Accountability and Anti-Corruption
The expectations that open data might serve as a strategic tool for reformers around the world to improve anti-corruption and accountability results has been a key driver behind the push by open data advocates for more and better open government data. The underlying theory appears straightforward: open data “can reinforce anti-corruption efforts by strengthening transparency, increasing trust in governments, and improving public sector integrity and accountability by reinforcing the rule of law through dynamic citizen participation, engagement, and multi-stakeholder collaboration”.1
Excitement over the promise of open data has been shared by large and small organisations alike. The G7 and the G20 have recognised its value, and multilaterals, such as the World Bank and the Inter-American Development Bank, have invested heavily in programmes to support open data. Bilateral aid agencies, including the Department for International Development (DFID) in the United Kingdom (UK) and the United States Agency for International Development (USAID), and philanthropic foundations, such as members of the Transparency and Accountability Initiative,2have also supported open data work. Additionally, multi-stakeholder initiatives like the Open Government Partnership, the Open Contracting Partnership (OCP), and the International Aid Transparency Initiative (IATI), among others, have facilitated and promoted efforts by government agencies, civil society, and media organisations across the world.
Current evidence about the impact of this work is relatively scant. Some argue that open data efforts have proven successful in “improving government by tackling corruption and increasing transparency, and enhancing public services and resource allocation”, and in “empowering citizens […] by enabling more informed decision making and new forms of social mobilisation”.3Yet, at the same time, others have pointed out that open data has not been widely used in corruption investigations.4Other research questions the linearity and simplicity of the assumption that data availability leads to results, arguing that “transparency, information or open data are not sufficient to generate accountability”.5It is fair to conclude that challenges exist in measuring the impact of open data to improve accountability and anti-corruption results. This raises questions about whether, and how, the open data community can convince the general public that greater access to open data is key to achieving results.
One reason why the evidence is patchy is that the relevant literature lacks common definitions of accountability and anti-corruption.6Definitions are often overly broad, defining accountability as the combination of answerability, the obligation to inform and justify public decisions, and enforceability, the ability to sanction or remedy contravening behaviour.7Corruption, in turn, is often used as an umbrella term to group behaviours related to the abuse of entrusted power, ranging from bribery and embezzlement to clientelism.8Both accountability and anti-corruption are about preventing, detecting, and disrupting abuses of power. Open data is a very powerful tool to reduce information asymmetries that lead to a power imbalance; however, more open information is not enough to actually negate the institutional and political dynamics that allow those in power to abuse it and remain impune.
Open data activists often assume that the solutions needed to strengthen accountability and to reduce corruption are already known by specialists, and that open data will increase the effectiveness of those working to implement such solutions. However, international development work focused on anti-corruption and accountability has been around only marginally longer than work on open data,910and the communities working on these issues have not yet reached consensus on several issues. Debates related to anti-corruption and accountability revolve around: concerns over how to prioritise and address corruption challenges in different contexts;11exploration of how to design, monitor, and implement interventions;12questions related to understanding and tracking changes in the political and technical dynamics that shape institutional reform and behavioural change;13discussions regarding how to identify and assess impact;14and ways to ensure that interventions actually empower marginalised groups and provide them with the means to improve their lives.15
Reflection on the overlap between the open data and the anti-corruption and accountability agendas offers important opportunities to methodically test underlying assumptions about the impact that power abuses have in practice and the role opening information can play in addressing these abuses. However, up to this point, such work has often been done by “pioneers” with little collaboration across agendas and with little attention given to the movement from simple data availability to using it strategically to address systemic or sectoral problems and achieving real impact.
This chapter will highlight the challenges, gaps, and progress made on key issues at the intersection between open data, accountability, and anti-corruption.
Open data for accountability and anti-corruption
In the mid-2000s, reformers pushing for open data began to demand the publication of data by governments in reusable formats that could be accessed by the general public. This effort later evolved toward identifying and then closing gaps in the publication of datasets,16with an additional focus on the implementation of data standards and data interoperability. Advocates have been successful in framing the open data agenda, advocating for standards, and convincing civil society, governments, and, to a lesser extent, the private sector to engage.
Open data initiatives have tended to focus on the release of data summarising existing government processes, while paying little attention to uses and users of the data, often treating open data as an end in itself. This has created momentum for the publication of datasets, but has also led to some governments focusing solely on transparency around selected issues without paying attention to opening up the underlying processes behind that data which are used internally to support transactions and decision-making. Open data and open government advocates have labeled these types of efforts as “passing off the release of inconsequential government-held data as transparency”17or “open-washing”.
The mostly implicit theory of change in many open data initiatives is that more information will (almost) automatically lead to its use by those working on anti-corruption and accountability and enable them to produce better outcomes and achieve impact. However, while information and technical improvements are great tools to better understand accountability and corruption challenges, they are not sufficient to address entrenched power structures that oppose governance reform and generate systemic changes.
In 2016, the Open Data Barometer found that a number of datasets relevant to anti-corruption work (e.g. budgets, company registries, spending, contracting, and land ownership) “still tend to be highly opaque, and often the least open”, and that important differences persist within and across regions.18A review of key datasets in five G20 countries also indicates that these relevant datasets are often not yet published, that public officials lack the skills to leverage open data, and that initiatives to strengthen citizen engagement using open data rarely link to anti-corruption or sectoral areas.19
In 2017, the “Open up guide: Using open data to combat corruption”20identified 30 key datasets21for fighting corruption (see sample in Figure 1), as well as standards that can make these datasets interoperable. The guide was tested in Mexico,22which produced evidence on the value of the guide for enabling government officials to open key datasets. It also highlighted the need to define clear data governance frameworks and to promote dialogue between data users and producers in government and civil society.
Top down and bottom up access to data
Efforts to open up data that is directly relevant to local accountability and corruption challenges are becoming more frequent, but they remain siloed, with a low degree of interoperability among released datasets that are often used by only a limited number of stakeholders active in a specific issue area. Such efforts are often led by civil society and, to a lesser extent, by governments. Examples of government-led efforts include the publication of commercial agreements, business relations, payments, and gifts to health providers by the private sector in France24and Germany,25as well as budget and/or spending data by many governments at different levels, often with support from international actors such as the Global Initiative for Fiscal Transparency,26the World Bank,27and Open Budgets.28These government-led efforts have also spread to government performance data, such as the publication of data on the use of public resources for natural risk management and response by Italy29and Mexico.30
Testing the “Open up guide: Using open data to combat corruption” – Mexico31
A joint effort by the government, Cívica Digital, Transparencia Mexicana, the Open Data Charter, and the Inter-American Development Bank tested the Open up Guide in Mexico by publishing a number of the key datasets it identifies.32This work provided insights into the challenges and opportunities of opening key datasets to fight corruption:
- Access to a list of key datasets and guidelines for data publication facilitates collaboration with institutions; however, this collaboration can be improved by prioritising data publication based on locally relevant corruption challenges and user needs. The process also provides entry points for opening datasets beyond the executive branch.
- Data publication needs to be complemented with capacity building for work on data and the provision of targeted support concerning gaps, legal challenges, and data use. The Mexico pilot enabled researchers to produce a process that can be used by governments elsewhere in their efforts to improve the publication of key datasets related to anti-corruption.
- Agencies with the mandate to open government data and civil society organisations are both key to ensuring the actual implementation of commitments to open data and to improving the processes and practices that underlie data production and use. This collaboration can be improved by instituting and/or strengthening formal data governance frameworks.
In other cases, civil society and media organisations have stepped in to close important gaps in the official publication of data related to accountability and anti-corruption. Most commonly, these efforts focus on those areas where governments have not indicated a willingness to act (or even explicitly oppose the publication of datasets) by using a wide array of strategies to achieve the release of information which is then transformed into open data. Such efforts often seek to pressure governments by accessing and releasing information in ways that will create incentives for government officials to publish the same information as open data. Some of the strategies used to access data when official open data is lacking include:
- Making public information requests33and publishing structured data from the results, such as the work by La Nacion newspaper on asset declarations.34
- Obtaining data from candidates running for public office and from government officials on assets, tax compliance, and interests, as with the work done by the civil society coalition behind the “tres de tres” initiative in Mexico.35
- Scraping documents and connecting different sources of data, such as with the publication of open data on political finance36in Peru by “Ojo público” and in Taiwan37by the Council Voting Guide.
- Transforming complex data into open formats as has been done by “Ciudadano Inteligente”38in Chile with regard to party financing.
- Turning information published by non-government actors (e.g. reports by private companies) into open data, as with the Data Extractors Programme by Publish What You Pay.39
- Combing through public records and linking up data to enable the investigations of potentially corrupt transactions, such as the work by the Open Data Institute (ODI) in Kenya40and the Organised Crime and Corruption Reporting Project in Eastern Europe.41
- Collating and systematising data from different sources and jurisdictions, such as the work by Open Ownership, merging public registers, government reports, and voluntary disclosure42to reveal beneficial ownership, or the work by Govtrack43with regard to the US Congress.
These efforts hold great potential, but have often faced challenges to translate data gathering into data use with tangible impact. Data often remains both siloed and dispersed, with information on the same topic being scattered across different agencies or levels of government, which provide the data in different ways and formats. Even where data can be collected and connected, concerns about its quality, completeness, usability, and sustainability are common. When working with data, questions of trust inevitably arise. Data users often doubt the reliability of the data and question whether the design and evaluation of public policies and decisions are actually based on that data. Finally, many potential data users face the emerging tendency of many governments to close civic space.44
Opening sensitive data in closed contexts
Most conversations around open data are based on experiences from those countries with some willingness to release open data on contentious issues, yet there are also efforts to open data for accountability and anti-corruption led by civil society mavericks in repressive countries with high levels of secrecy.
In Venezuela, the Transparency International chapter and the Instituto de Prensa y Sociedad de Venezuela have led an effort to compile, systematise, and publish open data45about regulations and decisions with regard to the use of public money. In Malaysia, the Sinar Project and the Web Foundation have produced and linked data about politically exposed persons46in an effort to shed light on how power is used and misused in the country. These admirable efforts challenge repressive and secretive governments and put issues of corruption and accountability up for public debate.
Over the last decade, progress and challenges in achieving accountability and anti-corruption results have led the community to gradually revise the theory and practice underlying their work on open data. Activists are now moving beyond models based on the supply and demand of data47to focus their work on more locally relevant problems, seeking to unpack the different elements needed to connect data production to use and impact. Some of the key ideas that may be coalescing into a revised theory of change include:
- The need to make explicit the steps needed to go from data production through to taking actions that can activate institutional responses.48
- A move from linear models to the use of cyclical and iterative approaches that enable a focus on specific governance challenges and the use of learning and adaptation.49
- Integrating open data into the operation of existing anti-corruption institutions and mechanisms.50
- Revising how to measure progress in the implementation of open data initiatives.51
The following sections will provide a deeper exploration of the different mechanisms connecting data availability and action with regard to existing anti-corruption systems and initiatives.
Moving from availability to use
Progress in the publication of data, even if uneven and patchy, has raised important questions about who will use that data, how they will use it, and what results can be achieved. There are no silver bullets when it comes to promoting the use of open data by local stakeholders to address corruption and accountability challenges. The approaches that have been used to bridge the gap between data production and use can be classified into three overlapping groups: those focused on data standardisation and technological tools, those focused on engaging users and particular problems, and those focused on changing government processes and practices.
First, those initiatives that have focused on standardisation and technological tools have paid great attention to the development of data standards and their implementation by governments. They aim to improve the quality and comparability of published data and enable the development of tools that can be adapted according to the needs of audiences in different contexts. These efforts have targeted a variety of areas from democratic processes to resource flows and, to a lesser extent, development results. Examples52include the IATI standard,53Fiscal Data Package,54the Popolo data specification,55the OpenCorporates schema,56and the Open Contracting Data Standard.57
The development and management of data standards related to accountability and anti-corruption have shown a similar trend to that of the broader open data space. Initially, standardisation was focused on finding ways to better present the information that was produced by governments, but later those leading the standards began to pay greater attention to data users’ needs, moving beyond representing government processes into using data to reshape those processes. Important challenges still remain in terms of the technical features and tools needed to make the implementation of data standards more useful and in relation to ensuring that stakeholders have the capacity to use the standards to address locally relevant challenges. Increased collaboration between standard developers, implementers, and data users at the global and national level is needed to develop technical solutions in a way that is sensitive to local capacities to produce the data and to put it to use within complex political systems.
Even though there are a number of stakeholders working to implement data standards, promote interoperability, and develop tools to facilitate data use, the actual use of open data has not increased proportionally. New projects that pay greater attention to supporting users trying to use data presented according to data standards are now emerging with strategies to promote data use58and to explore the use of open data to fight corruption in particular countries.
Second, those initiatives that have paid greater attention to engaging users and achieving particular outcomes have shown important results. A clear example is the work of journalists at the national and international levels involved in collaborative networks such as the International Consortium of Investigative journalists (ICIJ). Recent scandals, such as those exposed by the Panama59and Paradise60Papers, have not only uncovered corruption, but have led to the consequential launch of prosecutions and the resignation of public officers and even presidents.61After publishing such stories, data has been made available in open formats that can enable the work or analysis by others. While these examples could be used to question the value of open government data on politically salient issues when compared to data obtained through leaks, the disparity in outcomes may indicate more about the differences in the way this data is being produced, treated, and used.
Leaked data often includes full versions of documents that are then used to stimulate collaboration among networks of journalists, both online and offline. These networks review the data thoroughly to organise it, clean it, and make sense of it. The same networks then use the data to find leads that are further corroborated and developed through other sources, including open government data, documents, and on-the-ground research. This intense work is not focused on merely making the information available; it is aimed at making the information useful to further identify and expose illegal activities carried out by those in power.
Lastly, there are a number of initiatives that have focused on fostering and supporting changes in government processes and administrative practices. Some of this work relies heavily on data to explore the value of new technologies like machine learning, blockchain,62and algorithms;63however, using these tools to analyse open government data has not yet reached a widespread level of popularity.64Interest in these still emerging technologies leads inevitably to a range of challenges with regard to potential violation of privacy, the possibility of reproducing and increasing existing biases, and the threat of using automation to hide questionable decisions and practices.65
Other important work to promote change in government practices through the use of open data is led by multi-stakeholder initiatives on procurement, international aid, extractives, and public infrastructure. Even though these initiatives are at different levels in their uptake and maturity, all of them seek to alter long-established government processes. While some initiatives use formal multi-stakeholder forums for the production, verification, and use of open data, others promote the integration of open data into government processes beyond the simple publication of data. These initiatives have led to important, if not yet widespread, results,66ranging from identifying money flows in the extractives sector67and the misuse of public resources68to achieving savings and better service delivery through improvements in the planning and implementation of government processes (see the box below).
Open contracting: From open data to improved results
From saving millions in public resources69to fueling citizen mobilisation demanding accountability70and improving the implementation of service delivery programmes,71open contracting is one of the most successful uses of open data to improve anti-corruption and accountability results. Three features place the work of the OCP and its local partners72at the forefront of work on open data:
- Open contracting principles and the data standard to regularise the opening of procurement information were developed in collaboration with government reformers, lawyers, private sector companies, and the media.
- Sectoral efforts have gone beyond the development of a standard to focus on work with local reformers to address concrete challenges related to increasing value for money, strengthening public integrity, boosting market opportunities, enhancing internal efficiency, and improving the quality of goods and services.
- Reformers have used agile and adaptive ways for promoting the implementation of procurement reforms, user engagement, and the actual use of data, learning on the go and adjusting strategies as needed.
Opening information on government contracts is increasing the capacity of activists and journalists to understand and challenge existing structures and protocols that allow the siphoning of public resources and unfair contracting practices.
An example of this work in action is the joint effort by the municipal government of Bogotá and Colombia’s procurement agency, Colombia Compra Eficiente (CCE), to use open data to identify inefficiencies and corrupt practices in the delivery of school meals in the city.73The use of this data by government and suppliers has led to reshaping the way the programme is tendered, opening opportunities for more suppliers to participate, and enabling the busting of a price-fixing scheme for fruit. This improved the accountability of the process and enhanced the quality and timeliness of the meals provided.
The wide variety of approaches by government and civil society to address anti-corruption and accountability challenges should not be read as an attempt to identify the single best strategy to achieve results. Instead, the open data community needs to distill, share, and debate the lessons emanating from both successes and failures, reflecting on what these lessons mean for developing and implementing further projects moving forward. Additionally, it is not only a matter of choosing between approaches focused around a particular technology, stakeholder group, or government reform. It will be necessary to identify and explore, in practice, how a combination of these approaches can help to address particular corruption and accountability challenges within specific contexts.
Moving from use to impact
As discussed above, more data does not necessarily lead to a proportional increase in either the use of data or anti-corruption and accountability results; however, increased access to standardised, machine-readable, and reusable data has enabled sharper investigations into instances of corruption and abuses of power, additional research to identify inefficiencies in the use of public resources, and greater awareness of systematic biases against particular groups. Nonetheless, current advances are generally insufficient to address the root causes that underpin corruption and accountability challenges: the ways in which power is distributed in a given society and the subversion of existing (democratic) institutions for private gain.
There are several emerging efforts to improve openness beyond the executive branch7475and address corruption and accountability challenges76in other branches of government, including activities at the heart of the democratic process, such as monitoring elections and the undue influence of money in politics through campaign and party financing. Some initiatives, like those of organisations in the Openingparliament.org77network, have paid particular attention to the legislative branch.78These efforts to open and communicate information about legislators, how they perform their duties, and about legislation itself79have been at the centre of work to strengthen democracy, with consequent benefits for anti-corruption work. Yet, these emerging efforts often face important challenges in relation to the availability of data in machine-readable formats to support the accountability of members of parliament, as well as in relation to implementing lobbying reforms. While legislatures may be happy to see accessible data on aspects of government operations, they may be more resistant to opening up structured data on their own activities and interests.
A number of national governments have also been subject to interesting efforts to open up data about the judiciary80and oversight bodies, such as audit institutions. Their aim is to get a more complete picture of how cases are assigned to judges and how those cases progress until judgment is rendered. However, these efforts are not yet widespread and often face claims that they may hamper due process during trials.
Crucially, even when data is made available, initiatives tend to remain limited in their focus on particular branches of government or processes and generally have weak formal connections to the institutional systems in which they operate (e.g. the functioning of democratic institutions, the use of public resources, and the application of effective sanctions against those who engage in corrupt practices). This makes it challenging to follow cases of corruption from identification through to final resolution and sanctions, and, ultimately, hinders lasting impact and influence on future activities. Without ongoing scrutiny of democratic systems of power to enforce anti-corruption measures, individuals and institutions are able to continue to act with impunity, and the consolidation and replication of corrupt networks is facilitated.
The theory of change behind the idea of using open data for anti-corruption and accountability also highlights the potential value data can have in empowering citizens and enabling social mobilisation. Some organisations have used data to pursue an activist approach to crafting stories, uncovering wrongdoing, and identifying entry points that enable others to get involved. However, these approaches can put activists in peril, and, as of today, there are no established safety networks for this work, such as those that exist to protect human rights defenders or journalists. The absence of such safeguards, and the weaker links to established mechanisms for protection, may lead activists to take unnecessary risks and expose them to legal, reputational, or physical attacks.
Opening up the judiciary and advocating for greater accountability results
Due process and the effective management of evidence during trials is often seen by reformers as an excuse taken to the extreme by judicial bodies, preventing public disclosure of the most basic information on how cases are moved through the judicial system after a corruption scandal has been uncovered. However, some initiatives in this area have had an impact. One example of an open data initiative to obtain and use such information is the work done by the “Asociación Civil por la Igualdad y la Justicia” (ACIJ) in Argentina.
After years of litigation efforts to access information on corruption cases from the judiciary, and the burdensome work of turning hard copies into machine-readable data, ACIJ was able to create an observatory of cases.81This has enabled the public to demand greater accountability regarding the delivery of justice in corruption-related cases. Recently, this work has been further enabled by the opening of judicial information by the Argentine government.82Investigations into how corruption cases are allocated83have resulted in significant insights into how impunity is sustained, and there are now calls for reform to tackle more profound systemic issues in the judicial system.
Despite the emergence of various activist approaches, it is generally organisations that focus on governance, transparency, participation, and accountability that most frequently lead initiatives to address accountability and anti-corruption. These organisations play an important role, but still need to find effective ways to engage other key stakeholders, such as organisations working in particular sectors or territories, those working on rights protection, those active in social movements, or those working through other alternative mechanisms, such as strategic litigation. The minimal connections that often exist between open data initiatives and a broader range of stakeholders can deepen the challenges related to the usability, and actual use, of the data and hinder real impact in addressing problems that affect citizens.
The assumption that an improved capacity to identify instances of corruption leads to activating institutional oversight mechanisms is not necessarily wrong; however, assuming that those mechanisms will actually deliver results in the form of successful reforms, grievance redress, or sanctions without additional effort is, at the very least, an oversimplification. Open data can be a tool useful not only to identify instances of corruption, but also to engage, challenge, and reform the institutional designs and practices that enable corruption. Turning this potential into reality requires the use of approaches that consider the institutional and political environments in which data is produced and used. Adopting such approaches will enable sharper thinking about how the use of data can be more effective in practice, how to counter the forces that oppose openness (be they for private gain or from an aversion to change), and how to build stronger bridges between advocates for open data, activists working on sectoral and systemic challenges, and the democratic forces that can act on the findings and evidence obtained from the use of open data.
The open data community needs to explore and test innovative ways of using data that take all of these slowly acquired insights into consideration: to effectively challenge institutional mechanisms and practices that perpetuate impunity, inefficiencies, and the abuse of power; to reach out to unusual stakeholders by finding ways to integrate their needs and interests; to tap into existing social mobilisation processes; and to link the efforts of the different stakeholders engaging on reform with government branches and institutions.
Reflections and the road ahead
Over the past decade, reformers have used open data to create ripples and, in some cases, waves, in uncovering and prosecuting corruption. In a few cases, these efforts have resulted in the reform of systems where corruption had been the norm. Through their work, these reformers have generated insights that can help us to understand how to use open data more effectively to fight against corruption going forward. One of the key insights open data reformers have started to embrace is the value of adopting a problem-driven approach to the publication and use of data in order to address much more specific corruption and accountability challenges. These approaches also call for more collaborative models that are more grounded in the environmental context in which they are to be implemented, building on the needs and interests of local reformers and moving away from the replication of generalised practices toward the development of tailored “best-fit” solutions.
This shift in thinking on how to best use open data for accountability and anti-corruption does not represent a break with the ideals at the core of the open data movement, such as “open by default”, but it does call for some refinement of our thinking around how to articulate advocacy goals, learning aims, and the desired impact. There is, and will continue to be, value in demanding that governments open up data on key issues related to accountability and anti-corruption; however, these demands should be based on a clear understanding of the users and usefulness of data, as well as the technical, political, and institutional environments in which it will be used.
To grapple with the implications of these insights, stakeholders would benefit from engaging with each other to develop non-linear approaches to better address particular corruption and accountability challenges. Learning about other perspectives and approaches will provide useful insights to improve how we devise and test methods, monitor progress and results, and spur dialogue on how and why specific approaches might yield better results. In particular, it is important to explicitly address the following questions:
- How can the field facilitate and strengthen the work of local champions, including government, civil society, and the private sector, to generate and use data evidence to demand accountability and to lead in the fight against corruption?
- What are the needs of different local stakeholders with regard to using open data? How can these insights help to tailor technical tools and methodological approaches to better support stakeholders in different sectors and contexts?
- How can stakeholders build stronger and more effective connections among those working on open data, accountability, and anti-corruption, and those who work in sectors on specific issue areas?
- What are the potential risks associated with using emerging technologies, such as machine learning, and artificial intelligence, in relation to accountability and anti-corruption? How can these tools and methods be combined with the social mobilisation and institutional mechanisms needed to generate and sustain change?
- How can actors link the technical capacities needed to use open data with the political strategies needed to effectively change systems and ensure sustainable results?
Success in addressing these questions over the next decade will enable reformers to achieve significant accountability and anti-corruption results, but future work will require the community to develop holistic theories of change and the willingness to test them, implementing interventions in an iterative manner that enables reformers to ensure that open data is useful and used, while strengthening collaboration among stakeholders to achieve systemic reform and explicitly addressing entrenched power dynamics. In addition, the community must move beyond simple dichotomies that either highlight the production or the use of data toward models that start with identifying a specific problem to be solved, include the identification of the opportunities and challenges faced by local champions, and embrace learning and adaptation to develop solutions that are a better fit in specific environmental contexts.