Essays commissioned and edited by Thomas Carothers


Founded in 2010, the Transparency and Accountability Initiative (T/AI)—a collaborative undertaking of five private and public funders active in the governance field—seeks to support and generate innovative, practical work on transparency and accountability. T/AI believes in the potential to seize momentum generated by different strands of the thriving global transparency and accountability movement to have significant impact. Working with governments, foundations, NGOs, and other practitioners, we galvanize support for ambitious new ideas and promote better funding for this domain.

After T/AI completed five years of work last year, the T/AI steering committee began engaging in an effort to think through lessons learned from T/AI’s activities to date, identify promising frontiers for future work, and build new bridges between different sectoral and expertise “silos.” As part of that effort, in early 2016, we commissioned Thomas Carothers of the Carnegie Endowment for International Peace to collect and edit a series of short, forward-looking think pieces by leading scholars and activists on the future of the transparency and accountability movement. We commissioned the series for use at a T/AI steering committee retreat in March 2016, focused on defining a long-term vision of the collaborative going forward.

The prompt given to authors was, “Based on your understanding of how the field of transparency and accountability has evolved in recent years, what issues or areas of work would be especially important and fruitful for T/AI’s funders to address in the next several years?” 

In the spirit of T/AI’s commitment to knowledge sharing and openness, we decided after our March retreat to disseminate this collection of pieces more widely, in the hope that this set of essays, which has greatly helped our own thinking and practice evolve, will also be of use to a wider set of people and organizations working on transparency and accountability issues.

We wish to thank Thomas Carothers for curating, editing, and providing an analytic overview of the essays, as well as the authors for their ideas and inspiration in helping move the field forward.

Julie McCarthy, Open Society Foundations, and Martin Tisne, Omidyar Network, members of the steering committee of the Transparency and Accountability Initiative

Focus on Cross-Border Issues

Owen Barder, Vice President, Center for Global Development

“There’s serious evil in the world, and hardship and pain. And we should be humble and modest in our belief we can eliminate those things. But we shouldn’t use that as an excuse for cynicism and inaction. . . . We have to make these efforts knowing they are hard, and not swinging from naive idealism to bitter realism.” —U.S. President Barack Obama talking about aid to Africa with David Brooks in 2007, quoted in the New York Times.

As President Obama said, outsiders should be humble about our ability to improve governance or end corruption in other countries. Governance shapes and is shaped by a country’s political, social, and economic context: it is part of a complex adaptive system. We can’t heal the system, but we can perhaps help the system to heal itself.

Transparency is not an end in itself. It is only important if societies use the information to bring about social and economic change. Transparency does not automatically bring about accountability; nor does accountability automatically bring about progressive change. The attraction of T&A is that they do not attempt to bring about specific changes in governance; rather, they enhance the ability of a society to bring about changes itself.

Cross-Border Issues

It is tempting to try to jump-start or bypass the process by which societies use access to information to bring about change by putting more of our effort into stimulating civil society movements within developing countries. There are many reasons for this: we want to see faster progress, both because we want better outcomes and because we worry that, unless the movement gets some demonstrable results soon, transparency fatigue will set in and we will lose momentum.

In my view, this would be the wrong conclusion. We must continue to create the conditions in which internal forces can drive progressive change, including cajoling and encouraging other countries to embrace transparency. But if we try to create those movements from the outside, we risk isomorphic mimicry and the creation of astroturf movements. We should be wary of allowing our attention to shift from cross-border issues over which outsiders have most leverage and legitimacy, creating the circumstances in which local civil society organizations have the best prospects of emerging and creating progressive change.

This is important because the policies of industrialized countries on economic transparency and open government have four far-ranging implications for developing countries:

  • Information held in each country will be substantially more valuable when it can be linked to information held elsewhere. For example, linking trade data with unique identifiers for companies would reduce the scope for misreporting, which facilitates corruption, theft, and tax evasion. Other examples of the value of complementary information include company ownership data and exchange of information about tax.
  • Cross-border payments from one country to another, such as foreign assistance or payments for mineral rights, are more likely to be put to good use and less likely to cause a resource curse if those payments are transparent and open to scrutiny by the general public.
  • International open data standards significantly increase the usefulness of information while also reducing the transaction costs of publishing it for governments and firms. Because these standards make it easier to link data from different sources, their value rises exponentially in the number of organizations that publish to them.
  • The establishment and promotion of international norms about openness—such as open contracting by governments and transparency about the ownership of firms—can help these values spread more rapidly.

These four themes constitute a substantive agenda on which there is much more work to do.

Tackling this agenda not only benefits developing countries, it also directly and indirectly supports the policy objectives of rich countries, such as increasing tax revenues, helping to prevent money laundering and terrorist financing, reducing corruption, undermining opaque and autocratic regimes, and increasing the effectiveness of foreign aid.

Of these four themes, the interventions with the most potential relate to making cross-border payments transparent and to promoting global norms. Developing countries could not do this themselves even if they wanted to (which in the worst cases they do not). In many cases, industrialized countries are unwilling to do this alone, for fear of putting their own firms and citizens at a competitive disadvantage, so multicountry cooperation is needed.

T&A efforts should build valuable momentum by focusing on detailed reporting of payments to governments by companies and governments, including detailed aid transparency, and a global campaign for open contracts for all governments.

Neglected Actors, New Standards

Deborah Brautigam, Professor and Director, International Development Program, School of Advanced International Studies, Johns Hopkins University

In recent years, the governance and development subfield has seen an explosion of new data ranging from the microevidence of field experiments (J-PAL and Accountability Lab) to more robust government aid commitment and disbursement figures (International Aid Transparency Initiative) and extractive industry revenues (EITI and Revenue Watch). T/AI and networks like it have been useful catalysts of these developments. At the same time, two areas stand out as ripe for further work: (1) engaging key actors in China and India in creating opportunities for cooperation and potentially, progress; and (2) creating guidelines, common standards, or a system of peer review that would give users information on the relative quality and reliability of accountability data.

Engaging China and India

From looking at T/AI’s website, it appears that India and, especially, China are underrepresented in T/AI’s work, despite their enormous size and growing international importance. A search of T/AI’s website using “China OR Chinese” brought only 39 results, exactly the same as Kenya, while “Africa OR African” brought 169 results. Organizations like T/AI will get vastly more bang for the buck with greater focus on these two global behemoths—both on domestic transparency issues and on working with groups in those countries to foster greater transparency externally.

A key lesson for engaging with China is to be ready to learn from and understand Chinese constraints on, and opportunities for, change. This must include grasping how and why they do things the way they do, as well as what events, lessons, motivations lead to change, and not assuming we know better and/or that cooperation is about leveraging their money to do things our way.

There is much to be learned from parallel experiences of engaging China in the areas of corporate social responsibility and environmental protection. Above all, if done gradually and well such engagement can bear fruit. International environmental organizations (WWF, IIED, and Wild Aid) and UNDP have been working hard to involve Chinese civil society organizations and government agencies. These useful mutual learning exercises have had clear, promising results, including new guidelines on responsible overseas investment published by the government-linked China Chamber of Commerce of Metals, Minerals and Chemicals Importers and Exporters.

Although it appears that T/AI is already more engaged in India than in China, T/AI could build on its strong relationship with transparency and budget programs in South Africa and India as a bridge to Chinese actors whom the South Africans and Indians might know through various BRICS summits and events.

Data Quality/Standards

With the proliferation of organizations generating online data, we have come a long way since the World Development Indicators was the main source of development data. Yet some of the new efforts to generate data by methods, such as crowdsourcing (with data posted online cleaned by volunteers, in a kind of “wiki-database” methodology), have proven less promising than hoped. The urge to create new data, and to be transparent and open with this data (that is, publishing it very quickly), has sometimes run ahead of the responsibility for ensuring accuracy.

For example, the Land Matrix, a multidonor-sponsored effort, used media reports and crowdsourcing to create an online repository of an assortment of information about land acquisitions in developing countries. AidData is a donor- and philanthropy-funded effort to collect media reports and publish information about China’s overseas finances. Both hoped to foster greater accountability, but along the way they also created surprisingly flawed databases, which were then downloaded and used by other development organizations and scholars.

This suggests that although information might be made more available through these efforts, it can also easily be incomplete, inaccurate, inconsistent, or invalid. Perfection is an impossible standard in data generation, and significant concerns about quality have also been raised regarding some of the most oft-used development data, including national statistics in some developing countries. It may be time to go beyond data proliferation and support the establishment of a set of guidelines or standards, perhaps combined with a peer-review system, for “grading” the data used for accountability. An effort like this could start small, and on a voluntary basis, by a group of organizations involved in transparency (Global Witness? EITI? Transparency International?) with a focus on peer review of data used to assess transparency or levels of corruption.

Properly Analyze Systemic Corruption

Sarah Chayes, Senior Associate, Carnegie Endowment for International Peace

The interrelated problems of lack of government transparency and accountability and severe corruption are increasingly understood by government decisionmakers, civil society, and donor organizations as central drivers of many of the world’s most significant problems—from extreme poverty to violent conflict.

The very existence of T/AI testifies to that fact, as does the thrust of U.S. Secretary of State John Kerry’s January 22, 2016, speech at Davos. “The quality of governance,” he exhorted World Economic Forum attendees, “is no longer just a domestic concern. . . . All of you who are businesspeople who engage in . . . one country or another . . . you need to demand accountability.”

Yet while corruption is back in the spotlight (after a post–September 11 lull when other imperatives took priority), decisionmakers tend not to dwell on understanding the distinctive nature of the phenomenon as it is experienced today. They leap directly to the question of “what works,” searching for near-term tactical gains and fatally neglecting to acquire a more systemic understanding that would permit genuinely strategic approaches. As a result, many attempted remedies prove ill-adapted to local realities and achieve little.

Corruption today often represents the highly adaptive behavior of sophisticated structures that have deliberately bent or hollowed out key elements of state function to capture important revenue streams, ensure impunity for network members, and provide them with opportunities to secure and flaunt their gains. Corruption is not a feature of governing systems in many countries, but the system itself.

These networks frequently cross international boundaries and vertical echelons, and integrate public and private sector actors together with outright criminals, sometimes including terrorists. And yet their elaborate and purposeful nature has not been met by equivalently sophisticated analysis by those who would address the problem. In fact, officials of several Western governments and the EU have recoiled at my suggestion that they systematically study the personnel, structure, and mode of operations of kleptocratic networks abroad, objecting that to do so would be too “sensitive.” But the same officials have indicated that they would be grateful to have access to such an analysis—executed by a reputable outside institution known for rigor and impartiality—for planning purposes and to assist in bilateral conversations about corruption.

As it stands, however, international actors are operating almost blind in these environments. Donors may be canceling out the impact of their T&A programming with other “mainstream” efforts they are funding in the same country or with other policies pursued by non-aid parts of their governments. Beleaguered and underresourced civil society organizations may be scattering precious human and material resources, rather than concentrating them against high-value system vulnerabilities. Reformers may not be equipped to capitalize on a brief window of opportunity provided by a transition or by the actions of an independent judiciary. 

Some innovative, non-risk-averse aid providers would do a tremendous service for squeamish governments and civil society organizations alike if they were to fund systematic analysis of the kleptocratic networks in some portion of the more than 80 countries afflicted with systemic corruption. Such analysis would need to be country-specific, requiring intimate, on-the-ground familiarity with each country, including the identification and use of sensitive data and sources. The (updated yearly) analysis should include

  1. the structure of the networks, including their degree of vertical and horizontal integration (across public, private, and criminal sectors), and how they cross national boundaries;
  2. a network diagram depicting the identities and relationships among key network members;
  3. the mode of operations, including the elements of state function that have been deliberately bent to the purposes of the kleptocratic network(s), the instruments of force they favor, the agencies they have deliberately hollowed out by diverting their resources, and the key enablers or the network extensions overseas (such as banks, law firms, or contractors); and
  4. the most important revenue streams captured by these networks and the key destinations for the money.

Equipped with this type of information, civil society activists would be much better able to identify and exploit key system vulnerabilities; donors could better target their resources; businesses might have to take a broader view of “corruption risk” and “corporate social responsibility”; and governments faced with political trade-offs in addressing corruption would be better to equipped to accurately weigh the options and tailor responses, and have probing discussions with counterparts in affected countries. 

Learning, Linkages, and Frontiers

Paolo de Renzio, Senior Research Fellow, Open Budget Initiative

I am familiar with T/AI’s work, having used some of its products and publications and taken part in some of its activities. It has occupied an important space for discussion and sharing, and spearheaded some interesting initiatives. I would highlight four areas as possible future T/AI priorities.

Bolster Evidence and Learning

While rigorous research in the T&A field—for example, on the impact of T&A interventions or the conditions for T&A improvements—has increased in recent years, there is still a need for coordinated support for research that can advance serious learning throughout the sector, and bring it together in an intelligent way. Issues that come to mind include: (1) exploring the linkages between availability of information, use of information, and effective accountability/government responsiveness; and (2) specific challenges and opportunities in authoritarian environments. These issues are relevant in a number of relevant subfields, such as budgets, freedom of information, and service delivery. It would also be useful for research to explore different methods and approaches. Too much energy seems to be devoted to RCT-type studies, without adequately recognizing their limitations. In particular, there is a lot of scope for promoting rigorously designed comparative case study projects based on thorough process tracing methods. I hope to be proven wrong, but the large amount of money spent on the T4Dev, project seems to me like a lost opportunity, to some extent, given some of the possible alternative uses of the resources.

Improve Linkages Among T&A-Related International Initiatives

I often still get a sense that many of the existing T&A-related initiatives work in silos (OGP, EITI, International Aid Transparency Initiative, GIFT, and so on), despite being often funded by the same group of donors. T/AI could play an important role in bringing together these initiatives around specific common issues, and reach out to other multi-stakeholder initiatives that have a T&A component but not as their main focus. This would ideally result in some joint projects or activities, along the lines of the ongoing collaboration between GIFT and OGP on the Fiscal Openness Working Group, based on the recognition that often these initiatives share common objectives and involve similar stakeholders. An initial mapping of existing initiatives, their objectives, and possible areas of overlap and joint action could help in this respect.

Promote Collaboration Across the State/Society Divide

Too often T&A interventions are overly focused on either state actors (for example, working with government ministries in support of governance reforms) or nonstate ones (for example, civil society and the media), but inadequately take on the difficulty of promoting effective collaboration between the two sides, which ultimately is what makes T&A interventions effective. There clearly are interesting potential complementarities between the work done with governments and with CSOs that should be better recognized and exploited. In the area of budgets, an example would be working with governments to create opportunities for civil society participation in the budget process, or with parliaments and audit agencies to increase their collaboration with nonstate actors. Similarly, open data platforms would greatly benefit from ongoing dialogue between data providers and data users, to ensure relevance, user-friendliness, and user feedback. Program design across the T&A sector, to the extent possible, should take into account the need to promote better collaboration across the state/society divide.

Explore New Frontiers

T/AI is well placed to provide a space for thinking about new areas that would benefit from a focus on transparency and accountability, and where that need has not yet been fully recognized. It could help promote debate and support seed initiatives to create consensus and catalyze change. Climate finance is an example of an emerging area of global debate where efforts have led to some attention being given to its transparency and accountability needs and challenges. Other areas where more work could help are climate change more broadly defined, and the 2030 Agenda for Sustainable Development including SDG monitoring. Issues like the regulation of global finance or global inequality could also benefit from more debate around how the production and dissemination of information could shift the balance of power and help generate mechanisms for holding powerful actors to account.

More Research Needed on the Voice Link

Shanta Devarajan, Chief Economist, Middle East and North Africa, World Bank

The World Development Report 2004 identified the underlying cause of the poor delivery of basic services as a failure of accountability. It described two links where the “long route of accountability” breaks down. When policymakers or politicians are unable to hold service providers accountable, the “compact” has broken down. For instance, a minister of education may be unable to monitor, let alone sanction, teachers in rural schools. As a result, teachers in public primary schools in India, for instance, are absent 25 percent of the time. When citizens are unable to hold politicians accountable, a different link, “voice,” has broken. In some electoral democracies with large numbers of poor voters, politicians who fail to deliver on basic services continue to get elected and re-elected. The combination of the two linkage breakdowns—when citizens are unable to hold politicians accountable and politicians in turn cannot hold providers accountable—leads to bad basic services, especially for the poor.

Since the 2004 report, considerable research has probed the compact between policymakers/politicians and service providers. RCTs have been used to examine the effects of different incentives to reduce teacher or doctor absenteeism. While this research has generated interesting results (mostly confirming that high-powered incentives work), it begs the question of whether politicians actually want to hold these providers accountable. That is, whether the voice link, between citizens and politicians, is working. There has been less research on this link, and what exists yields conflicting or ambiguous results.

Some research examines interventions that affect both links. For instance, Bjorkman and Svensson showed that in Uganda, publicizing the relative quality of health service delivery in different clinics led to a reduction in infant mortality in the treatment group. It is difficult to discern, though, whether the effect was due to politicians knowing that citizens were better informed, to politicians holding doctors in clinics accountable, or to citizens directly monitoring service providers. A comparable study led by Banerjee on primary education in Uttar Pradesh showed very little difference between the treatment and control groups. Neither study unbundled the effects to identify whether the results were due to changes in the voice or the compact relationship.

Other research, aimed at the voice relationship, assesses the effect of information campaigns on voter behavior and, ultimately, outcomes. Again, the results vary. Keefer and Khemani used a radio campaign in Benin on the importance of sending children to school and having them sleep under insecticide-treated bed nets as a means of encouraging parents to put pressure on the public education system to get teachers to show up, and on the health system to get health workers to stop stealing bed nets and selling them on the open market. In the treatment group, children’s learning outcomes were higher and malaria deaths lower. But these results were obtained through the opposite mechanism from the intended one: on receiving the radio messages, parents sent their kids to private schools and bought malaria nets in the open market. Instead of enabling citizens to hold politicians accountable, the radio campaign made parents accountable for the education and health of their children.

More research is needed to understand and unbundle the voice link. The role of information in strengthening or weakening citizens’ ability to hold politicians accountable is one major issue, but not the only one. I would suggest that a multidonor collaboration concentrate on a few other topics in the framework of the voice link that are underresearched. One such topic is the role of norms in determining the behavior of politicians. It could be that politicians in India tolerate absentee teachers in public schools because all other Indian politicians do. Meanwhile, in a different country, persistent teacher absenteeism would be considered scandalous. How do norms evolve, and how can they be shifted?

Another topic is what determines the type of person who runs for election and wins. Why would a person who does not share the prevailing norm choose to run for election? And how would he or she win an election where other candidates represent the prevailing norm? These are key questions for understanding citizens’ ability to hold politicians accountable. But they are political questions. Thus international organizations may have difficulty taking them on. A multidonor coalition could provide the neutral ground on which this critical research can be conducted.

Building and Empowering a Global Practitioners Community

Nikhil Dey and Aruna Roy, Mazdoor Kisan Shakti Sangathan (MKSS), India


From its initial local struggles to the building of a national campaign—which ultimately fueled a broad-based people’s push for a new law—India’s Right to Information (RTI) campaign has increased the participation of India’s citizens in the country’s democratic processes. The law is widely used, helping the movement sustain itself in the face of trenchant and powerful opposition. The outcome goes much beyond the law—it has changed citizen-state relations by democratizing governance.

Other recent areas of progress include: (1) social audits in the National Rural Employment Guarantee Act—deepening and institutionalizing the concept; (2) spreading social audits to other areas; (3) building janata (“people’s”) information systems, and not being content with management information systems; and (4) raising the need for a comprehensive, bottom-up accountability law.


Perhaps the biggest challenge today is to establish the functional connection between transparency and accountability, and start making modes of accountability more tangible and real. Others include

  • defining the progression from the right to information to the right to demystified usable information;
  • internalizing the rights-based framework of information access;
  • expanding use of the RTI and T&A platforms by the poor and marginalized—establishing and activating the intrinsic connection between the right to know and the right to dignity;
  • institutionalizing T&A in public sector delivery (building “people’s” information systems, instituting mandatory disclosure of information rather than just open data, and institutionalizing social/public audits); and
  • mediating the contradiction between independence and accountability. How do you make independent commissions accountable?

Technology is a double-edged sword that can be used for empowerment on the one hand, or centralization of power on the other. In both cases it may seem that transparency and accountability are being advanced. An important question is, who does technology serve? This needs to be studied and understood, because far too often technology is presented as benevolent aid and even a game changer by technocrats, whereas most often the mandate is political.

Potential Areas for Future Multidonor Collaborations

Looking at these processes in terms of tangible achievements and time lines (for the purpose of evaluating progress) runs completely counter to the organic nature of these processes. India’s RTI would have never taken off if it had been donor driven, due to the burden of donor requirements relating to reporting, documenting, evaluation, and analysis. The RTI law campaign was free from all these expectations and time lines, enabling people’s power to define the campaign.

What sorts of strategic collaborations with multidonor networks would therefore be fruitful?

  • Supporting a means to ensure that even small T&A efforts in different parts of the world are understood and shared in a community of practitioners from around the world. This would give practitioners an opportunity to learn from each other, but use local wisdom to contextualize the relevant needs.
  • Disseminating knowledge of social audits, which have appeal and relevance across the world. Donor collaborations can facilitate the process of people visiting areas where such initiatives are taking shape to forge knowledge sharing. This could include supporting a series of workshops on the meaning and scope of social audits and people-centered mechanisms of grievance redress (the “Janta Information System”).
  • Building a global discussion platform where regional and international meetings of activists are held, wherein activists set the agenda of what donors should do, rather than vice versa, and in which the role of donors would be to observe the discourse.
  • Helping create a community of T&A practitioners to (1) serve as a solidarity network for supporting local efforts and sustain sharing of knowledge through exposure visits, workshops, and discussion forums; and (2) protect local T&A initiatives by galvanizing international pressure. A collaboration of donors and activists could also build an alternate people-centered vocabulary of words like accountability, transparency, participation, governance, and good governance that have been otherwise co-opted by a techno-management-oriented donor leadership.
  • Establishing a 1 percent norm for funding for transparency and accountability in all social sector programs, to make people’s participation and monitoring sustainable.
  • Endorsing a charter that lays down the minimum principles of T&A that should apply to donor funding itself.
  • Endorsing a charter that lays down principles of T&A that should be a minimum set of demands on government, extractive industries, other businesses, security establishments, and so on. The minimum principles of T&A adopted by the Ministry of Rural Development, Government of India, to be applicable to its public programs in rural areas is one such example that can be built upon.

Moving From New Knowledge to New Practice

Walter Flores, Community of Practitioners on Accountability and Social Action in Health (COPASAH), Guatemala

For about ten years now, my team and I have been working on strengthening the accountability of public services for rural indigenous populations in Guatemala. During this time, we certainly have seen an evolution in the theory and practice of accountability. We started with an aim to strengthen the democratic governance of public polices and services in a postconflict context with historical inequity and social exclusion. To us, it was clear we were aiming at a relatively long process of engagement. The emerging frameworks and tools on social accountability and legal empowerment were useful to understand how to implement, monitor, and evaluate our interventions, which included improving working relationships among rural indigenous citizens, authorities, and public service providers and improving access to public services.

In addition to the increased interest from donors in accountability, we also witnessed a mushrooming of “social accountability projects,” which were technocratic interventions attempting to solve highly complex citizen-state interactions through two- or three-year projects. Fortunately, it seems these types of projects are much less frequent now.

In terms of the theoretical advance, in the past couple years we have seen the research by scholars who take an explicit approach to the politics of accountability, including power relations (Fox, Tembo, Joshi, and Halloran). For us, practitioners working in postconflict contexts with high levels of impunity, this theoretical advance speaks better to our experience than the technocratic view that dominated the past decade.

Despite the theoretical advance described above, we have yet to move forward on our operational tools and implementing frameworks. For instance, although many donors and large international NGOs recognize the complexity of accountability and the politics involved with it, they still work through self-contained projects and logframes. Projects would not be a problem per se if they were cycles in a broader agenda rooted in a strategic framework. But usually they are not.

I would highlight three current challenges and opportunities.

Going From Knowledge to Practice

In my organization, we follow the principles of action research, which hold that learning occurs when one changes his or her own practice. I would say that we have accumulated very important knowledge in the accountability field but are only just starting to learn (demonstrating changed practices).

Building a Genuine Collaboration Between Researchers and Practitioners

Most of the research in the field has consisted of top-down initiatives by donors and academic researchers. There are very few examples in which a research agenda has been developed and implemented with the active collaboration of practitioners. In 2015, I participated in two international meetings in which this situation was broadly recognized and there was a true interest in developing effective collaboration. To make this intention a reality, donors can help by supporting a continuous engagement between researchers and practitioners (going beyond one single meeting) and even supporting the development of a strategic research agenda that would be developed by interested researchers and practitioners together.

Connecting the Accountability Field With the SDGs Movement

There is an interest in the SDGs community to implement citizen participation as a way to ensure national and subnational accountability. However, many calls from within that field are referring to “citizen voice” as the key driver. The past decade of research and knowledge from the accountability field has shown us that citizen voice is not sufficient to affect the behavior and decisionmaking of those in power. This means that the accountability field has much to contribute to the SDGs movement. I see a unique opportunity to build collaboration between both fields. T/AI could develop a specific stream of work to support those researchers and practitioners who are building in-country and international collaboration about SDGs accountability, including citizen-led accountability.


I feel strongly that there is a need to continue advancing the theoretical work, frameworks, and tools related to the politics of accountability. At the core of citizens’ demands for transparency and accountability lies the need to build and strengthen democratic institutions. These efforts are for the medium and long term. It would be good therefore to have explicit goals in this regard and the framing of an action plan.

The Need for Vertical Integration

Jonathan Fox, Professor, School of International Service, American University


The link between transparency and accountability seems opaque. A steady stream of evidence from the mushrooming domain of T&A initiatives at local, national, and global levels finds some progress with information disclosure, yet limited traction toward achieving accountability. That’s without even addressing the challenge of shrinking civil society space. Yes, “it’s too soon to tell.” But the challenges appear to lie deeper.

Falling Short

The scholarly literature on how countries actually reach sustained accountability underscores systemic change. Yet tactical, tool-led approaches have not triggered “big bangs.” Successful top-down initiatives are rare. Where “people power” movements for accountability do take off, the T&A field faces challenges maximizing synergy. So far, few T&A initiatives have managed to pinpoint the incentive structures or weaken the vested interests that perpetuate impunity.

Sunshine has often not managed to disinfect. Conventional approaches to social accountability and open government tend to be locally circumscribed (for example, community scorecards), to rely on national agencies to disclose (often general and less than actionable) official data, or to make unrealistic assumptions about citizen demand for the products of CSO infomediaries. Local interventions remain localized, rarely spreading horizontally or extending leverage vertically by influencing higher-level authorities. Meanwhile, civic tech platforms in national capitals need to bolster leverage by connecting more with offline civic action. Plus, policy monitoring and public interest advocacy initiatives often are not well-articulated with each other, falling short of their potential for synergy.


A conceptual reboot is in order, to inform a new generation of strategies that take entrenched institutional obstacles more fully into account by “doing accountability differently.” This involves focusing on pro-citizen power shifts as the goal. One such strategy worth more serious consideration involves the vertical integration of coordinated civil society policy monitoring and advocacy, to leverage the power shifts at local, subnational, national, and transnational levels necessary to produce sustainable institutional change. The vertical integration proposition responds to a missing link in the field: the challenge of how to bolster impact by taking scale into account.

The core rationale for trying to monitor each stage and level of public sector actions is to reveal more precisely where the main causes of accountability failures are located and also their interconnected nature. After all, the forces of impunity are usually already vertically integrated. Understanding as many links in the chain of public sector decisions as possible will help inform possible solutions and empower the coalitions needed to promote them—including bolstering the public checks and balances institutions of horizontal accountability (if they are merely weak rather than actually captured). By attempting to take scale into account, vertical integration puts coalition-building between social and civic actors with complementary strengths at the center of the strategy—for example, infomediaries plus membership-based civic organizations, plus independent media, plus insider allies (if available).

That said, vertical integration is easier said than done. Coordination among very different kinds of actors, across scale, while bridging monitoring with advocacy—faces numerous challenges. Scaling up is daunting for bottom-up initiatives—even though some degree of “partial vertical integration” can bolster leverage. Scaling down to coordinate with local actors takes national initiatives outside their comfort zone. Conventional coalitions may be limited to groups that already know each other. Broad-based membership organizations with civic muscle and more technical policy analysis infomediaries may face cultural gaps. CSO actors that do policy monitoring and those that prioritize advocacy bring very different repertoires to the table and may lack experience with collaboration. Insider allies may fear any public criticism of government. These multiple gaps underscore the potential need for shared interlocutors to build and sustain diverse coalitions. Yet if the underlying goal is a pro-citizen power shift, then strategies are needed that can produce a whole that is greater than the sum of the parts.

Efforts at vertical integration are going on already yet are often off the radar screen of international actors. Doing more to identify and capture the lessons of such experiences can inform future strategies.

In short, vertically integrated, multilevel coalitions between CSOs, broad-based social organizations, and public sector allies (where available) can combine bottom-up independent policy monitoring with the civic muscle needed to use evidence effectively for public interest advocacy.

(This text draws from the introduction to an Issue Paper prepared for U4.)

Four Frontier Issues for Further Investigation

Archon Fung, Ford Foundation Professor of Democracy and Citizenship, Kennedy School of Government, Harvard University

As a domain of scholarly inquiry and development practice, the field of transparency and accountability has grown and deepened enormously in recent years. I would highlight four frontier areas for further investigation.

Domains and Dynamics

Are there particular areas in which T&A efforts are likely to be more or less effective? Transparency strategies that provide information for individuals (for example, consumers) may face fewer challenges than informational efforts that rely upon social or collective action (such as most efforts to hold governments or corporations accountable).1 There may be other characteristics that make transparency more apt for some kinds of problems than others. Some kinds of information may be generally more salient or comprehensible to the relevant individuals. Information for some kinds of public problems may be more easily acted upon than in other policy domains or contexts. For example, public budgeting information may be more easily understood, actionable, and salient when it pertains to local governments rather than to national government. In the public health domain, it may be easier to act on information that comes from education campaigns—like encouraging pregnant women to take nutritional supplements and where to get them—than on information about the corruption or indolence of local health officials.

Role of Governments

Government occupies a fundamentally ambiguous space in this domain.2 On the one hand, parts of government—corrupt frontline workers such as teachers or clinic workers; nonperforming agencies and departments; and procurement officers, licensers, police, or politicians on the take—are the targets of transparency. Many T&A efforts suppose that providing information about what they are doing (or not doing) is a key to reform. But on the other hand, other units of government are usually key to collecting, and often providing, the very information that constitutes transparency, such as absenteeism records, public spending records, or data about contract recipients. Furthermore, affirmative governmental actions in forms such as freedom of information laws and public records systems are usually necessary for transparency efforts to reach scale. How should those who favor T&A address these contradictory roles of government?

I believe the T&A field would benefit from a sharper understanding of the political dynamics that make some officials and government organizations favor transparency policies and motivated to implement them rigorously. Sometimes, positive government action comes from successful advocacy. It often requires reform-minded officials at higher and lower levels. How does this “reform-mindedness” come about? In some cases, supporting reform is good politics—it garners a political constituency. Others are more intrinsically motivated—perhaps from the view that transparency is an essential component of modern good governance. How do alliances between officials and components of government who favor transparency, on one hand, and social activists, on the other, work?

Users of Information3

All T&A efforts suppose—explicitly or implicitly—that if information is made publicly available, someone will use that information to attempt some social, political, or economic improvement. I would urge that across the field, it become standard practice to immediately articulate, in every single effort, who that someone is. Whom do reformers believe will use transparency information, why will they use it, and what are the hoped for effects of their use? Laying out this everyday logic model would clarify thinking on the ground and provide the basis for post hoc assessments by funders, researchers, and reformers themselves. The next step in this investigation is to understand the capabilities that different kinds of users need to utilize information from transparency policies effectively. That is, what sorts of expertise, capability, and orientation do potential users like journalists, advocacy groups, political opposition parties, and groups of public service users need to understand and deploy information from transparency efforts?

The Role of Technology in Transparency and Accountability

I have begun to address that question elsewhere.4

Can Advances in Identification Technology and Systems Enhance Transparency and Accountability?

Alan Gelb, Senior Fellow and Director of Studies, Center for Global Development

Many countries are implementing new identification programs or strengthening existing ones. Some of these programs are “foundational” or multipurpose; others are “functional” or directed toward particular uses. Donors have supported at least half of them. The largest is India’s Aadhaar program that has so far enrolled some 960 million people, but very large programs have been rolled out in Indonesia, Pakistan, and other countries.

Especially in the many countries with weak civil registration systems, this portends a sea change in the relationship between citizens and states. The new technologies ensure unique identities in ways not possible before, including to those previously not formally recognized, who are invariably the most marginalized groups. The technologies provide methods for individuals to authenticate themselves remotely, establishing a clear one-to-one link between individuals and also between governments and people. Transfers and payments and the delivery of many other services can be fully tracked from source to individual recipient. This potentially increases T&A in service delivery. Such technology has been used in various settings, ranging from the South African social security system (SASSA) and disaster relief in Pakistan (Watan) to the reform of LPG subsidies in India (PAHAL-DBTL).

The Question

It is still unclear, however, how the new ID systems will influence development outcomes and governance more broadly. The horse of new ID technology has left the stable—which way will it run? One view sees it further increasing the “legibility” of a state and enhancing state power over individuals (Scott) and enabling surveillance (Privacy International). Another view is that legal identity and recognition are essential if individuals are to realize their rights (SDGs). Yet another is that the new systems should be seen in the context of efforts to improve public administration, enhance the delivery of services, and reduce the scope for corruption.

What Do We Know So Far?

Impact can work in any of these ways depending on the political economy of the country, and impact may be asymmetric, affecting the poor and rich differently. The potential of the new ID systems to monitor service delivery for timeliness, accuracy, and quality is barely being used by developing country governments. An approach along this line could have done much, for example, to increase T&A for the use of funds in Haiti (Ramachandran, Subramanian, Gelb, and others on CGD’s website). One of the innovations in the remarkable rollout of the Aadhaar program itself has been the use of advanced process and service monitoring tools (Gelb and Clark).

A process to monitor uses of the Aadhaar program is in the works. Very little rigorous performance data on other country cases exists, and few use the potential of their management information systems to provide accessible public data on their achievements. The next phase should emphasize the use of these technologies to enable a highly transparent process of holding service providers to account. Donors ought to support this through the design of their programs and by offering standards and examples to governments ready to support such innovation.

What Could Be Done?

CGD has been one of the very few places trying to understand, on a cross-country basis, the policy issues associated with new ID technology. Donors are now turning their attention to a more systematic approach toward the topic, including through a global initiative, ID4D. There is also increasing attention in the countries themselves: ID4Africa, a new initiative launched to help countries learn from each other brought together some 350 people from 27 countries in June 2015; the second meeting is planned for May 2016. These meetings provide useful information on many initiatives, but there is still little systematic work on the T&A implications of the new systems.

What is needed is a concerted initiative to take advantage of new technology to enhance the transparency of service delivery through a multidonor (partner) initiative. Cash transfer programs could be a place to start, since quite a number use these approaches. But there are many other possible entry points, for example, partnering between established groups engaged in development-related areas and groups such as GSMA, the Biometrics Institute, and ID4Africa that come out of the technology side as well as some private companies with very highly advanced process and service monitoring capacity, such as Federal Express, UPS, and so on.

Getting Closer to the Ground

Blair Glencorse, Executive Director, Accountability Lab

Progress in the field of transparency and accountability has been patchy. Funders have focused largely on transparency (and to a lesser extent accountability), and the associated efforts around open government. This has been valuable—transparent information should provide the basis for more effective policymaking and service delivery. Efforts around freedom of information and citizen monitoring of budgets, for example, have led to tangible, positive changes. At the multilateral level, initiatives like the Open Government Partnership have truly pushed the field forward and created frameworks for countries to monitor progress and share lessons.

At the same time, as T/AI has itself reported, the longer-term impact of this work in improving people’s lives has been mixed. Often, assumptions are untested, processes of change are vague, and contexts are insufficiently analyzed. Equally, efforts to show how transparency can then lead to greater accountability are often misunderstood or missing, and “political will” still seems difficult for us to understand and to generate. Technology frequently becomes the goal, rather than part of a larger strategy to engage people and governments in effective ways. Lessons from what works are often not shared and are rarely implemented. In short, the shift from open government to open governance is very much in its nascent stages.

Drawing on Accountability Lab’s experience in helping young people in developing countries develop T&A tools and link practice and learning to larger strategies and donor processes, I would recommend the following topics for greater T/AI future focus.

Using the Local Context

The T&A field still lacks a real, highly ethnographical approach to context (see Quaggiotto) and a rigorous, deeply nuanced understanding of the relationships and incentives that lead to behaviors within a specific political space. We need to work with organizations that build trust in meaningful ways (let’s think religious organizations or labor groups, not just NGOs), can map local assets in nontraditional manners, and use better and different types of data to make decisions (less surveys and statistics and more sense making of narratives, social media mining, and so on).

Understanding Varied Pathways for Change

The frequent emphasis on theories of change has led to linear mental models. The increasing emphasis on adaptive learning is useful, but we still tend to test singular assumptions. Our experience indicates that donors may want to move toward ways to test multiple theories of change at the same time (for example, more design thinking, rapid prototyping, and accountability incubators) as a way to discover next-generation ideas and push the field as a whole forward.

Building Accountability Movements

Too much T&A work is tactical and projectized rather than strategic and collaborative (see Fox). T&A efforts in a given context are frequently disparate and fail to cohere. The stakeholder ecosystem (see Halloran) is often not fully appreciated. Accountability must be understood as a generational movement that will require long-term, collective, and often nonlinear support. This means really working to bolster a shared shift in thinking and ensuring that change-makers understand that there is a community to support them and from which they can learn.

Move Toward Integrity

The inevitable result of the use of words and phrases like transparency, accountability, and even open government is that the starting point is a focus on monitoring, compliance, and enforcement. This makes the politics and the energy of the change process difficult from the outset. Reframing this work as building integrity and “naming and faming” rather than “naming and shaming” is far more constructive. It allows us to expand zones of effectiveness and support individuals who are inclined to reform and create a compelling change narrative. Donors should look for organizations that are thinking in these ways and mobilizing youth behind them.

Rethinking Multidonor Approaches

Efforts like T/AI and Making All Voice Count have been important in furthering understanding of the field, but often operate as echo chambers among traditional development actors. We are finding that the forward-thinking elements of the private sector, for example, are increasingly interested in how to build integrity (see their role in pushing for OGP in Pakistan). A useful step might be to bring in these potential donors in a more coherent manner, which would also allow processes like OGP to better operationalize.

Promoting a Culture of Transparency and Accountability

Frederick Golooba-Mutebi, Uganda- and Rwanda-Based Independent Researcher and Analyst

My reflections are influenced by observations and experience in the two African countries that I know best, Rwanda and Uganda.

Freedom of Information

Both these countries enacted freedom of information laws in recent years to much popular acclaim. The idea was that the measures would make it easy for citizens to access information and, should they encounter difficulties, allow them to invoke the law to compel whoever might wish to deny access to facilitate it instead.

In both countries, the law has emboldened some users of information, such as journalists, to make demands on public officials to provide them with what they need. In some instances, the information has been made available. In other cases, however, reality has turned out to be very different from theory. The process of trying to gain access is usually so protracted in that even where there is no will to obstruct, those seeking information give up. In theory they could go to court and seek redress. Such is the backlog of unresolved cases in local courts, however, that only the most determined and well-resourced take the legal route.

Increasing Popular Participation in Decisionmaking

Both countries have also made some moves to increase popular participation in decisionmaking, particularly at the local level. In both instances this was the result of the ideological orientation of the elites that rose to power after civil wars. They believed that popular participation would ensure decisionmaking took into account the wishes of members of the public, and that it was a safeguard against misrule. These efforts have fared better in Rwanda than in Uganda. The Rwandan government is committed to the idea, and attendance at public meetings is high (not least because of the obligatory monthly communal work), as is responsiveness by officials to popular concerns. In Uganda, however, it has long fallen victim to participation fatigue, partly because attendance at public meetings is voluntary, leading to very small numbers of people turning up, and partly because responsiveness by officials is low, itself a disincentive to participation.

Much energy and effort have been put into promoting participatory budgeting in both countries. In Rwanda, such participation has become commonplace because of the high attendance at public meetings and the public spiritedness of officials. In Uganda, the enthusiasm with which it was greeted in its early days has petered out. A combination of participation fatigue and the waning of enthusiasm by officials guaranteed a lack of sustainability.

Increasing Demand and Compliance

These observations point to two challenges. One is how to broadly increase citizen demand for transparency and accountability in societies with long histories of subservience and hierarchy. Another is what should be done in contexts where, despite the enactment of freedom of information laws, governments or officials remain reluctant to facilitate such access in practice. These questions take us to the issue of what donor collaborations might be useful.

I believe that international actors, often intent on near-term, narrowly defined gains, should not neglect the broader task of helping promote a culture of transparency and accountability. Although of course cultures grow from local roots, external efforts to help them evolve in certain ways can be meaningful. Efforts to promote transparency and accountability are being carried out in societies that have long seen an emphasis on the importance of public officials safeguarding state secrets. Reversing this culture of secrecy is not easy and requires concerted efforts.

One productive avenue is supporting public education campaigns through media, of which radio broadcasts are the most accessible, especially in rural communities. These should be geared at sensitizing public officials to the value and importance of transparency and accountability and specifically aimed at changing their inclination from wanting to hide information to encouraging people to demand it.

Another is building collaboration among key users of information, such as civil society groups and media, to compel officials to provide access to information, via litigation, if need be. Litigation costs time and resources that most users of information struggle to afford. Here is where donors ought to come in, to finance public interest litigation designed to encourage governments and officials to implement freedom of information laws in a timely fashion.

Investigate the Technological Gaps

Gus Hosein, Executive Director, Privacy International

Being less familiar with the transparency and accountability field than many of the other contributors, perhaps I can offer an alternative viewpoint, from a civil libertarian who works at the intersection of technology and society.

Reaching Out to Other Parts of Civil Society

The transparency sector needs to work on better engagement across civil society. This isn’t a criticism of existing work—we all work in our own domains. But transparency is necessary to study every part of every power activity. Even though across civil society people are working on holding the powerful to account, we don’t see many opportunities opening up for true collaboration between transparency experts and other civil society actors. 

That’s not to say that the transparency sector has to start partnering with other domains of civil society. But inspiring and engaging and also building capacity across civil society would be advantageous to everyone involved. It need not be about specific issues nor about co-opting other civil society institutions, but about the opportunities that transparency can provide in their own agendas.

Such collaborative activity is much needed in the Global South but not only there. The problems we are facing are everywhere. Europe (all regions) is a desert for resources of this nature and much more work is required there—I worry that the lack of resources in Europe is in part why Europe is reacting to the current threats from the self-proclaimed Islamic State and migration in such wild and diverse ways. More generally, funders should work together to finally fill the gaps between the funding that is destined for one region or another.

Exposing the “Social Within the Technological” Promise

While there are already great levels of interest within the transparency sector in registers and open data, I suggest a different approach to questions of transparency relating to data and technology.

In the privacy domain, we are always encountering promises that grand new technological systems and policies will fix society by reducing fraud and friction and by enabling efficiency and inclusion. Every aspect of these systems and policies needs the insight of the transparency field. Privacy advocates are already working hard to understand what data is being gathered and generated, as well as the legal powers of institutions and to identify the secret interpretations of laws that enable them to accumulate our data—though they struggle to understand how the devices, networks, and services collect and use that data. 

But the transparency sector could be helpful in looking at the other dynamics. Is there corruption within the procurement system for these new systems? What do proponents of the policy or system have to gain? What cases of misuse and abuse have arisen despite the claims of technologies solving graft? Whether it is an ID system that claims to reduce fraud and mismanagement in the disbursement of government aid, or a surveillance system that fails to function according to specifications, more information is required to match the billions spent against the real outcomes met.

This will help temper the political tendency toward a kind of intoxication with technological solutions. For instance, European politicians are speaking often these days about the need to strengthen borders and build passenger surveillance systems—yet last month the UK National Audit Office released a report on how the UK’s e-Borders scheme has been a wholesale failure. Currently, the job of doing these kinds of analyses lies with no sector of civil society.

Exploring Automated Power

In the future, the administration of power and decisions that impact our lives will be increasingly made by complex technological systems using algorithms and vast amounts of data. As societies, state institutions, and industry, and individuals too, come to rely on these systems, who is holding those systems to account? This is popularly referred to as “algorithmic transparency” at the moment, but there is more to it than that. Who is auditing the data input into machine learning systems? Who is carefully auditing how technological systems make decisions and who is impacted negatively and who gains disproportionately? Tomorrow, the transparency of our criminal justice, welfare, and national security systems will require these kinds of investigations.

The profession for this kind of work does not even yet exist. The legal regimes to undertake these kinds of investigations do not yet exist. The technological capabilities to break apart functioning systems are not fully formed. Yet nonetheless we are building and deploying these systems.

The challenge of transparency for tomorrow requires much preparation.

Multiple Ways to Go Deeper and Wider

Varja Lipovsek, Director, Learning, Monitoring, and Evaluation, Twaweza, Tanzania

T&A as a stand-alone field has had most resonance at the international level. It’s compelling to compare countries on indicators and useful to have initiatives like OGP, which unifies signatory countries under a set of guiding principles and provides some international peer pressure. The international trends can usefully be applied (so the argument goes) to influence national development trends. In practice, not surprisingly, there is a big difference between internationally agreed standards and principles and nationally binding ones.

Two Challenges

  • T&A rhetoric and nominal practice are often used for whitewashing, particularly the transparency component. It’s fairly easy to make information available but far more difficult to get to where the information is a common tool used in consultative deliberations between citizens and the state.
  • The T&A community often focuses on the national level, but governance is most meaningful at subnational levels, where it’s deeply specific and contextual, and touches directly on lived experiences. Even OGP is talking about “OGPx”—applying the principles to subnational levels. Translating international standards to local governance practices, via national frameworks and legislations, is a long process, requiring serious commitment.

What’s Important for the Future of the T&A Field

  • Link with and bring back the human rights agenda. We do T&A because we believe these principles should shape the social contract between the state and citizens. There is a powerful legal framework available for this (relevant nationally and internationally). Why have we backpedaled in the last ten years away from using it? Our new concerns about civic space—which is all about rights—indicate we are circling back to a rights focus. Why not make this move much more deliberate?
  • Get into institutional reform space. We believe that sectors that function according to T&A principles deliver better service (for which there is some evidence, though we could use a lot more). But in order to bring about sectoral reform we have to think about systems and institutions. It’s not only about changing the behavior of the people who staff the institutions (as is the fad these days), it’s about changing institutional processes and norms and how they are applied (down to the nitty-gritty of job descriptions, oversight roles, incentives, budget allocations, and so on). So work directly with governments, particularly in convincing governmental institutions to not be afraid to adopt an experimental approach—getting the social contract right includes not being always sure what the best mechanism is, and trying out different things.
  • Get deeply contextual. This means focusing on the subnational level, customizing initiatives more and worrying less about external validity, seriously understanding local political realities, history, and geography, and focusing on the long term and completely getting out of the short-term project cycle, except for serious piloting.
  • Look to widen spaces for citizen participation. Support spaces that are already bubbling; get into the thick of how civil society/citizen movements are working, avoid romanticizing “the grass roots,” and instead use anthropology feelers to understand power dynamics, such as marginalization of some groups and elite capture.
  • Finally admit that technology is a tool, not a transformative agent. Abandon the tool-driven approach to T&A of hammers (a scorecard, a mobile phone, and a platform) looking for nails. Set the norm that multiple tools should be tried and refined through iterative improvements.
  • Measure! Good evidence is paramount for any learning. Follow keenly the processes, intermediate steps, short-term results, but honestly measure long-term outcomes and impact (preferably ten years). Let the nature of the problem dictate the method, not the other way around.

What Sorts of Multidonor Collaborations Might Be Valuable?

Those that:

  • align T&A funding to a clear and joint learning agenda: generating rich, contextual evidence of what works (whether T&A is pursued as the means or the ends);
  • allocate funding (or help source funding) for supporting innovative measurements/evaluations, which bring together the best of the intellectual research thinking with in-country practitioners;
  • apply the T&A standards to their own practices, and demand the same from similar entities—multilaterals, bilaterals, and foundations; and
  • are unafraid to make links to the difficult questions on how we apply T&A principles to the global community (not just within countries)—particularly the role of the “global north” on issues of taxation, migration, and intellectual property.

Focus on “Rational Ignorance”

Paul O’Brien, Vice President for Policy and Campaigns, Oxfam America

As the poorest half of the planet sees that just 62 people have more wealth than all of them, collective frustration at extreme inequality is increasing. To rebalance power and wealth, many in our community are turning to transparency, accountability, participation, and inclusion. Interrogate that “development consensus,” however, and opinions are fractured over the benefits and costs of transferring power from the haves to the have-nots.

In truth, our theories of change often diverge. Most development organizations may agree on the need to advocate for more investment, innovation, information, and incentives as well as strong institutions, but some organizations are genuinely committed to only one of those i’s, and that can be problematic. Oxfam often finds itself choosing and moving between the relentless positivity of politically benign theories of change (for example, we just need more “investment” or “innovation”), the moderation of those who focus exclusively on transparent “information” with no clear pathway to ensure its political relevance, and the relentless negativity of activists that think the only way to transform “institutions” or realign the “incentives” of elites is to beat them up in public.

Oxfam’s challenge is both to be explicit in our theory of change and to show sophistication and dexterity in working across that spectrum. If Oxfam’s theory of change is based on a citizen-centered approach to tackling global systemic challenges like extreme inequality, then our opportunity may be engaging the “rational ignorance” of citizens and consumers.

Ignorance is rational when the costs of gaining knowledge outweigh the benefits. Rational ignorance is why I don’t read the Washington, DC, education budget even though my child goes to a public school here. It is why I never used to pick General Mills or Kellogg’s cereals based on their social justice performance. Until that answer was here and easy to find, my gut said it would take too much time, be irrelevant to my personal needs, and not give me a pathway to action, or when it did, wouldn’t be worth it. Even professionally, as more donors fund transparency work, there is just too much data, too and too many indices, to follow it all. In short, I would like to think my growing ignorance is mostly rational.

Rational ignorance is a profound threat to our theory of change. Oxfam recognizes that without systems for public accountability and active citizenry, states tend to forget their primary duty to regulate opportunity and power among their people. Instead power is captured. Oxfam’s relevance depends on our ability to overcome the rational ignorance of potentially active citizens. Our challenge is urgent—active citizenship cannot wait until states have capacity, elites get comfortable, or political rights open up. Like Acemoglu and Robinson, we think public institutions work best when political power is distributed at the same time as states build legitimate institutions.

That is why we celebrate when Ghanaian farmers march and present 20,000 signatures to successfully increase Ghana’s agriculture budget, or when consumers take more than 700,000 actions to get a slew of corporate reforms from the world’s biggest food manufacturers.

That doesn’t mean we are winning. As information channels grow, so does the rationality of ignorance, and our task is to make active citizenship more worthwhile. Peixoto’s and Fox’s findings are useful in this respect: (1) institutional responsiveness to citizen engagement tends to happen when online and offline support are blended; (2) donor-driven transparency citizen “voice” initiatives rarely yield institutional responsiveness, the ideas have to be owned by local institutions—government, CSOs, or both; and (3) exclusively demonizing the very elites from whom power must be distributed may not work.

Other lessons on overcoming rational ignorance include

  1. Translating data into relevance for citizens requires not just cutting-edge broadcast communication skills (see, for example, Tanzania) but interactive dialogue that allows citizens to shape debate, strategy, and outcomes (see Burkina Faso, among others); and
  2. Timing matters. See how citizens increased engagement in Zambia and the Dominican Republic before elections.

Our field is awash with slick terms that re-describe but fail to resolve old challenges. “Rational ignorance” may be one such term. But if it signals that consumers and citizens will ignore transparency and accountability efforts unless and until those efforts meaningfully engage the personal self-interest or civic energy of the people we ultimately serve, then it is worth chewing on.

(This essay has also been published on the World Bank’s blog People, Spaces, Deliberation, at

Accountability, Accounting, and Accounts: The Human Heart Is Not Transparent

Lant Pritchett, Professor of the Practice of International Development, Kennedy School of Government, Harvard University

I have been part of the “accountability” movement since before my role as one of the authors of the World Development Report 2004 that highlighted accountability and introduced the “accountability triangle” into the mainstream development discourse. I worry however that this movement has gone badly wrong as it has conflated “accountability” with “accounting” on the premise that the important and relevant features of the world are “transparent.” This has stripped the much-deeper notion of the “account” out of “accountability.”

An account is central both individually—the narrative I tell through which I as a person explain and justify myself to people whose views I care about—and organizationally—here is how we together construct meaning and legitimize our organizations existence and actions. Some part of an account is accounting—that is, reducible to “thin” descriptions of the world. But I would argue that most of the account needed in accountability is “thick” (in the sense of Geertz).

I realize this might seem odd, if not bizarre, coming from an economist, but the Beatles basically had it right: “I don’t care too much for money, money can’t buy me love.” Money cannot buy one love because the human heart is fundamentally not transparent, and hence one cannot make or enforce a contract for love. There is no reduction of love to a set of thin, hard, objectively verifiable indicators—whether these are inputs, outputs, or outcomes—such that accountability for love could be reduced to an accounting without an account.

I have lived in India recently, working on education there. The government education system is in dire straits as learning outcomes are poor and deteriorating. There are massive problems with teacher absenteeism and poor performance of teachers even if they do come to school.

So one approach is to pour money into the system for expansion and funding of inputs. This is the kind of “thin” approach that lends itself well to a pure “accounting” rhetoric of “transparency and accountability” as it is easy to document enrollment rates, input availability (for example, how many schools have books), and where budgets are going. The budget per pupil has expanded massively, the federal spending on education has increased tenfold, and new management information systems have been created that (in principle) keep track of all the thin inputs from the school level up. The problem is, none of this goes anywhere near the heart of the problem, which is the hearts of the teachers. In the state of Tamil Nadu one can use the MIS data to show inputs have gotten much better on every front. And yet, over the last ten years the government’s school enrollments have fallen by over 1 million students as parents have continued to vote with their feet out of the public into private schools.

Of course one could imagine increasing “transparency and accountability” aimed at inputs and teacher attendance. But recent experiments in the health sector are not at all promising that forcing attendance can improve services—the titles of two recent papers on schemes to increase attendance say something of the results—“Putting a Band-Aid on a Corpse” and “Deal With the Devil.” Once organizations have declined into dysfunction, a key problem is that formal mechanisms of accountability have ceased to have traction on the normative account of frontline providers’ behavior. Attacking that problem through “accounting” and “transparency” assumes one can beat a turtle into moving—that is, penetrate the hard defensive shell from external pressures that dysfunctional organizations have created.

If one needs to change the behavior of teachers so that they act with concern to achieve learning objectives and that a normative narrative guides their behavior, then one has to focus on the account. Why am I a teacher? What does a teacher do? How do I account for my behavior as a teacher to others whose esteem I value? It is not at all clear that more narrow and thin accounting-based accountability is on the path to changing the account of accountability that is needed for high performing education systems (and, by extension, a whole range of governmental activities). In fact, it may be not just a detour but a diversion.

Double Down on Technology

Ben Scott, Senior Advisor, Open Technology Institute, New America Foundation

The integration of digital technologies into public sector institutions has moved with astonishing speed, in no small part thanks to T/AI. The concepts of open government, open data, and digital services that were novel in 2010 are now accepted by many governments as essential to adequate public services. Many motives are driving these changes—including a desire to demonstrate higher levels of government accountability to the citizenry through digital transparency. Even for governments that pursue technological change for other economic reasons, digitization in public sector institutions creates a kind of path dependency toward transparency and accountability.

The use of technology—in a McLuhanesque sense—becomes a public message about T&A. Technology is not deterministic (it can be turned against people in disturbing ways), but its promise to deliver greater transparency, citizen engagement, and effective services breaks old paradigms and opens opportunities for progress.

Core Message

My message here is simple: double down on technology as a catalyst for change. Why?

First, digitalization has not yet achieved anywhere near its full potential. Open data/open government are in their infancy. To critics, practical change for real people is still anemic. Yet it is too early to judge. We should stick with it with renewed purpose. Even when corrupt governments appear to use new technologies as fig leaves to cover bad behavior—stick with it. This is a long game. When governments begin to digitalize record keeping, reporting, transactions, and engagement with citizens, good things can happen—people will want more.

Second, political evolution may accelerate as the technologies mature. In public sector technology implementation, we have not yet begun to see the returns of exponential progress common in the private sector—for example, search algorithms, interactive maps, autonomous vehicles, AI, and drones. But it is coming—the tools of “gov-tech” will become transformative (rather than just a “nice to have”). When that happens, we want T&A to be built into the DNA of the technology, which will require steady investment by T/AI and its partners.

Third, as noted above, technology offers a way for T/AI and its partners to work with governments that are not actually interested in either T or A. The benefits of T&A that are correlated with digital services will be externalities in this “economic logic” of digital transition.

Four Proposals for Future Investment

Help Grow Markets for Homegrown Gov-Tech

Though there are large tech companies with a significant public sector business, they do not totally dominate the sector like they do in other IT markets. Public sector tech has natural disadvantages for large corporations. It pays less than private sector. It is often atomized, so a sale to one local government institution doesn’t promise a pathway to a scale business. This market failure opens the door for homegrown business to work with its own government. T/AI can work from the top down to encourage government to be an anchor tenant in gov-tech markets, and it can work bottom-up to showcase homegrown talent (whether a social enterprise or conventional business).

Build Coalitions Around Gov-Tech

Promoting technology in governments should be done in partnership with private sector and civil society. The former should be there to have products to deliver; the latter to ensure the values of T/AI are persistently pushed in the processes of deliberation and implementation.

Encourage Experimentation

We should be looking for the harbingers of exponential change. For example, public blockchain technologies. The basic idea is to have a digital public ledger for recording time-stamped transactions that are verified through encrypted authentication and logged in a tamper-proof data file that is decentralized and incorruptible. What started as a radical strain of technology (for the exchange of cryptocurrencies) has created a data platform tailor-made for gov-tech. It offers 100 percent transparency in a record of transactions that is secure, decentralized, and highly cost-effective. Already, governments including Mexico, Colombia, and the UK are building pilot projects or exploring options. T/AI should be out in front of this trend.

Privacy and Security

As we argue consistently for the virtues of open gov/data, we should increase our awareness of privacy and security issues for citizens. As gov-tech moves into the “big data” world, there are new challenges for protecting data and the rights of people to their data.

Time for New Networks and Coalitions

Jeremy Weinstein, Professor of Political Science, Stanford University

We’ve learned that transparency and citizen participation aren’t silver bullets to produce accountability, but can be useful. The challenge now is to build on the evidence gathered so far, and put T&A issues into their appropriate place in the global development landscape. I recommend attention to three mutually reinforcing elements:

Double Down on Motivated Policymakers and Technocrats

T&A interventions work best when there are motivated policymakers and technocrats who see value in creating effective feedback loops, and are positioned to use the information or pressure generated from citizen engagement to change incentives inside a government bureaucracy. This continues to be an area worthy of investment on at least three fronts:

  • Regional and international efforts that create a common space, in which motivated policymakers and reformers can self-identify, connect with one another, and gain high-level political support for their efforts, continue to be worthwhile. In this sense, OGP’s openness to engaging subnational governments is exactly the right direction. OGP will likely be flooded with interest from governors and mayors, who will be ideally positioned to sustain the momentum of OGP.
  • It is essential to find ways to support efforts for governments and citizens to co-create new approaches to citizen engagement. As with the movement in cities in the United States, looking for ways to broaden collaborative problem-solving seems a worthwhile investment.
  • There is value in exploring how to use this network of motivated policymakers and technocrats to work on tough, crosscutting problems that multiple governments share, such as transparency related to crime and policing, taxation and revenue collection, or massive new inflows of support for climate finance. The efforts to push forward fiscal transparency, open data, and open contracting build on this insight.

Invest in Building Broader Coalitions Where Political Will Is Limited

If there is disappointment in the T&A community with progress, it likely comes in part from the recognition that the technical and the political cannot be separated. Improving the quality of government services means changing the incentives of those in power, and typically there are a set of actors who benefit from the status quo. Interventions are running aground where citizens, while informed, choose not to activate or mobilize, or where elites in political systems make adjustments to work around technical fixes to counter absenteeism or vote fraud. The donor push on T&A has led to a lot of box-checking by governments lacking motivated policymakers and reformers, so it is no surprise the results are often disappointing.

The challenge now is to get more concrete about exploring causal pathways in which other actors help turn information into empowerment and ultimately political action. At the individual level, this means understanding the conditions under which information mobilizes citizens, and how it can be delivered to facilitate action. At the community level, it requires focusing on intermediaries and coalitions. Citizens are rarely likely to mobilize on their own—except as it may relate to very local outcomes. An essential next step therefore is understanding how transparency efforts can be used to shape/leverage independent media and political party platforms and, perhaps most importantly, contribute to broad-based, issue-focused coalitions that energize key actors such as unions, churches, and so on. Resources that support practitioner/researcher partnerships to explore political mobilization would be well spent.

Use the New Consensus on the SDGs to Make Concrete Progress on a Few Promising Fronts

The adoption of Sustainable Development Goal 16 presents an enormous opportunity. For all the flaws of the new agenda, the recognition of the central role of governance to development was important, overdue, and hard fought. But this is the goal most likely to be orphaned in the years ahead, given the challenge and sensitivity of work on these issues. T/AI could play a critical convening role in figuring out how to make Goal 16 meaningful going forward.

For example, if we can use Goal 16 to make progress on commonly accepted, regularly generated measures of the quality of governance, it could be a game changer and become a focal point for national discussions, donor engagement, and more. Alternatively, T/AI could identify some subset of the Goal 16 agenda (for example, around information) and play a galvanizing role in making governments aware of interventions that have proven effective, helping national governments build partnerships to design new interventions and facilitating connections between researchers and governments to evaluate the impact of SDG-motivated commitments on governance.

Analytic Overview

Thomas Carothers, Vice President for Studies, Carnegie Endowment for International Peace

Second-Generation Guideposts

Although the contributors to this exercise range widely in their contributions across the domain of transparency and accountability work, a number of overarching ideas and themes appear in many of their essays. Taken together, these ideas and themes can be understood as basic guideposts of what might be called a “second generation” approach to T&A, one that the contributors believe should replace the core features of T&A work during its initial surge. It is not clear whether the assumed elements of the first generation approach were ever in fact quite as present as the analyses suggest—as with most sets of “second generation” recommendations in policy analysis, some tendency exists toward portraying the first generation as having been more simplistic than it probably was in practice. Yet setting up the two contrasting frameworks nevertheless provides a useful overview:

First Generation Second Generation
Inadequate attention to particularities of local contexts Deeper understanding of local contexts
Short-term projects Longer-term, more iterative, “organic” engagements
Blanket assumption re value of transparency More focus on how transparency translates into accountability
Fascination with technological tools; treating them as ends in themselves Viewing technological tools as means, not ends; experimenting with multiple tools
Act first, learn later Greater attention to accumulating and applying learning
Tactical aims Strategic approaches
Small-scale, fragmented efforts Building larger movements and coalitions

Beyond these general imperatives, the contributors’ recommendations span four principal categories: (1) recommended research; (2) new linkages; (3) improved methods; and (4) new substantive areas.

Recommended Research

The contributors put forward many ideas regarding further research.

A core idea mentioned by several authors is the need for more research on how precisely transparency in different issue areas and contexts translates into accountability. The authors emphasize that there may not be an overarching answer to this central question, but rather many different answers that activists and funders will have to grasp and incorporate. De Renzio frames this broadly, “There is still a need for coordinated support for research that can advance serious learning throughout the sector, and bring it together in an intelligent way.” Glencorse describes it as the need for understanding “pathways for change” and notes that the frequent emphasis on theories of change has led to “linear mental models.” Adaptive learning is useful, he comments, “but we still tend to test singular assumptions.”

Some of the writers emphasize the need for more research on particular kinds of contexts for transparency and accountability work. Chayes for example highlights what she believes is a serious lack of detailed understanding of kleptocratic systems in the many countries where they appear. She argues that such understanding is critical to effective work in such contexts, yet is frequently seen by external actors as being too sensitive to pursue. De Renzio stresses the need for greater understanding of how transparency and accountability play out in authoritarian contexts.

Other contributors highlight the need for more research on certain elements of the commonly assumed causal chain in accountability work. Devarajan for example notes how the “voice link” is crucial to effective accountability, yet is often not really understood in practice. The voice link involves not just the ability of citizens to use information in holding politicians accountable, but other issues, such as the role of norms in determining how politicians respond to pressure from citizens. He notes that research on the voice link inevitably raises highly political questions, and is thus avoided by multilateral organizations that try to stay clear of explicitly political work. Fung similarly highlights the need for work on how reform-mindedness arises among officials and other power holders: “The T&A field would benefit from a sharper understanding of the political dynamics that make some officials and government organizations favor transparency policies and motivated to implement them rigorously.”

Another major research suggestion comes from Flores who contends that most T&A research is the result of “top-down initiatives by donors and academic researchers.” There is a strong need in his view for donors to support the development of a strategic research agenda that would be jointly developed by practitioners and researchers.


Many of the authors point to the need for more linkages to be forged in T&A work. In some cases, the recommended linkages are within the domain of T&A actors. In other cases, they are between T&A actors, and actors outside the standard T&A field.

Recommended linkages to be strengthened or developed within the field:

  • Ties among different actors working on social audits in diverse parts of the world (Dey/Roy)
  • A global discussion platform for activists across different regions, “To ensure that even small T&A efforts in different parts of the world are understood and shared in a community of practitioners from around the world” (Dey/Roy)
  • Common space for motivated policymakers from different national contexts to connect with each other (Weinstein)
  • Bringing different T&A initiatives (such as OGP, EITI, International Aid Transparency Initiative, and GIFT) together around specific concerns, encouraging joint activities among them (de Renzio)

Recommended linkages to be strengthened or developed between the T&A field and others:

  • Ties between the T&A field and the SDG community, to help show SDG-focused actors how T&A can contribute across all of the SDGs (Flores)
  • Doing more to bring private sector actors into T&A efforts and processes (Glencorse, Scott)
  • Connecting T&A concerns to the civil society domain more generally—making it a feature of all civil society development work rather than allowing it to be seen as a specialized sector of civil society work (Hosein)

Improved Methods

Many contributors suggest possible improved methods in T&A work.

A starting point for many discussions of methods is the need for better context analysis by funders and practitioners. Glencorse for example argues that “the T&A field still lacks a real, highly ethnographical approach to context . . . and a rigorous, deeply nuanced understanding of the relationships and incentives that lead to behaviors within a specific political space.” Pritchett warns that in its focus on tangible outputs and markers, T&A work is neglecting deeper contextual issues, such as why service providers are not doing their job. For Lipovsek, getting more contextual “means focusing on the subnational level, customizing initiatives more and worrying less about external validity, seriously understanding local political realities, history, and geography.”

Following this, some writers stress the need for better theories of change. Fung notes that all T&A efforts explicitly or implicitly suppose that if more information is made publicly available, someone will use that information to attempt some social, political, or economic improvement. He urges that “across the field, it become standard practice to immediately articulate, in every single effort, who that someone is.”

O’Brien goes further into the need to do more to ensure that citizens act on the information that becomes available. Overcoming the “rational ignorance” of citizens is crucial, he notes, and he posits ways around it, such as better translation of data into relevant forms.

On how to get greater traction on accountability, several authors call for increased attention to the building of pro-accountability coalitions or movements. Weinstein for example says that an essential next step is concentrating on how transparency efforts can be used to “contribute to broad-based, issue-focused coalitions that energize key actors.” Arguing that T&A efforts “are frequently disparate and fail to cohere,” Glencorse says that “accountability must be understood as a generational movement that will require long-term, collective, and often nonlinear support.”

Fox takes the idea of coalitions or movements one step further, advocating that more attention be paid not just to greater horizontal linking up among actors but to vertical integration of T&A efforts: “Coordinated civil society policy monitoring and advocacy, to leverage the power shifts that local, subnational, national, and transnational levels necessary to produce sustainable institutional change.”

Two authors present other broad ideas for changed methods. Lipovsek calls for greater work at the subnational level, where traction may be more possible. Golooba-Mutebi argues that international actors are often intent on near-term, narrowly defined gains, and as a result neglect “the broader task of helping promote a culture of transparency and accountability.” Public education campaigns through media, as well as strategic litigation to compel officials to provide access to information, are two ways he suggests that donors can do this.

Suggestions regarding methods also concern the realm of monitoring and evaluation. Several authors urge funders not to let monitoring and evaluation methods (as opposed to learning from monitoring and evaluation) shape program design and implementation. More generally, Lipovsek urges funders to apply transparency and accountability norms and practices to themselves.

New Areas

Various contributors suggest new areas that T/AI could take up. Some of these are relatively specific issues:

  • Gelb outlines the important spread of new citizen identification programs in developing countries and notes that there has been little systematic work on the T&A implications of these systems. He recommends “a concerted initiative to take advantage of new technology to enhance the transparency of service delivery through a multidonor . . . initiative.”
  • Scott urges T/AI to develop a role helping grow markets for gov-tech: “T/AI can work from the top down to encourage government to be an anchor tenant in gov-tech markets, and it can work bottom-up to showcase homegrown talent.”
  • Hosein highlights how states and industry are increasingly relying on complex technological systems using algorithms and vast amounts of data, yet it is not clear who is holding these systems to account. He advocates a role for T/AI work to help create the analytic and legal foundations for investigating such systems.
  • Brautigam suggests that funders look for ways to help ensure quality of new streams of data becoming available through transparency work: “It may be time to go beyond data proliferation and support the establishment of a set of guidelines or standards, perhaps combined with a peer review system, for ‘grading’ the data used for accountability.”
  • Scott identifies public blockchain technologies as a promising new data platform for gov-tech that T&A funders should be paying attention to.
  • Dey and Roy note the need for figuring out better ways to mediate the tension between the accountability imperative and the independence of some state institutions, such as independent commissions.

Other ideas for new areas of work point to very broad areas of engagement:

  • De Renzio identifies climate finance as an emerging area of global debate where more attention to T&A would be warranted.
  • He also points to the regulation of global finance or global inequality as areas that “could also benefit from more debate around how the production and dissemination of information could shift the balance of power and help generate mechanisms for holding powerful actors to account.”
  • Brautigam advocates more attention to and engagement with China on transparency and accountability issues, arguing that there is much to be learned from the parallel experiences of engaging China on corporate social responsibility and environmental protection.
  • Barder advises that T&A funders concentrate on the policies of industrialized countries that affect T&A, such as information on company ownership and taxation, cross-border payments for mineral rights, and international open data standards.
  • Dey and Roy recommend greater funder focus on expanding the use of T&A platforms by the poor and marginalized.


1 See Archon Fung, “Infotopia: Unleashing the Power of Democratic Transparency,” Politics and Society 41, no. 2 (June 2013): 183–212.

2 For related reflections, see Stephen Kosack and Archon Fung, "Does Transparency Improve Governance?," Annual Review of Political Science 17 (2014): 65–87.

3 An extended discussion of users in the context of transparency in advanced industrial countries can be found in: Archon Fung, Mary Graham, and David Weil, Full Disclosure: The Perils and Promise of Transparency (Cambridge: Cambridge University Press, 2007). The challenges facing individual and organizational users are obviously different in developing country contexts.

4 See Archon Fung, Hollie Russon Gilman, and Jennifer Shkabatur, “Six Models for the Internet + Politics,” International Studies Review 15, no. 1 (March 2013): 30–47.