on a range of service-delivery outcomes—pri- The study found some impact on accountability and marily in education and health. Box A.5.1 summarizes little evidence of impact in the other four areas: quality some of these findings. of advice or records for government decision-making were A research project undertaken by the University Col- not significantly different; there was little impact on pub- lege London163 analyzed the statement of policy-makers to lic understanding of government decision-making; evidence identify six broad objectives of FOIA and used a combina- on increasing political participation was sparse, with less tion of a review of official literature and records, interviews than 0.1 percent of the population making FOI requests, with officials and other stakeholders, an online survey of and many of those who did were already engaged in the RTI requesters, and an analysis of media articles to assess political process; and little change in trust levels. In an- its impact on these objectives. First, regarding transparency, other perceptions survey of mid-level federal public officials a majority of the official respondents felt that FOIA made undertaken in Mexico by IFAI in 2007, 59 percent of the public authorities more transparent and led to greater pro- respondents believed that Mexico’s 2002 transparency law active disclosure in such areas as expenses, details of gifts, had contributed to increase transparency and access to in- salaries, overseas travel, and hospitality. But the shift toward formation, but only 7 percent said that it had resulted in openness was not only FOIA-related; it was also connected reduced corruption.164 to other drivers, such as information and communications 53 Annex 6 . Findings on Use and Compliance This annex summarizes the findings from the study on use • In Peru, according to PCM data, the number of requests of RTI, responsiveness, and proactive disclosure. increased from about 56,000 in 2004 to 68,000 in 2010. • In Albania and Moldova, the absence of monitoring Request Volumes data makes it difficult to assess the absolute number of requests, but interviews suggested that awareness and In the five countries where data was available (for differ- use of RTI was very limited. ent years since the passage of the law), there was a fairly • In Uganda, apart from civil society and media groups significant increase in the number of RTI requests filed filing requests, often to test the efficacy of the ATIA, year-to-year since the institution of the law. It is difficult there was no evidence of use by the general population. to get a measure of the extent of the use of RTI in three of the countries—Uganda, Albania, and Moldova. Data While the study could not access data on regional re- about the number of requests for information that are filed quests, if requests at the subnational level are taken into as RTI requests are not available at either the state or cen- account, the numbers are likely to increase substantially. In tral levels. India, for example, while the number of requests in 200708 was about 250,000 nationally, taking into account requests • In Mexico, the number of RTI requests to federal gov- at both the central and state levels, the number was more ernment entities increased from 8,000 in 2003 to about than three times higher, approximately 800,000. A recent 50,000 in 2009. A large percentage of requests involved study in the United Kingdom found that while requests to information generated by the agencies, such as conces- the central government had not risen significantly in recent sions, legal procedures, and statistics. Other requests years, requests to local governments had increased dramati- include personal data, remuneration of public officials, cally from 2005 to 2009. In 2005, according to the survey, subsidy programs, list of beneficiaries, and eligibility there were approximately 25,000 requests to the central requirements. Requests were also made on agencies’ governmentand 60,000 requests to local government. In activities, such as work plans, project results, and pro- 2009, there were 40,000 requests to the central govern- curement information—this could have been by civil ment and 165,000 requests to the local government.165 society groups. These numbers might not exactly correspond to the num- • In India, statistics showed a steady increase in RTI ap- ber of people filing requests because many people—espe- plications to the central government from about 24,000 cially CSO representatives and journalists—are likely to in 2005–06 to 260,000 in 2007–08. The RaaG study file several requests. But they are a close approximation. showed that between 2005 and 2008, 2.5 million peo- In countries where people might not be able to get ac- ple submitted RTI requests across the country (includ- cess to even basic public information, an effectively-en- ing central and subnational) or about 800,000 per year. forced right can be useful to gain access to information that • In the United Kingdom, official data showed that the might otherwise be difficult to obtain. But they also suggest number of requests has grown steadily from approxi- that RTI is a specialized instrument. As a percentage of the mately 16,000 in 2006 to 44,000 in 2010. population, the request numbers are almost miniscule, in the 0.03–.04 percent range. And this is not reflective of • In Romania, official data showed that the number of lower capacity or the newness of the law. The annual num- requests has remained consistent at approximately ber of RTI requests in even higher capacity countries with 600,000 since the law was adopted. a longer experience base in the implementation of the law 55 ANNEXES is still very small, ranging from 0.1–0.4 percent. Hazell and • subsidy programs, list of beneficiaries, and eligibility re- Worthy (2010) point out that by definition, FOI requesters quirements; and are extraordinary—they represent a small fraction of the • activities of agencies, such as work plans, project re- population and are usually activists or professionals. sults, and procurement information, which could have There can be several reasons for this relatively low been by civil society groups. percentage of requests to population. Information man- In Romania, a large number of the requests were related agement and information sharing is a general government to legislative acts, a reflection of the absence of adequate function, and people might not need to use RTI to gain measures for proactive disclosure. Information on budget access to most of the information they need. In most coun- allocations and the use of public funds represented about tries, citizens might be able to get personal information 25–30 percent of the information requested and could (about, for example, a health benefit or an adverse decision have been from civil society groups. In Albania, interview- on school admission or immigration status) without neces- ees suggested that almost 90 percent of the requests are for sarily invoking the law. Further, filing RTI requests is fairly basic information like access to drugs or medicine, proce- technically demanding, and even in countries with estab- dures to get diploma accreditation, and deadlines for appli- lished disclosure laws, making a request requires technical cation for the universities, possibly arising from the absence understanding of the process. Since RTI laws provide ac- of proactive disclosure in these areas. cess to “records,� prior knowledge about the specific kinds of records required is important. The U.K. study166 also found that in 2009, individu- als accounted for 37 percent of local government requests, businesses accounted for 22 percent, and the media ac- Who Uses RTI and for What? counted for 33 percent. Over the past five years, the per- centage of individuals and businesses making requests had Requests for information can be separated into three broad dipped slightly, while the percentage of media making local categories. The first is personal requests for information re- requests had risen significantly, from 11 percent. While quested by citizens. For citizens and communities in many users who were surveyed generally thought that RTI leg- countries, where officials tend to be unresponsive to even islation at the local level had significantly increased the basic information about their personal benefits, the exis- transparency of local government, officials felt that the leg- tence of an RTI law can be important. The second broad islation had only had little impact. set of requests is on information about government deci- sion making, likely to be made by NGOs, advocacy groups, In order to prevent discrimination among different re- and media outlets. The effectiveness with which civil soci- quester categories, most laws are “requester-blind,� provid- ety groups are able to use RTI for these purposes determines ing for anonymous requests, but some data emerged on the its usefulness as an instrument for social accountability. socioeconomic and demographic characteristics of users The third big category of users is of businesses, with an in- from surveys, interviews, and monitoring data when avail- terest both in personal and business environment-related able. In India, for instance, the RaaG survey showed that information. awareness is very limited among the rural and poorer sec- tions and that men file many more requests than women. Where requests are monitored, some data are available But civil society groups have filed requests on behalf of poor on the kinds of information that are requested under RTI. and rural communities. In Mexico, the study was only able In Mexico, the reports categorize the types of information to access data collected by IFAI based on self-reporting of requests as: user data by the requesters at the central level, and the pic- • information generated by the agencies, such as conces- ture that emerged is that of fairly educated and technically sions, legal procedures, and statistics; savvy requesters. • personal data and remuneration of public officials; 56 ANNEX 6. FINDINGS ON USE AND COMPLIANCE Other research has found that, in the more-developed Responsiveness countries, the most frequent categories of requesters are pri- vate citizens, the media, and businesses.167 In the United In the five countries, where data was available, there was a States, a report by the Coalition of Journalists for Open fairly high level of procedural compliance. Government168 that analyzed 6,439 FOIA requests to 11 cabinet-level departments and six agencies in 2005, found • In Mexico, between 2003 and 2010, 96.9 percent of that over 60 percent of requests came from the private sec- the requests received a response: either the information tor, including professional data brokers working on behalf was provided, a reason for refusal was provided, or the of clients, seeking information such as asbestos levels on requestor was directed to where the information could navy ships, cockpit recordings from crashed airliners, and be found. The administration also responded to public background data on prospective employees. The rise of pro- information requests in 12.8 days, well below the limit fessional data brokers is thought to be due to the desire of of 20 days established by the law. These responsiveness companies to mask their interests in obtaining particular rates are comparable to higher-income countries, which pieces of information, and it is somewhat controversial be- of course have greater capacity and resources. cause it is not necessarily the type of government account- • In the United Kingdom, data suggested that between ability-driven FOIA usage envisioned by the drafters and 75–90 percent of all requests receive a response within supporters of the law. The second-largest group of request- the required timeframes, and among all monitored bod- ers, at roughly 33 percent, was comprised of private citizens ies, 91 percent of requests are handled within the 20- seeking a wide variety of personal and other information. day deadline or with a permitted extension. Requests from the media accounted for only 6 percent of the total. CJOG noted that many journalists felt that the In Romania, according to government data, central and long wait times for official response made FOIA less useful local public institutions respond favorably to 95–98 per- for regular newsgathering than it was for longer investiga- cent of the requests received,171 but interviewees from the tive pieces. Another earlier study of U.S. usage of FOIA monitoring agency suggested that agencies might be inflat- found that 40 percent of total requests came from corpora- ing figures to look good. tions and 25 percent from lawyers.169 In India, based on RTI applications filed by RaaG, the An earlier survey of Canada, New Zealand, and Aus- central government has a compliance rate of 81 percent. tralia also showed a high percentage of personal requests Government data also suggests that a relatively small num- as opposed to accountability-related requests. Statistics ber of requests have been rejected by public authorities in from Australia and Canada showed that between 80 and the central government, although data on the kinds of in- 90 percent of all requests were for access to personal files, formation requests that have been rejected is unavailable. with 90 percent of Canadian requests being received by In Peru, the compliance rate for processed requests— just five institutions: efense (from servicemen), corrections that is, all requests responded to with a positive or nega- (from prisoners), archives (former servicemen and public tive answer—over the last few years has been around 90–95 servants), police (criminal records), and employment and percent. Immigration (immigration records).170 While this type of usage does not preclude transparency outcomes, it suggests In Uganda, Albania, and Moldova, civil society groups a focus on immediate personal—and, in the case of civil regularly attempted to test the effectiveness of RTI by float- servants, personnel—matters rather than broader govern- ing requests using it. But in all three countries, according ment accountability goals. to interviews conducted with these groups, they met with no or limited success. In Moldova, according to a study conducted by Access Info, a local NGO focused on RTI, rates of response were very low, although there was a marginal improvement in 57 ANNEXES recent years from 22.6 percent in 2007 to 34.6 percent in The law usually requires that information needs to be 2010.172 In interviews with public officials across ministries, disclosed “in the format in which it exists.� In many case they claimed that responses to RTI requests are processed study countries, interviewees suggested that agencies reply in a regular manner and that there is adequate institutional to requests by issuing information that is difficult to in- capacity to process routine requests. terpret; published in PDF format that is difficult to reuse; or lengthy, with pages of legalistic argumentation that is In Albania, civil society surveys, such as that by the Cit- virtually impossible to decipher. Budget reports might be izens Advocacy Office173 suggest that information requests released at a high level of aggregation. are often ignored, there are delays in answering requests, and there is absenteeism among information officers. An- A number of global studies have been recently under- other report found numerous cases of administrative silence taken that attempt to look at the level of responsiveness or mute refusals.174 by officials to information requests. In 2004, a study con- ducted by the Open Society Justice Initiative called “Trans- In Uganda, while the 2005 ATIA was not used by citi- parency and Silence�175 submitted a total of 140 requests zens for the most part, a study carried out by the local NGO in 14 countries.176 The “Ask Your Government!� study was Hurinet on the generic levels of information requests and compliance showed that of the survey participants that had led by Access Info Europe, the Centre for Law and De- requested information from a public institution—police, mocracy, and the International Budget Partnership, with local government, and the ministry of Education—as many cooperation from local civil society organizations who sub- as 70 percent had not received a response. In a study car- mitted six questions about budget openness in three the- ried out by the Uganda Debt Network, requests were sent matic areas: maternal health, development assistance, and to different ministries over five months; the energy ministry the environment. A total of 1,061 requests for information was the only one that even acknowledged receipt of the were made to 80 countries, accompanied by phone calls, request, and the education ministry eventually provided additional letters, faxes, emails, and sometimes personal the information requested for. Organizations such as the visits. Data from these studies provide an interesting com- Anti Corruption Coalition Uganda have floated numerous plement to the findings from the case studies. These global requests under ATIA, with a lot of back-and-forth com- studies typically use the methodology of floating informa- munication with ministries, but no release of information. tion requests to government agencies across countries and drawing comparative conclusions based on responsiveness. But even in countries with high officially-reported com- pliance rates, these numbers only tell part of the story. For These studies suggest find that responsiveness across instance, disaggregation by the responsiveness on specific countries continues to be low. Less than half of the requests kinds of requests as well as specific requesters will enable resulted in information being provided. In the Open Bud- a more comprehensive understanding of the biases of of- get study, only one in four requests led to full information ficials. Noncompliance could range from illegal measures being provided, and only 45 percent yielded any informa- like refusing to respond properly to requests, deliberate tion at all. Fully 42 percent of the requests met with re- nonrecording of information, or the destruction of infor- sponses that were not compliant with RTI standards. The mation to deliberately responding only at the end of the level of mute refusals was a high 38 percent, even after up established timelines and only making available minimal to three attempts to get a response. They also found that information. On specific requests that might be considered there is a positive correlation between having an RTI law more “sensitive,� officials often stonewall requests. Com- and responsiveness. Countries with RTI laws were more pliance rates might also denote procedural compliance to responsive than countries that did not. The case of Chile routine requests, but there might be resistance to comple- is interesting. When the “Transparency and Silence� study tion of requests that were of a particularly sensitive nature. was carried out in 2004, Chile did not have an RTI law and High compliance rates might also mask the many subtle performed badly, with a level of mute refusals at 69 percent. mechanisms public officials use to deny information re- In the intervening time, civil society has been galvanized, quests while still formally complying with the law. and Chile has both adopted a law and recognized RTI in 58 ANNEX 6. FINDINGS ON USE AND COMPLIANCE its constitution. There was a huge improvement in the per- major challenges. In India, for instance, while several de- formance in Chile in the “6 Question Campaign,� where it partments put information enumerated in Section 4—the provided positive answers in response to five out of the six proactive disclosure section under the RTI Act—online, requests. maintaining these websites became a challenge. Informa- tion tended to become out-of-date fairly quickly. For exam- ple, the RaaG study found that while 65 percent of urban Proactive Disclosure public authorities had published the details of their orga- nizations on their websites, only 45 percent had published In the United Kingdom, the goal of establishing a virtuous information on public information officers (PIOs), and circle between publication schemes and the right of access only 25 percent had published information on budgets and has been realized to some extent. Pressure from civil society salaries. PIOs are also often unaware of their obligations organizations has encouraged the proactive, routine publi- to update and upload this information. Responding to the cation of some information that was initially handled on a case-by-case basis—for instance, information such as sal- PWC survey, 43 percent of PIOs reported that they were ary information for officials employed by local authorities not aware of the proactive disclosure provisions. All four who earn over £100,000. FOIA has also been successful in departments assessed under the India case have RTI links facilitating the disclosure of information about how public on their websites, providing citizens with some basic infor- funds are spent because of a combination of pressure from mation on the RTI, the list of PIOs, and various other kinds organizations such as the Taxpayer’s Alliance, which was of information as mandated by the law, although informa- founded in 2005 partly to take advantage of the opportuni- tion gaps remain. In Romania, websites lacked interactiv- ties offered by the FOIA, and a series of decisions by the ity, and the quality of information was not user-friendly.177 commissioner and the tribunal that restricted the oppor- While core government ministries published information tunities to withhold this information. For instance, when required under FOIA, ten years after the law was passed, salary information of public officials was first requested, this other publicly-funded entities, such as public schools and information was considered to be sensitive by authorities hospitals, were still not aware of their obligations. such as local councils. Over time, a consensus has emerged In Albania, while progress is being made in some that information concerning salaries of senior staff are le- areas, such as electronic procurement and publication of gitimate matters of public interest. Many councils now pro- procurement notices and tender dossiers on the website actively publish this information on their websites, and the of the Public Procurement Agency (PPA), publication of current conservative government’s transparency agenda administrative instructions by the sectoral ministries is in- has extended this even further by mandating the proac- adequate, according to the 2009 CPII monitoring report. tive disclosure of all local government spending above According to another study conducted by CPII in June £500 as well as a wide range of data on central government 2010, aside from Tirana Municipality, none of Albanian spending. municipalities publish their city council’s decisions on of- But in most countries, information disclosed under ficial websites or local gazettes. Civil society pressure can proactive disclosure provisions face two sets of problems: have some effect: an “extra� edition of “138 Council of sustainability and incompleteness. Both sustainability and Ministers Decisions� was published following the publica- completeness of proactive disclosure provisions have been tion of the CPII study.178 59 Annex 7 . Examples of Use of RTI for Governance and Service Delivery Ration Cards—India company delivering the water. The request brought to light A field experiment found that slum dwellers who submit- the fact that there was no such agreement or contract—a ted information requests under India’s RTI received their breach of South Africa’s public finance legislation; it re- ration cards almost as fast as those paying “speed-money� sulted in the municipality being reported to the Auditor to the administration (four months versus two-and-a-half General for investigation.181 months on average), while most of those that applied for Water Treatment—Mexico the card in the standard prescribed manner did not receive In 2006, Maderas del Pueblo del Sureste—a nongovern- the card during the one-year window in which the research mental environmental organization supporting indigenous was conducted.179 people and rural communities in Chiapas, Mexico—filed Land Entitlements—India access to information requests using the federal transpar- In 2006, a coalition of NGOs in Orissa called Soochana ency law, seeking information about a sewage project that Adhikar Manch mounted a campaign that resulted in was negatively impacting a village that was receiving waste 42,000 FOIA requests being submitted to authorities. One from a neighboring town and had no access to clean water. of the submissions made during the campaign related to an Information released through these requests showed that application filed in 2002 by thirty-two landless claimants to the water treatment system was not properly designed and receive their land entitlements; within a few days, twenty- needed a filter system that had not been installed. The Cin- six of them received their land allocations. talapa sewage project was halted, and authorities publicly acknowledged that changes had to be made to ensure water Public School Access—India was properly treated.182 In Delhi, a NGO helped low-income families obtain access to a public school by placing RTI requests about the avail- Farm Subsidies—Mexico ability of seats for the poor and eligibility norms.180 The in- In 2007, a Mexican NGO—FUNDAR—requested from formation obtained helped prove that the school in question the Ministry of Agriculture the list of beneficiaries of PRO- was not making places available for low-income students as CAMPO, the largest federal farm subsidy program in the mandated by law; as a result, more low-income students country that is designed to support the poorest farmers and were accepted, giving them access to public education. reduce inequality in the rural sector. The Ministry of Ag- riculture did respond to this information request, but the Municial Water—South Africa information was incomplete and issued in unreadable for- Villagers in Emkhandlwini used South Africa’s FOIA law, mats. When more complete and legible information was with the help of the Open Democracy Advice Centre, to finally made available at the direction of the information request the minutes of council meetings about the provision commission—IFAI—it showed that the bulk of farm sub- of water, the municipality’s integrated development plan sidies were not being allocated to the country’s poorest (IDP), and its budget. The documents, which were released and smallest farmers but instead to the richest and most after a six-month delay, showed that the village was sup- productive farmers, and there was evidence of nepotism posed to receive access to clean water. Villagers were able and patronage. FUNDAR and other NGOs created an on- to apply pressure on the municipality by getting the media line repository of this information183 to keep pressure on to cover the issue. In response, the municipality installed the ministry. The Minister of Agriculture was removed fixed water tanks and delivered mobile ones in the commu- from his post, but no changes to the list of beneficiaries nity. When the mobile water tank supply became erratic, or to the program’s rules of operation were made. When villagers utilized the FOIA again, this time to request the subsequent investigations revealed continuing evidence of service-level agreement between the municipality and the corruption and nepotism, more resignations followed, and 61 ANNEXES the government imposed ceilings on eligibility for subsi- related to the customs department, the ministry refused to dies. But abuses have continued.184 Although the actions reveal financial information on these transfers. In 2005, the on PROCAMPO have resulted in more widespread audits IFAI ruled that the information had to be released because and study of the program, including ones by international the origin of the resources was public even if invested in agencies, the information obtained through the transpar- a “private� trust fund. Information subsequently released ency law was crucial in the uncovering of corruption and revealed that the trust fund managed resources in a highly misallocation of subsidies. discretionary manner and that there were resources miss- ing to the tune of US$750 million. No public official was Municipal Contracts—India sanctioned. As a result of this scandal, the reforms to the In 2002, information procured under the Delhi RTI Act Transparency Law proposed by a Congressional Committee 2001 by Parivartan, a Delhi-based NGO, as well as sub- in 2011 included a provision forbidding trust funds from sequent public hearings, revealed massive corruption and appealing to Bank secrecy laws. However, these reforms embezzlement of funds in 64 of the 68 Delhi municipal have not yet been approved by the congress.186 corporation contracts; Of the Rs. 13 million officially sanctioned for improving civic amenities, items worth ap- Corruption—South Africa187 proximately Rs. 7 million were nonexistent. The incident After several reports on corruption in provincial adminis- prompted the local municipal councilor to offer full trans- tration in South Africa in the 1990s, a civil society group parency in public works programs, and the MCD agreed to created and made publicly available a database of audit a series of corrective measures, like displaying information queries, which demonstrated that although 90 percent of about public works projects at worksites and in offices and the budget was questioned in the official audit, corrective local communities. The Delhi high court directed the po- action was taken in only 10 percent of cases. The findings, lice to investigate allegations of corruption, but the police which were published on the Internet, were also publicized were unable to collect evidence. in the mainstream media. After a cabinet team was estab- lished to follow up on the findings, the portion of the budget Diversion of Funds—Mexico with audit disclaimers decreased to 54 percent, but after the In 2002, FUNDAR was able to procure information from team was disbanded, the portion of the budget questioned in the Ministry of Health on the arbitrary allocation of spe- audit rose again to nearly 90 percent. cial funds earmarked for the purchase of retro-viral HIV/ AIDS medications to an anti-abortion NGO—Pro-Vida. Public Works—India188 Together with a coalition of six other civil society orga- The Indian state of Rajasthan passed an RTI law in 2000 nizations, FUNDAR requested information on all the fi- requiring similar disclosure of information by public offi- nancial reports Pro-Vida had submitted. The information cials of the state. In 2001, suspicion of diversion of public showed numerous irregularities in the use of funds, such as works funds caused citizens to protest. The protest resulted payments to fictitious organizations, frivolous expenditures, in the exposure at a public hearing of what amounted to excessive expenditures on publicity campaigns, and prefer- US$140,000 in fraud by a government official. However, as ential diversion of funds to Pro-Vida. Following a media of 2006, there had yet to be any criminal charges filed in con- campaign, Pro-Vida was forced to return the funds received nection with this matter. from the Ministry, was banned from receiving public re- Corruption in Medicine Procurement—India189 sources, and was charged a fine of 13 million pesos, but no In 2006, RTI applications with the PIO of South Eastern public official was investigated or sanctioned.185 Coal Fields Ltd. (SECL) in Madhya Pradesh and Chhat- Trust Funds—Mexico tisgarh procured information through the RTI Act that In Mexico, trust funds have traditionally been very resis- revealed that fictitious companies were listed as suppliers tant to disclosing information about financial operations, of luxury items in the name of procurement of medical sup- arguing that they are not technically public institutions plies, and that there were overcharges. These details were and are thus protected by bank secrecy laws. When the published in a local magazine, after which the vigilance Supreme Audit Institution identified a significant devia- unit instituted a formal investigation and action against tion of resources in financial transfers to some trust funds the officials concerned was initiated. 62 ANNEX 7. EXAMPLES OF USE OF RTI FOR GOVERNANCE AND SERVICE DELIVERY Public Procurement Contracts—Romania of contracts, the existence of a guarantee clause, the num- In Romania, when the prominent NGO, APADOR-CH ber of times that guarantee clause was invoked and against requested information on public procurement contracts for which companies, and copies of the contracts themselves. road construction and repairs during 2000–05 from Bucha- In response, the Bucharest Municipality, invoking con- rest Municipality, the response only communicated the list fidentiality clauses, communicated about the concluded of contracts concluded; the municipality refused to send contracts but refused to send copies of them; all other ques- copies of the contracts, invoking confidentiality clauses tions were left answered. The court found that the way in as the reason. The courts found that the way that public which public money is spent—including through procure- money is spent, including through procurement, is infor- ment—is information in the public interest that cannot be mation that is in the public interest. This was the first time “hidden� from citizens because of confidentiality clauses. a court found that confidentiality clauses in public procure- This was the first time a court found that confidentiality ment contracts do not supersede the public’s right to know. clauses in public procurement contracts do not supersede the public’s right to know. Parliamentary Expenses—United Kingdom In the United Kingdom, several MPs have been convicted, Court Fees—Romania reprimanded, or suspended in 2010, as a result of claims APADOR-CH requested information from the MoF on they made under the parliamentary expenses system that the number of applications to waive court stamp fees and were subsequently judged to have been illegal or unethi- about which payment exemptions, reductions, reschedul- cal. The origin of this scandal was a series of FOI requests ing, postponement, or other subsidies had been granted be- lodged by NGOs and investigative journalists that were tween 1990–2002 as well as how many challenges against resisted by Parliament for a considerable period and were court fee amounts were received and approved in the same even the subject of an aborted attempt to amend the FOIA period. Initially, the MoF did not respond to the request. to exempt the material. Although the details were initially When APADOR-CH filed an administrative appeal, the leaked to the Telegraph rather than released under the MoF denied the information on the grounds that they did FOIA, this leak occurred only when it became clear the not possess it, that processing the data would take longer information would be released. than 30 days—the legal time limit for answering FOIA Public University Websites—Romania requests, and that the data was personal and not of pub- In Romania, SAR, a Bucharest-based think tank,190 ran a lic interest. In response, APADOR-CH took the MoF to series of projects for approximately 5 years to measure the court and throughout the appeals process, the courts found transparency and quality of public information on public in favor of APADOR-CH, ruling that the requested in- university websites. Evidence from the second round of formation is not personal, that it is of public interest, and evaluations of the same universities showed marked im- that it is the MoFs responsibility to organize itself in such provement as a result of this monitoring. The same exercise a way as to be able to provide answers to public infor- was performed in several rounds on a sample of around 300 mation requests. After the verdict, the MoF still did not “core� public authorities (ministries, territorial agencies, fully comply with the court decision in a timely manner. and local governments), either by testing them on their APADOR-CH took MoF to court yet again. The court’s compliance with the general FOIA provisions or by explic- final decision found that “administrative abuse is obvious� itly requesting more “sensitive� information, such as the and that the MoF had to pay APADOR-CH compensatory full record of public procurement from the previous budget (“moral�) damages for not complying with the initial court year. In all cases, rankings were prepared and published by sentence in time and for violating APADOR-CH’s right to sector and institution, and these became widely debated in information.191 the national and local media. Surveillance—Romania Road Construction and Repairs—Romania One of the most interesting and high-profile cases of using APADOR-CH requested information on public procure- FOIA in sensitive areas is APADOR-CH’s request for infor- ment contracts for road construction and repairs during mation from the Romanian Secret Services (or Romanian 2000–05 from Bucharest Municipality, including the value Information Services—SRI) and the General Prosecutor’s 63 ANNEXES Office (GPO) on secret surveillance of Romanian citizens in a timely fashion, demonstrating that court decisions that was requested, approved, conducted, and which resulted have impacts on agency behavior by clarifying what kind in the prosecution and conviction of the monitored individ- of information must be released and by changing agency uals.192 Both the SRI and the GPO refused to release the in- expectations about how FOIA will be enforced. formation on the grounds that it was a state secret. The court Advertising Contracts—Romania found that, according to the Law on Classified Information, In another case, the Center for Independent Journalism all information on activities of the SRI is classified, rejecting (CJI) requested information on public institutions whose APADOR-CH’s argument that the SRI must prove that the advertising contracts were approved by the prime minister release of information would harm national security and that and the minutes of meetings in which the awarding of such the Law on Classified Information forbids the classification contracts was discussed. This was a reaction to a press article of information in order to cover up breaches of the law— about a memo issued by the then-Prime Minister Adrian which would mean that information revealing possible un- Nastase noting that all advertising contracts of ministries lawful surveillance cannot be classified. This case highlights required his prior approval, presumably in order to direct the limits of FOIA as well as the contradictions between the them toward supporters of the ruling party. “Avoidance Law on Classified Information, the law governing the SRI, strategies� employed by the prime minister’s office included and the FOIA. Even though APADOR-CH continued its the issuing of a press release that was then presented as an advocacy for aligning these laws with the FOIA, these limi- answer to the information request despite the fact that it did tations persisted. not provide any of the requested information, questioning However, the lawsuit against the GPO achieved better the legal standing of the plaintiff and of the Chancellery of results. Even though the initial administrative appeal was the Prime Minister (CPM) itself to initiate and stand trial, rejected, all courts involved in the process—including the and arguing that the information requested was not in the Supreme Court of Justice—found that the information re- public interest was classified. After a first court finding in quested could not be classified because it represented ag- favor of CJI, the CPM reached an agreement with the NGO gregated statistical data, and its release would not endanger and released the information. national security. This represented a significant break- Conflict of Interest—Romania through because it was the first time that information on In 2004, an individual member of APADOR-CH requested the secret surveillance of persons was made public by Ro- from the Prime Minister’s Control Department informa- manian authorities. However, even after the final verdict tion about the number of controls for conflict of interest of the court, the GPO did not provide the full information performed by the institution, the persons verified, and the requested by APADOR-CH, and what information was re- outcomes for each case. The CPM refused to answer on the leased was delayed by over 200 days. But even this partial grounds that by the time the request for information was information was revealing, and APADOR-CH used it to received the control department no longer existed. After a further its human rights advocacy, for example, by point- court struck down this argument by finding that the CPM ing out that the prosecutor’s office never turned down a was responsible for communicating the information, the surveillance request received from the SRI, that there was a CPM changed its reason, arguing that the information was very large number of people under surveillance, that many not in the public interest. Again, the court struck down of these people were monitored over a long period of time, this argument, finding that the information was in the pub- and that only a small fraction of them were convicted of a lic interest and requiring CPM to release it; the court also crime as a result of the surveillance.193 When APADOR- awarded the defendant compensatory (“moral�) damages CH filed another request for information regarding secret for having her rights breached. surveillance of persons, GPO released partial information 64 Notes 1. Labeled in different countries as “Right to Information,� “Freedom 12. A local NGO, the Human Rights Network Uganda (HURINET- of Information,� and “Access to Information� laws. The term “free- U) conducted a study on more generic requests for information (not dom of information� was the most common label for the earlier set using ATIA), and found that 70 percent of the participants in the of laws, and also reflected the close link with freedom of expression. survey reported not having received a response (HURINET 2010). The idea of freedom is tied to the concept of liberties, or freedom 13. Citizens Advocacy Office 2003; Access-Info 2011. from restraints from censorship, for example. The idea of “right� on the other hand, connotes a positive right to public information. 14. These numbers might not correspond exactly to the number of The view of access to information as essential to enabling citizens to people filing requests because many people—especially CSO repre- monitor the performance of public agencies and hold them account- sentatives and journalists—are likely to file several requests, but, they able, suggest that access to information is more usefully characterized are a close approximation. In some countries, only data at the na- as a positive right, not a negative freedom. “Access to information� tional level could be accessed, and if requests at the subnational level is a more generic category, and while several laws are labelled “access- are taken into account, the statistics on use would be even higher. to-information laws,� it does not denote the same principle as a legal In India, for example, while the number of requests in 2007–08 was entitlement. Rather, the right is supposed to enable access. As more about 250,000, nationally—taking into account requests at both cen- developing countries view the idea of citizen access to information tral and state levels, the number was more than three times higher, as a fundamental right for participatory and open development, the approximately 800,000. term “right to information� has become more commonly used. This 15. Hazell and Worthy 2010. report uses the terms “right to information� in preference to the other terms, but it is synonymous with “freedom of information�). 16. Ibid. 2. Freedominfo.org 2012; Banisar 2006. 17. Hazell and Worthy 2010. Although there is little detailed analysis of how RTI is used in different countries, in developed countries, 3. Article 19 of the Universal Declaration of Human Rights recog- a majority of people can presumably get information—especially nized the right to information as a fundamental human right. Clause information on personal data—through simple requests for informa- 2 of Article 19 of the International Covenant on Civil and Political tion or through online databases rather than having to use RTI. In Rights states that the “right to freedom of expression� includes “free- developing countries, lower awareness about RTI might explain the dom to seek, receive and impart information and ideas of all kinds.� small percentage of requests. 4. Bennett 1997. 18. Satark Nagrik Sangathan 2007. Grigorescu 2002. 19. Journalists working for The Sunday Times (Heather Brooke and 5. Among them: The Carter Center, Open Society Initiative, Com- Jon Ungoed-Thomas) and the Sunday Telegraph (Ben Leapman). monwealth Human Rights Initiative, and Article 19. 20. Samia Costa Tavares 2007. http://www.deccanherald.com/con- 6. Jenkins and Goetz 1999. tent/184246/power-rti-act-cracking-land.html. 7. The Indian grassroots NGO, Mazdoor Kisan Shakti Sangathan 21. For instance, priorities might include focusing on implementing (MKSS), widely credited with being the driving force for the RTI minimum standards such as creating updated lists or registers; priori- movement in India, epitomizes this idea of “the right to know� as tizing archiving documents of the greatest value to citizens, such as fundamental to the “the right to live� as well as the exercise of socio- those related to budgets, education, and justice; and setting in place economic rights for better livelihoods and access to better services clear roles, responsibilities, and coordination mechanisms as another from the state. Roy and Dey 2004; Roy and Udupa 2010. The MKSS short-term priority. was founded by Aruna Roy, a retired Indian civil servant; Nikhil Dey, 22. As the 2009 Global Integrity Report pointed out, countries that a lawyer who left his studies in the United States to take up rural are among the largest recipients of international donor assistance activism; and Shankar Singh, an expert in rural communication. tend to have the largest implementation gap, suggesting that aid- 8. Lord Meghnad Desai at the Fourth Annual Convention on RTI, dependent countries might establish laws and institutions to meet New Delhi, October 2009, quoted in Roberts (2010: 3). donor requirements but do not necessarily implement them, which leads to a proliferation of legal and regulatory reforms on paper with 9. Narayanaswamy 2010. no actual benefit. 10. Banisar 2006; Bennett 1997; Mendel 2004, 2009, 2011; Pud- 23. Roberts (2006) points out that the countries that initially adopt- dephat 2009. The RTI Ratings Index framework developed by two ed disclosure laws had a long tradition of respecting citizen rights and leading institutions in this area—Access Info Europe and Center for the rule of law, a lively popular press, and healthy and independent Law and Democracy, and has developed a good framework to assess NGOs, and that the efficacy of RTI in countries with a poorer record the quality of legislation and provides a useful guiding framework on key governance dimensions, such as political and civil liberties, to identify the key dimensions on implementation. http://www.rti- rule of law, and accountable governance, had yet to be tested. rating.org/. 24. Kreimer 2008; Roberts 2006. 11. Hazell and Worthy 2010. 65 NOTES 25. http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTP 51. RaaG 2009; Access-Info 2009; Citizens Advocacy Office 2003. UBLICSECTORANDGOVERNANCE/0,,contentMDK:23253090~ 52. Interview with CSOs, September 2010. pagePK:148956~piPK:216618~theSitePK:286305,00.html. 53. Roberts 2006. Poor organization also made tracking down specific 26. In some instances, the right extends to all interested parties, pieces of data impossible. including noncitizens. 54. Neuman and Calland 2007. Institutional weakness, lack of 27. Weber 1922. resources, cutbacks in civil service, and poor maintenance have made 28. Government of India 2010. the management of records a major challenge in many of the case study countries. 29. Chapter III, IV, and V; RTI Act 2005. 55. RaaG 2009: 25. 30. As evident, for instance, in the file-notings controversy (see endnotes 57 and 80). 56. Access-Info Europe and the Open Knowledge Foundation 2011. 31. The original nodal agency was the Lord Chancellor’s Depart- 57. In India, the government attempted to introduce amendments ment, which was responsible for administering the courts and judicial to restrict from the purview of the law—�file notings�—notes that system. This was merged into the Department of Constitutional Af- officials make on policy files that are critical to understanding the fairs in June 2003, and then renamed the Ministry of Justice in 2007, policy-making process; it also attempted to introduce other amend- following the transfer of responsibility for probation, prisons, and ments. Venkatesan 2010. prevention of reoffending from the Home Office. 58. Over 40 percent of rural and 15 percent of urban applicants 32. Department for Constitutional Affairs 2002. polled in a survey reported facing harassment and threats. 33. The Council of Ministers repeatedly dragged its feet on pro- RaaG 2009. Safeguarding the Right to Information: Report of the mulgating the implementation regulations, even though these were People’s RTI Assessment 2008. Executive Summary.� Raag and Na- prepared and submitted by the department of information in charge tional Campaign for People’s Right to Information (NCPRI). http:// of implementing the 2005 ATIA. The regulations were finally passed rti-assessment.org/exe_summ_report.pdf in April 2011, gazetted in May, and published and publicly released 59. In New Zealand, research showed that straightforward, non-polit- in July 2011. Even after regulations were adopted after significant ically-sensitive requests resulted in on-time satisfactory compliance, delay, potentially high costs, procedural complexities—including while politically sensitive requests were delayed, transferred, and multiplicity of forms and processes, and lack of adequate guidance refused. Palmer 2007. for implementing agencies have been identified as particularly weak. World Resources Institute 2011. 60. Neuman 2009. 34. In interviews, officials in charge of monitoring implementation in 61. Ibid. agencies pointed out that agencies might be mis-reporting figures, but 62. Interview with officials, 2010. there are no resources to audit or cross-check their validity. Interview with DGS, August 2011. 63. For instance, in the United Kingdom, the ruling of the ICO can be appealed to the Information Rights Tribunal, a specialist quasi- 35. Bookman and Guerrero-Amparan 2009; Ministry of Justice 2011. judicial body. Further, appeals against the tribunal’s decisions can 36. In Albania, for instance, ministers continued to use political be appealed to the Administrative Appeals Chamber of the Upper spokespersons to deliver information to the media rather than ap- Tribunal. Finally, appeals against the decisions of the Administrative proaching information dissemination from the perspective of RTI. Appeals Chamber lie with the Court of Appeal. Such a complex structure requiring building capacity at multiple levels of the enforce- 37. Interview with CSOs, August 2010. ment system might not work in countries with more limited financial 38. Interview with officials, 2010. and human capacity. In Australia, a hybrid model is used, where requesters may appeal procedural shortcomings in relation to the 39. Cozonac, Guja, and Munteanu 2004. processing of their request to the ombudsman, but they have to file 40. RaaG 2009. an appeal with the Administrative Appeals Tribunal if they want to challenge the substantive decision by the agency in its application of 41. Centre for Development and Democratization of the Institutions the act’s exemptions. Requesters in Hungary have the option to go 2003. directly to court for a binding order or to the less powerful informa- 42. Probabilística 2007. tion ombudsman. Most choose the information ombudsman route 43. Interview with officials, August 2011. specifically because the courts take so long to determine cases. 44. RaaG 2009. 64. IFAI 2009. 45. Last published annual report (2010) compiled by the Department 65. Fox, Jiménez, and Haight 2009; Gozzo 2006. of Government Strategies; 69 percent in 2003; 92 percent in 2004; 66. There are only a few cases of the noncompliance being taken to 79 percent in 2007; and in more than 50 percent of cases in all year. appeals. One high-profile case, Tullow Oil case, demonstrated the 46. In India, the variations in the rules framed by the central and weakness of the judiciary in addressing RTI issues. In 2009, a number state governments have resulted in as many as 88 different RTI rules of CSOs filed a case against the government’s refusal to release currently in operation in India. certified copies of oil exploitation agreements because of alleged confidentiality clauses. The CSOs argued that the information was 47. RaaG 2009: 10. of public interest. But the chief magistrate said in his ruling that the 48. Probabilística 2007. petitioners had not proved the public benefit from disclosure of the information—a considerable misinterpretation of the law. Interview 49. Ministry of Justice 2011. with CSOs, September 2010. 50. Interview with DGS, August 2011. 66 NOTES 67. The record of judicial appeals on the Tullow Oil case, and more The early years of FOIA in the United States were marked by execu- broadly on accountability issues, shows that the judiciary does not tive branch distrust of the legislation, leading to amendments in have the specialist technical capacity to address ATIA issues. 1974, 1976, and 1986. More recently, following the attacks on the United States on September 11, 2001, the United States govern- 68. http://www.ipp.ro/jurisprudenta544/; Moraru 2011. ment tried to significantly narrow the scope of the FOIA. Waning of 69. Ibid. congressional oversight from 2001 to 2006 may have also led to the slowing of the processing of FOIA requests. FOIA request delays in 70. In 2003, the IFAI was allocated close to US$18 million. This 2005 increased by 11 percent over those in 2004, while the backlog budget increased minimally over the years. By 2009, the IFAI of unprocessed FOIA requests grew from 20 percent in 2004 to 31 obtained a little over US$22 million (official information obtained percent in 2005, even though the total number of requests dropped from the IFAI). (Relya 2009). 71. Ley Federal de Protección de Datos Personales en Posesión de los 80. In 2005, the Moldovan authorities drafted two pieces of legisla- Particulares, Chapter VI, Article 38. tion that severely limited access to information: one on state and 72. Central Information Commission 2006. service secrets and another new law on access to information. 73. A study found that the failure to impose penalties cost the gov- Media and NGO groups, supported by the international community, ernment Rs. 86 crores (Tiwari, Ruhi, and Anuja 2010). The Indian including the Organization for Security and Cooperation in Europe law provides for the imposition of financial penalties, compensation vehemently opposed the draft, which was subsequently withdrawn. In to applicants for any loss or detriment suffered, and disciplinary ac- 2005, coordinated action between civil society groups and the inter- tion against public information officers for noncompliance, but data national community prevented the government from adopting some suggests that less than two percent of penalties imposable under the controversial laws. The Information and Security Service proposed a RTI Act are actually imposed. draft “Law on State and Official Secrets,� which was approved by the government on December 2, 2004, and submitted to the parliament 74. The RTI law provides for a maximum penalty of Rs. 25,000, for approval. In addition, the Ministry of Informational Development whereas several high courts, such that of Calcutta and Gujarat, elaborated a draft law on information. Civil society groups, Transpar- impose a maximum penalty of Rs. 1,000. See Chauhan 2012. ency International Moldova, and Access-Info actively opposed these 75. In 2010, an executive decree gave the IFAI the power to sanction laws on the grounds that they would limit access to information. individuals who misuse the IFAI’s personal databases. Yet, it still does After coordinated pressure, the Moldovan government withdrew not have the power to sanction public officials who do not comply both draft laws from parliamentary consideration. with the IFAI’s decisions (Barajas 2010). In India, File notings are essentially the opinions and notes of civil 76. Kreimer 2008; Roberts 2006. servants on government files that sum up the decisions taken on a particular matter. The unease with the disclosure of file notings 77. Lindstedt and Naurin 2010. appears to extend to the highest levels of government, with even the 78. Access Info Europe and Centre for Law and Democracy 2011. president of India expressing his concerns about this to the prime In fact, a study assessing the level of responsiveness across countries minister in a string of official correspondence. In 2006, soon after the found that among countries with longer-standing RTI legislation, law was enacted, the DoPT issued a notice on its website stating that newer democracies performed better in terms of greater openness. file notings were not to be disclosed under the RTI Act. Civil society groups and activists were quick to protest this move and challenged 79. In countries ranging from the United States, to Ireland, to New the notice before the central and state information commissions who Zealand, there has emerged a pattern of initial enthusiasm and use of supported the view that notings could be accessed under the law. The RTI legislation, followed by official government retrenchment, either government prepared a draft RTI amendment bill, the main purpose through attempts to formally amend the law or through bureaucratic of which was to exclude file notings from the purview of the act. Civ- procedures designed to blunt the impact of politically-sensitive in- il society groups and leading RTI activists rallied against this bill and formation requested by civil society. This official resistance, whether launched a major campaign with the support of the media, success- manifested through delays, refusals, or other means, appears to be fully stalling the government from pushing through the amendment. more of a reaction to politically-sensitive requests filed by journalists Again, in 2009–10, in response to an RTI request, the government or opposition politicians rather than personal requests. confirmed that amendments to the act were under consideration to In Ireland, when newspapers published stories revealing official improve the functioning of the law and prevent its misuse. These impropriety and misconduct, there was evidence of officials “prere- included exemptions of frivolous and vexatious requests for informa- leasing� information requested by opposition politicians to friendly tion, discussions on policy decisions, and information from the office journalists. In Canada, in the 1980s, senior officials hid background of the chief justice of India. After civil society groups addressed papers, sometimes secretly shredded records, and resisted requests for letters to the prime minister and Sonia Gandhi to contest these pos- results of public opinion polls, and attempted to limit the effects of sible amendments, the government decided to shelve them, assuring the law in practice, shifting tasks to quasi-governmental organiza- activists that they would be considered only after consultations with tions not covered by RTI—thus excluding data from public exposure, a range of stakeholders. In this way, civil society groups have not and adopting elaborate procedures for handling politically-sensitive only played a crucial role in the passage of the legislation, but also as requests, which ultimately led to a delay in their processing (Roberts watchdogs of the RTI Act, remaining vigilant and responsive to any 2006). government pushback on the law. One analyst of New Zealand’s law concluded that there were in prac- 81. In 2006, the Albanian Helsinki Committee successfully inter- tice two laws: one that applied to straightforward, non-politically- vened, sending a letter to the speaker of the parliament requesting sensitive requests, which resulted in on-time satisfactory compliance; that she stop approval of the government-proposed amendments and one that applied to politically-sensitive requests, which were to Law 8503 that would have added more categories of restricted delayed, transferred, and refused. (Palmer 2007). information. In 2008, CSO groups, with support from Open Society Foundation Albania and the Centre for Development and Democra- 67 NOTES tization of Institutions, proposed amendments to Law 8503 to bring 100. The legal foundation for the ATIA was provided by the 1995 it more in line with advanced standards of European legislation. The Constitution that incorporated a guarantee on access to information amendments were prepared in open meetings with representatives of to all citizens. This came directly from the concerns of the progres- the public administration, members of parliament, interest groups, sive Uganda Constitutional Commission led by Justice Benjamin media, and civil society, but these efforts have still not succeeded. Odoki, for the human rights abuses of the preceding decades. Article In 2008, CSO groups, proposed amendments to Law 8503 to bring it 41 of the 1995 Constitution reflects the Commission’s belief in the more in line with advanced standards of European legislation. The importance of fundamental freedom of expression and the right of amendments were prepared in open meetings with representatives of every person to information as core to the rule of law and democracy: public administration, members of parliament, interest groups, media “Every citizen has a right of access to information in the possession and civil society. But the changes never took place. of the state or any other organ or agency of the state except where the release of the information is likely to prejudice the security or 82. Fox, Jiménez, and Haight 2009; Fox and Haight 2010; Jenkins sovereignty of the state or interfere with the right to privacy of any and Goetz 1999; Singh 2007; Sulaimani 2006. other person.� 83. APADOR-CH 2006. 101. Ghana, Sierra Leone, Kenya, Tanzania, Zambia in Africa, Sri 84. Such as FUNDAR, Colectivo por la Transparencia, Fundar, Free- Lanka, and the Philippines in Asia are a few examples. dom of Information—Mexico (LIMAC), and Center for Research 102. United Nations General Assembly Resolution 59(l). and Teaching in Economics (CIDE). 103. African Charter on Human and People’s Rights: http://www. 85. Such as Uganda Debt Network (UDN), Advocates Coalition achpr.org/english/_info/charter_en.html. for Development and Environment, HURINET, the Foundation for Human Rights Initiative, Anti-Corruption Coalition of Uganda, 104. European Convention for the Protection of Human Rights and the umbrella and the Coalition on the Freedom of Information and Fundamental Freedoms. http://www.echr.coe.int/nr/rdonlyres/ (COFI). d5cc24a7-dc13-4318-b457-5c9014916d7a/0/englishanglais.pdf. 86. Freedom House 2012. 105. American Convention on Human Rights. http://www.cidh.org/ Basicos/English/Basic3.American%20Convention.htm. 87. Ibid. 106. Chapter 2 of the 1776 Swedish Freedom of the Press Act en- 88. Jenkins and Goetz 1999. titled “On the Public Nature of Official Documents� provides for ac- 89. Such as Friends of the Earth, Greenpeace, and Campaign against cess provisions. http://www.riksdagen.se/templates/R_Page____6313. the Arms Trade. aspx. 90. Interviews with CSOs. 107. Roberts 2006. 91. For instance, the nodal coalition of health NGOs—UNHCO— 108. Szekely 2007. sits on the Health Advisory Committee and has access to policy 109. Florini 1999; Roberts 2006. making. 110. Bennett 1997, 2001; Grigorescu 2002. 92. In Africa, groups such as the Kampala-based African Freedom of Information Centre have encouraged media organizations to play 111. Roberts 2006. a more strategic—and therefore less overt—role in campaigns for 112. Interviews with stakeholders involved in the adoption of the greater openness. law. 93. Ackerman 2005. 113. Republic of Uganda 2005. 94. Fox 2007. 114. Roberts (2006) suggests that the very reasons that many devel- oping countries—what he, quoting Gunnar Myrdal, characterizes as 95. http://www.opegovernmentpartnership.org/. “soft states�—are adopting RTI legislation, even though they don’t 96. PETS, conducted initially in 1996, showed that when informa- have the preconditions of Western liberal democracies, is precisely tion about funding allocation to district schools was publicized because they are soft states—because they are more easily susceptible (for instance, in national and local newspapers), leakage of funds to pressures from international organizations. reduced from 74 to 20 percent—that is, more than 80 percent of the 115. For instance, auditing and monitoring of the declared assets of funds were received by local primary schools after the publication of elected officials have proven ineffective despite a strong asset disclo- data as opposed to about 26 percent before it was published. While sure legal regime. the Uganda PETS initiative is widely cited as a success, following attempts at PETS in other countries did not meet with the same suc- 116. Freedom House 2012. cess. Donor studies have concluded that local context is important 117. OSCE 2006. and governments may not be willing to share unfavorable findings with the public, as was demonstrated in a Tanzanian PETS project. 118. Article 19(1)(a). Sundet 2008: 8. 119. For a detailed history of the campaign for the right to infor- 97. World Bank 2004. mation see Singh 2007, 2009; Jenkins and Goetz 1999; Goetz and Jenkins 1999; and Banisar 2006. 98. World Bank 2007. Governance and Anti-corruption Strategy. 120. Singh 2010. 99. http://data.worldbank.org. 121. Certain provisions of the law came into immediate effect. 122. For a comprehensive overview and analysis of the formulation and approval of the transparency law in Mexico, see Luna Pla 2009: 61. 68 NOTES 123. Such as IDIS-Viitorul Foundation. Interview with Igor Mocanu. 150. Jenkins and Goetz 1999. 124. Law No. 982 from May 11, 2000, regarding access to infor- 151. Workers and Farmers Union. Roy and Dey 2004; Roy and mation, published in the Official Gazette No. 88–90, article 664 Udupa 2010. The MKSS was founded by Aruna Roy, a retired Indian from July 28, 2000. civil servant; Nikhil Dey, a lawyer who left his studies in the United States to take up rural activism; and Shankar Singh, an expert in 125. Neicovcen 2000. rural communication. 126. This law was subsequently modified by Law 27927 of February 4, 152. Hazell and Worthy 2010. 2003, systematized through a Single Revised Text, adopted through Supreme Decree, and later developed through regulations in Supreme 153. For example, CDDI in Albania and Access-Info in Moldova. Decree. 154. The RaaG study used a triangulation of methods, floating 127. www.cdep.ro/pls/proiecte/upl_pck.proiect?idp=2517. requests, conducting surveys, and accessing CIC data. Individual interviews were conducted with 18,918 persons across ten states and 128. Law 182/2002. Delhi, including 1,415 PIOs and heads of offices and departments. 129. Which derives its name from Article 19 of the Universal Decla- A total of 630 focus group discussions were organized in villages and ration of Human Rights. districts. The total number of participants in the study was 37,704 including 18,786 people who participated in these groups. More than 130. Mendel 2004. 1,000 public authority offices were inspected in both rural and urban 131. Mendel 2003: 96–8. areas. Over 800 RTI applications were filed with various public au- 132. Ostermann 2005. thorities across the country. Data from over 25,000 RTI applications was analyzed. Over 60 papers and magazines were analyzed for con- 133. Prime Minister 1969. tent and coverage. Over 5,000 case studies were extracted, depicting 134. Kettle 2010. successes, failures, and peculiarities of the RTI regime. 135. Singh 2010. 155. The RaaG study, for instance, was funded by Google Founda- tion. 136. Outlook India 2006; IBN Live 2006. 156. Islam 2003; Kaufmann and Bellver 2005. 137. Banisar 2005. 157. Kaufmann and Bellver 2005. 138. Law 28664 on the National Intelligence System and the Na- tional Intelligence Directorate. 158. Islam 2006. 139. For a comprehensive analysis of all state access-to-information 159. Reinikka and Svensson 2005. and transparency laws see, Diego Ernesto Díaz Iturbe, coord. “Métrica 160. Hubbard 2007. de la Transparencia.� CIDE, 2007. 161. Joshi 2010. 140. Views expressed in discussions with various stakeholders. 162. Ibid. 141. In these cases, the results are mixed. Very few institutions have established effective mechanisms to receive requests for information; 163. Hazell, Worthy, and Glover 2010; Worthy 2010. only the supreme court is credited with designing a good system. 164. Probabilística 2007. 142. http://articles.timesofindia.indiatimes.com/2010-04-13/in- 165. Worthy, Amos, Hazell, and Bourke 2011. dia/28130376_1_judiciary-chief-justice-information-act; http://ar- 166. Ibid. ticles.timesofindia.indiatimes.com/2010-01-13/india/28135423_1_cji- hc-verdict-judicial-independence; http://www.thehindu.com/news/ 167. Holsen 2007. national/article2013124.ece. 168. Coalition of Journalists for Open Government 2011. 143. George Akerlof, Michael Spence and Joseph Stiglitz received 169. Ibid. the Nobel Prize in 2001 for their analysis of how imperfect informa- tion lead to market failures. Also see Akerlof 1970; Spence 1973; and 170. Hazell 1989. Stigler 1961. 171. There is one discrepancy: the percentage of refusals went up 144. Islam 2006 and others. spectacularly, from 1–3 percent before 2006 to 39 percent in 2008, but no one from central institutions interviewed was able to explain 145. Stiglitz 2001. why. 146. World Bank 2004. 172. Access-Info 2011. 147. Ibid. 173. Citizens Advocacy Office 2003. 148. Jenkins and Goetz 1999. 174. Open Society Justice Initiative and Centre for Development 149. Accountability has been characterized as being comprised of and Democratization of the Institutions 2007. “answerability� on the one hand and “enforcement� on the other. 175. Open Society Justice Initiative 2006. Access to information by itself enables answerability or “the obliga- tion of public officials to inform about and to explain what they are 176. Argentina, Armenia, Bulgaria, Chile, France, Ghana, Kenya, doing.� But enforcement requires “…. accounting agencies to impose Macedonia, Mexico, Nigeria, Peru, Romania, South Africa, and sanctions on power holders who have violated their public duties.� Spain. Ackerman quoting Schedler 1999; Fox, 2007. 69 NOTES 177. For example, staff charts or budgets that are scanned and 191. Cases taken from APADOR-CH (2006) and from http://www. attached as graphic files have poor resolution; budgets cannot be apador.org/acc_cases.php?tip=8&order=date%20desc. converted into spreadsheets for analysis. 192. Specifically, APADOR-CH asked the Romanian Information 178. State Publication Center. Service General Prosecutor’s Office for information about requests for secret surveillance measures against members of political parties, 179. Peisakhin and Pinto 2010. civic associations, and journalists filed with the general prosecutor’s 180. McCall and Wilde 2007. office; warrants granted and extensions; total duration of surveil- lance in each case; commercial companies set up or taken over; and 181. Dimba 2008. their revenues and expenditures. Requests for secret surveillance 182. Martinez-Moralez 2009. measures filed with GPO between 1990 and 2002 on grounds of 183. http://cic.gov.in/bestpractices.htm. national security and criminal investigation; warrants granted and extensions; persons subject to secret surveillance; persons prosecuted; 184. www. subsidiosalcampo.org.mx. final convictions; persons notified of being under secret surveillance 185. Fox and Haight 2010: 140. after expiration of warrant; complaints against secret surveillance activities or warrants; number of those found legitimate; and longest 186. See Gozzo 2006. duration of an authorized surveillance. 187. El Universal, March 30, 2011. 193. “Of the 5,489 warrants for monitoring, only 1,451 persons were 189. Sundet 2008. sued and only 692 sentenced definitely. It results that the privacy of 4,500 persons was violated without any reason.� See http://bignews- 190. http://rti.gov.in/rticorner/studybypwc/progress_made.pdf. magazine.com/2003/12/apador-ch-discontent-with-information-from- 190. www.sar.org.ro. prosecution-office/. 70 References Access-Info. 2009. “Public Procurement: Transparency or Opacity, Bookman, Zachary, and Juan Pablo Guerrero-Amparan. 2009. Monitoring Report,� Access-Info, Chisinau. www.acces-info.org. “Two Steps Forward, One Step Back: Assessing the md/upload/akizitsii_ultima.pdf. Implementation of Mexico’s Freedom of Information Act.� Mexican Law Review. Nueva Series I (2), January–June. ———. 2011. “Access to Information and Transparency in the Decision-Making Process: Attitudes, Perceptions and Chauhan, Chetan. 2012. “High Courts Altering RTI Rules, Tendencies.� In Annual Monitoring Report. Chisinau: Access-Info. Penalties: Study.� Hindustan Times, June 17. http://www. hindustantimes.com/News-Feed/TopStories/High-courts- Access Info Europe and Center for Law and Democracy. 2011. altering-RTI-rules-penalties-Study/Article1-873792.aspx. “Ask Your Government! The 6 Question Campaign: A Comparative Analysis of Access to Budget Information in Citizens Advocacy Office (CAO). 2003. “Report of the Access to 80 Countries.� International Budget Partnership (September). Information.� Albania. http://www.access-info.org/en/budget-transparency/ Constitution Unit. 2011. “The Best Things in Life are Free: 200-6-question-campaign. How Much Does FOI Cost?� Blog. http://constitution-unit. Access-Info Europe and Open Knowledge Foundation. 2011. com/2011/04/11/the-best-things-in-life-are-free-how-much- “Beyond Access: Open Government Data and the Right to does-foi-cost/. (Re)Use Public Information.� January 7. http://www.access-info. Cozonac, Cornelia, Iacob Guja, and Petru Munteanu. 2004. org/documents/Access_Docs/Advancing/Beyond_Access_7_ Monitoring Access to Information in Moldova, 10–19. Chisinau: January_2011_web.pdf. Transparency International Moldova. www.transparency.md/ Ackerman, J. M. 2005. “Social Accountability in the Public Sector: Docs/TI_monitorizarea_informatie_ro.pdf. A Conceptual Discussion.� Social Development Papers 82, Central Information Commission. 2006. “Government of India, World Bank, Washington DC. New Delhi, Annual Report 2005–2006.� http://www.cic.gov.in/ Akerlof, George A. 1970. “The Market for ‘Lemons’: Quality AnnualReports/AR-200506/MainReport.pdf. Uncertainty and the Market Mechanism.� Quarterly Journal of Citizens Advocacy Office. 2003. “Report of the Access to Economics 84 (3): 488–500. Information.� Albania. APADOR-CH. 2006. Access to Information of Public Interest in Coalition of Journalists for Open Government. 2011. “Frequent Romania. Selected Cases. Vol II. Bilingual Edition. Bucharest, Filers: Businesses Make FOIA Their Business,� http://www.cjog. Romania: Open Society Justice Initiative. http://www.apador.org/ net/documents/who_uses_foia__cx.pdf. en/publicatii/foia_2006_en.pdf. Department for Constitutional Affairs. 2002. “Work of the Banisar, David. 2005. “Comments on the Moldovan Draft Law on Advisory Group on Implementation of the [Freedom of Information.� OSCE: The Representative on Freedom of the Media, Information] Act,� section 6. The Stationery Office, London. September. www.osce.org/fom/16546. Dimba, M. 2008. “Access to Information as a Tool for Socio- ———. 2006. Freedom of Information Around the World 2006: Economic Justice.� International Conference on Right to Public A Global Survey of Access to Government Information Information, Carter Center, February 26–29, Atlanta, Georgia. Laws. London: Privacy International. http://www. humanrightsinitiative.org/programs/ai/rti/international/laws_ Ferraz, Claudio, and Frederico Finan. 2008. “Exposing Corrupt papers/intl/global_foi_survey_2006.pdf. Politicians: The Effects of Brazil’s Publicly Released Audits on Electoral Outcomes,� Quarterly Journal of Economics Banerjee, A., R. Banerji, E. Duflo, R. Glennerster, and S. Khemani. 123(2): 703–746. 2008. “Pitfalls of Participatory Programs: Evidence from a Randomized Evaluation of Education in India.� Policy Research Florini, Ann. 1999. “Does the Invisible Hand Need a Transparent Working Paper 4584, Impact Evaluation Series 21, World Bank, Glove? The Politics of Transparency.� Paper prepared for the Washington, DC. Annual World Bank Conference on Development Economics, Carnegie Endowment for International Peace, Washington, DC. Bennett, C. J. 1997. “Understanding Ripple Effects: The Cross National Adoption of Policy Instruments for Bureaucratic Fox, Jonathan. 2007. “The Uncertain Relationship between Accountability.� Governance 10(3): 213–234. Transparency and Accountability.� Development in Practice 17(4): 663–671. ———. 2012. “The RTI Ratings Index Framework.� http://www. rti-rating.org. Fox, Jonathan, Carlos García Jiménez, and Libby Haight. 2009. “Rural Democratization in Mexico’s Deep South: Grassroots Bjorkman, Martina, and Jakob Svensson. 2009. “Power to the People: Right-to-Know Campaigns in Guerrero,� Journal of Peasant Evidence of a Randomized Field Experiment on Community- Studies 36 (2). Based Monitoring in Uganda.� Quarterly Journal of Economics 124: 2. 71 REFERENCES Fox, Jonathan, and Libby Haight. 2010. “Farm Subsidy Recipient Jenkins, Rob, and Anne Marie Goetz. 1999. “Accounts and Lists: A Case of Clear or Opaque Transparency.� In Subsidizing Accountability: Theoretical Implications of the Right-to- Inequality: Mexican Corn Policy since Nafta, edited by Jonathan Information Movement in India.� Third World Quarterly 20(3), Fox and Libby Haight. Woodrow Wilson International Center 603–622. http://www.archonfung.net/docs/pal218/JenkinsGoetz. for Scholars. AccountsAccount.pdf. Freedom House. 2012. “Freedom in the World Country Ratings, Joshi, Anuradha. 2010. “Review of Impact and Effectiveness of 1972–2012.� http://www.freedomhouse.org/report-types/ Transparency and Accountability Initiatives: Synthesis Report— freedom-world. Annex 1: Service Delivery. � Prepared for the Transparency and Accountability Initiative Workshop, October 14–15, Institute Freedominfo.org. 2012. “93 Countries Have FOI Regimes, Most of Development Studies. http://www.ids.ac.uk/files/dmfile/ Tallies Agree.� October 19. http://www.freedominfo.org/ IETAAnnex1ServicedeliveryJoshiFinal28Oct2010.pdf. 2012/10/93-countries-have-foi-regimes-most-tallies-agree. Kaufmann, D., and A. Bellver. 2005. “Transparenting Transparency: Government of India. 2010. Office Memorandum No. 2/10/2010-CR, Initial Empirics and Policy Applications.� World Bank Policy July 9. Ministry of Personnel, Public Grievances and Pensions, Research Working Paper, World Bank, Washington, DC. DoPT. Kettle, Martin. 2010. “Tony Blair Interview: The Full Transcript.� Grigorescu, Alex. 2002. “European Institutions and Unsuccessful Guardian September 1. Norm Transmission: the Case of Transparency.� International Politics 39(4): 467–489. Kreimer, S. F. 2008. “The Freedom of Information Act and the Ecology of Transparency.� Public Law and Legal Theory Gozzo, Gaia. 2006. “El caso PROVIDA: deudas pendientes con la Research Paper Series, University of Pennsylvania Law, 08–06 transparencia y la justicia.� FUNDAR, Centro de Análisis e (September). Investigación AC, Mexico. www.fundar.org.mx/fundar_1/site/ files/CasoProvida.pdf. Lindstedt, Catharina, and Daniel Naurin. 2010. “Transparency Against Corruption.� International Political Science Review 31 (4): Hazell, Robert. 1989. “Freedom of Information in Australia, Canada, 301–322. and New Zealand.� Public Administration 67: 189–210. Luna Pla. Issa. 2009. “Movimiento Social del Derecho de Acceso a Hazell, Robert, and Ben Worthy. 2010. “Assessing the Performance la Información.� Instituto de Investigaciones Jurídicas, UNAM. of Freedom of Information.� Government Information Quarterly México City. 27(4): 352–359. http://www.sciencedirect.com/science/article/pii/ S0740624X10000614#bbb016. Malesky, Edmund J., Paul J. Schuler, and Anh Tran. 2011. “The Adverse Effects of Sunshine: Evidence from a Field Experiment Hazell, Robert, Ben Worthy, and Mark Glover. 2010. The Impact of on Legislative Transparency in an Authoritarian Assembly,� the Freedom of Information Act on Central Government in the UK: October 4, 2011. Does FOI Work? Basingstoke and New York: Palgrave Macmillan. Mander, H., and A. S. Joshi. 1999. “The Movement for Right Holsen, Sarah. 2007. “Freedom of Information in the U.K., the U.S. to Information in India: People’s Power for the Control of and Canada.� The Information Management Journal (May/June). Corruption.� Paper presented at the Commonwealth Human http://www.arma.org/bookstore/files/Holsen.pdf. Rights Initiative Conference on “Pan-Commonwealth Advocacy Hubbard, Paul. 2007. “Putting the Power of Transparency in for Human Rights, Peace and Good Governance in Africa,� Context: Information’s Role in Reducing Corruption in Harare, January 21–24. http://www.humanrightsinitiative.org/ Uganda’s Education Sector.� Working Paper 136, Center for programs/ai/rti/india/articles/The%20Movement%20for%20 Global Development, Washington, DC. http://www.cgdev.org/ RTI%20in%20India.pdf. files/15050_file_Uganda.pdf. Martinez-Moralez, Emilene. 2009. “Documents in Action: FOI Human Rights Network-Uganda (HURINET). 2010. An Analysis Success Stories in Mexico,� FreedomInfo.org, March 20. http:// of Laws Inconsistent with the Right of Access to Information. www.freedominfo.org/2009/03/mexico-success-stories/. http://right2info.org/resources/publications/uganda-analysis-of- Mendel, Toby. 2003. Freedom of Information: A Comparative Legal laws. Survey. New Delhi: UNESCO, 96–8. IBN Live. 2006. “Government Backs Off, Won’t Harm RTI Act.� ———. 2004. “Legislation on Freedom of Information: Trends http://ibnlive.in.com/news/govt-to-drop-rti-amendment- and Standards.� PREM Note 93 (October), World Bank, act/19031-4.html. Washington, DC. Instituto Federal de Acceso a la Información Pública (IFAI). 2009. ———. 2009. The Right to Information in Latin America: A 7 Informe de Labores. http://www.derechoasaber.org.mx/data/ Comparative Legal Survey. Quito: UNESCO Office Quito. http:// arch_segui/i110810-1.pdf. www.unesco.org/new/en/communication-and-information/ Islam, Roumeen. 2003. “Do More Transparent Governments Govern resources/publications-and-communication-materials/ Better?� World Bank Policy Research Working Paper 3077, publications/full-list/the-right-to-information-in-latin-america-a- World Bank, Washington DC. comparative-legal-survey/. ———. 2006. “Does More Transparency Go Along with Better ———. 2011. “Amending Access to Information Legislation: Legal Governance?� Economics and Politics 18 (2): 121–167. and Political Issues.� Working Paper, World Bank, Washington, Iturbe, Diego Ernesto Díaz. 2007. “Métrica de la Transparencia.� DC. http://wbi.worldbank.org/wbi/Data/wbi/wbicms/files/ CIDE. http://www.metricadetransparencia.cide.edu/ drupal-acquia/wbi/Mendel%20-%20Amending%20ATI%20 MetricaTransp2007.pdf. Legislation,%20legal%20and%20political%20issues_0.pdf. 72 REFERENCES McCall, Elizabeth, and Alexandra Wilde. 2007. “Rights to Pandey, P., S. Goyal, and V. Sundararaman. 2009. “Community Information and the Poor: Experiences from India.� In Participation in Public Schools: Impact of Information Making Democracy Deliver. Innovative Governance for Human Campaigns in three Indian states.� Education Economics 17: 3. Development, 105–131. New York: UNDP. Ministry of Peisakhin, Leonid V., and Paul Pinto. 2010. “Is Transparency an Justice. 2011. Freedom of Information Act 2000—Statistics on Effective Anti-Corruption Strategy? Evidence from a Field Implementation in Central Government 2010 Annual and Q4: Experiment in India.� Regulation and Governance 4(3): 261–280. October–December 2010—Full Statistical Data. London: Ministry of Justice. Population Reference Bureau. 2011. World Population Data Sheet. Moraru, Adrian. 2011. “Free Access to Public Interest Information Probabilística. 2007. La cultura de los servidores públicos alrededor de 10 Years after the Adoption of the Law. An Assessment of los temas de la transparencia y acceso a la información: Encuesta. Jurisprudence on the Law Regarding Free Access to Information México: IFAI/ CETA. of Public Interest.� Institutul pentru Politici Publice (IPP), Puddephat, Andrew. 2009. “Exploring the Role of Civil Society in Bucharest, Romania. http://ipp.ro/protfiles.php?IDfile=125. the Formulation and Adoption of Access to Information Laws. Narayanaswamy, V. 2010. “Responsibility of Political Leadership The Cases of Bulgaria, India, Mexico, South Africa, and the in Promoting RTI.� The Fifth Convention on the Right to United Kingdom.� Access to Information Working Paper Series, Information Act, New Delhi, September 13. http://cic.gov.in/ World Bank, Washington, DC. convention-2010/Speeches/Narayanasamy.pdf. RaaG. 2009. “Safeguarding the Right to Information: Report of the Neicovcen, Vasile. 2000. Meeting minutes of a speech by member of People’s RTI Assessment 2008. Executive Summary.� Raag and parliament regarding that ATI law in the parliament’s session, National Campaign for People’s Right to Information (NCPRI). April 20. http://rti-assessment.org/exe_summ_report.pdf. Neuman, L. 2009. “Enforcement Models: Content and Context.� Reinikka, R., and J. Svensson. 2005. Fighting Corruption to Improve Access to Information Working Paper Series, World Bank, Schooling: Evidence from a Newspaper Campaign in Uganda.� Washington, DC. Papers and Proceedings of the Nineteenth Annual Congress of the European Economic Association. Journal of the European Neuman, Laura, and Richard Calland. 2007. “Making the Law Economic Association 3(2/3): 259–267. Work: The Challenges of Implementation.� In The Right to Know: Transparency for an Open World, edited by Ann Florini. Relya, Harold C. 2009. “Federal Freedom of Information Policy: NY: Columbia University Press. Highlights of Recent Developments,� Government Information Quarterly 26: 314–320. Nguyen, T., and G. Lassibille. 2008. “Improving Management in Education: Evidence from a Randomized Experiment in Republic of Uganda. 2005. “Democracy and Governance Madagascar.� http://www.poverty-action.org/node/4219. Assessment.� http://allafrica.com/download/resource/main/main/ idatcs/00011037:9bb6a95ddb73e6ba77d985dea14fdc73.pdf. O’Connor, N. 2010. “An Economic Argument for Stronger Freedom of Information Laws in Ireland.� TASC Discussion Paper. Roberts, Alasdair S. 2006a. Blacked Out: Government Secrecy in the Information Age. New York: Cambridge University Press. Olken, B.A. 2007. “Monitoring Corruption: Evidence from a Field Experiment in Indonesia.� Journal of Political Economy 115(2). ———. 2010. “A Great and Revolutionary Law? The First Four Years of India’s Experience with Right to Information Act.� Public Open Society Justice Initiative. 2006. Transparency and Silence. Administration Review 70 (6). http://papers.ssrn.com/sol3/papers. A Study of Access to Information in Fourteen Countries. Open cfm?abstract_id=1527858. Society Institute. http://www.soros.org/initiatives/justice/articles_ publications/publications/transparency_20060928. Roy, Aruna, and Nikhil Dey. 2004. “The Right to Information: Facilitating People’s Participation and State Accountability.� Open Society Justice Initiative and Center for Development and Workshop at the 10th International Anti-Corruption Democratization of the Institutions. 2007. “Proposals for Conference, Prague, Czech Republic, October 7–11. Changes in the Albanian Law on the Right to Information.� Presented at the Civil Society Forum Tirana (March 6). Roy, Aruna, and Nachiket Udupa. 2010. “Mass Job Guarantee,� Himal South Asian (October). www.himalmag.com/The-mass-job- Organization for Security and Cooperation in Europe (OSCE). guarantee_nw4749.html. 2006. “Constitutional Debate, Discussion at the Parliamentary Commission on the Drafting of the Constitution.� Volume 1: Singh, Shekhar. 2007. “India: Grassroots Initiatives.� In The Right to Introduction. Know: Transparency for an Open World, edited by Ann Florini. New York: Columbia University Press. Ostermann, Christian F. 2005. “CWIHP and Its Partners Seek Greater Access to Albanian Cold War Files.� Newsletter (April). Spence, M. 1973. “Job Market Signaling.� Quarterly Journal of http://www.shafr.org/passport/2005/april/ostermann.htm. Economics 87 (3): 355–374. Outlook India. 2006. “Protest against RTI Amendments Gaining Stigler, G. J. 1961. “The Economics of Information.� Journal of Momentum.� August 8. http://news.outlookindia.com/item. Political Economy 69 (3): 213–225. aspx?405250. Stiglitz, Joseph. 2001. “Information and the Change in the Paradigm Palmer, Geoffrey. 2007. “A Hard Look At The New Zealand in Economics.� Nobel Prize for Economics Lecture, Prize Lecture Experience with the Official Information Act After 25 (December 8), Columbia Business School, Columbia University, Years.� Address to International Conference of Information New York. Commissioners, Wellington, United Kingdom, November 27. 73 REFERENCES Support for Improvement in Governance and Management Vleugels, Roger. Overview of All FOIA Countries, Fringe (SIGMA). 2010. “Assessment Albania 2010.� A joint initiative Special (September 7, 2009), http://right2info.org/resources/ of OECD and European Union. http://www.oecd.org/countries/ publications-1. albania/46465284.pdf. Wantchekon, Leonard. 2009. “Can Informed Public Deliberation Sulaimani, Kamran. 2006. “Paid Just Rs 4 Per Day under Rural Job Overcome Clientelism? Experimental Evidence from Benin.� Scheme, Widow Moves Gujarat HC.� The Indian Express (June Working Paper, Department of Politics, Princeton University. 15). www.indianexpress.com/news/paid-just-rs-4-per-day-under- Weber, Max. 1922. Economy and Society. Berkeley: University of rural-job-scheme-widow-moves-gujarat-hc/6515/. California Press (1979 printing). Sundet, Geir. 2008. “Following the Money: Do Public Expenditure World Bank. 2004. World Development Report: Making Services Work Tracking Surveys Matter?� Bergen: Chr. Michelsen Institute U4:8. for Poor People. Washington DC: World Bank. Tavares, Samia Costa. 2007. “Do Freedom of Information Laws ———. 2007. “Strengthening World Bank Group Engagement on Decrease Corruption?� Rochester Institute of Technology, Governance and Anti-Corruption.� World Bank, Washington Rochester, NY. http://www.meida.org.il/wp-content/ D.C. uploads/2008/01/FOI%20and%20corruption.pdf. World Resources Institute. 2011. “Uganda’s Access to Information Tiwari, Ruhi, and Anuja. 2010. “RTI Penalties for delay in Regulations: Another Bump in the Road to Transparency.� information are rarely imposed.� Live Mint.com, January 13. Comments Paper (June 30). http://pdf.wri.org/uganda_access_to_ http://www.livemint.com/2011/01/13225141/RTI-penalties-for- information_regulations_2011-06-30.pdf. delay-in-inf.html. Worthy, Ben. 2010. “More Open but Not More Trusted? The Effect Szekely, Ivan. 2007. “Central and Eastern Europe: Starting from of the Freedom of Information Act 2000 on the United Kingdom Scratch.� In The Right to Know: Transparency for an Open Central Government.� Governance: An International Journal of World, ed. Ann Florini. New York: Columbia University Press. Policy, Administration, and Institutions 23(4): 561–582. United Kingdom, Prime Minister. 1969. Information and the Public Worthy, Ben, Jim Amos, Robert Hazell, and Gabrielle Bourke Interest, Cmnd 4089. Her Majesty’s Stationery Office. London. 2011. “Town Hall Transparency? The Impact of Freedom of United Nations. 1948. The Universal Declaration of Human Rights, Information on Local Government in England,� Constitution Article 19. http://www.un.org/en/documents/udhr/index.shtml. Unit, University College, London. http://www.ucl.ac.uk/ constitution-unit/research/foi/foi-and-local-government/town- Venkatesan, J. 2010. “Government Moots Amendments to the hall-transparency.pdf. RTI Act,� The Hindu, June 10. http://beta.thehindu.com/news/ national/article451129.ece. 74 The fundamental right of citizens to access information about their governments was first recognized in Article 19 of the United Nations Declaration on Human Rights and was also subsequently integrated into the constitutions of several countries. For much of the twentieth century, the right remained primarily theoretical and aspirational, except in a handful of countries—mostly western industrial democracies—that adopted legislation to operationalize the right to information. Over the last two decades, however, the number of countries with laws guaranteeing the right to information and enjoining the government to put in place necessary measures to enable the exercise of this right has grown exponentially—from less than 13 before 1990 to more than 90 by 2012. The adoption of an RTI law is an important statement of a country’s move toward more accountable governance. Access to information might be cast in intrinsic terms—as a democratic right of citizens to access information about the state’s functioning and the democratic responsibility of the state to make this information public. But the rapid spread of RTI legislation in developing countries has led to a focus on RTI in more instrumental terms to address some of the most difficult governance, welfare, and development challenges, by opening up decision-making and functions of the state to public scrutiny. However, while there is considerable literature on good practices and standards of legislation, there is little empirical research on how these provisions worked in practice, especially in developing-country contexts, if they have been effectively enforced, or on the effectiveness of RTI laws in fulfilling stated claims of improving transparency, accountability, good governance, or service delivery. The research on which this report is based was undertaken as a means of bridging this gap and on understanding the dynamics of implementation and effectiveness of RTI laws. It looked at a number of issues on the the effectiveness and impact of RTI laws and the factors that might explain how effectively the law works. What happens after the law is passed? Does it actually enable citizens to gain access to information, and what kinds of information? What evidence exists of the broader impact of RTI laws—for instance, in strengthening accountability of public officials, in anticorruption, and so on? In countries where there is evidence of the law working well, what were the strengths of the factors that might explain this? In countries where it is not being used, what constraints and challenges can be identified? The project looked at these issues in eight countries that have all passed RTI laws in the course of the last decade or so, span across different regions, and represent a range of income levels and varying political and administrative traditions. Although the number of countries under review is small, their experience offers useful lessons on potential challenges and constraints in setting in place the institutional changes, capacity enhancements, and normative shifts necessary for RTI laws to function effectively. The report draws lessons from the experience of these countries to inform policymakers and officials charged with implementing RTI in other contexts, who must be cognizant of and make concerted and explicit efforts to address these challenges if the Right to Information is to become a reality. The World Bank 1818 H Street, NW Washington, DC 20433 USA www.worldbank.org