Expert Commentary

Does transparency improve governance? Evidence from 16 experimental evaluations

2014 study in the Annual Review of Political Science establishing clear definitions around transparency initiatives and analyzing 16 leading studies on field experiments.

The idea that transparency can make institutions more effective and provide greater accountability and better results for the public seems uncontroversial on the surface. But scholars and bureaucrats who have been involved in the wave of transparency initiatives over the past decade continue to debate the particular merits of various approaches. Some commentators have been troubled that as a reaction to scrutiny, malfeasance and inefficiency could increasingly be kept hidden and transparency could erode public trust in institutions and personal privacy.

The many types of transparency initiatives around the globe are often confused, making sharp distinctions all the more essential. A 2014 study published in the Annual Review of Political Science, “Does Transparency Improve Governance?” elucidates fundamental concepts and, after reviewing 66 studies on transparency and development outcomes, analyzes the 16 studies that were based on empirical data from experiments. The authors, Stephen Kosack of the University of Washington and Archon Fung of the Harvard Kennedy School, note that the research literature “has thus far yielded frustratingly mixed conclusions: some studies conclude that transparency has huge effects; others show little or no effect.” Among the 16 experiments examined, 11 were deemed relatively successful.

Kosack and Fung begin by distinguishing four fundamental types of transparency: (1) The right to government information, embodied by laws such as the U.S. Freedom of Information Act (FOIA); (2) transparency around private organizations and corporate behavior, often spurred by consumer campaigns; (3) government regulatory transparency such as “financial disclosures of corporations, product safety disclosures”; and, (4) “transparency for accountability,” which can help improve delivery of services in fields such as education and health. The researchers focus their analysis on “transparency for accountability” because it reflects “the evolution of transparency from an end in itself, or an ingredient with important but nonspecific benefits for democratic governance, into a tool for dealing with increasingly practical and specific concerns of government performance.” This area has increasingly become an area of focus for policymakers, foundations and other interested parties.

The study sets out three basic frameworks for understanding programs:

  • The action cycle: A successful transparency for accountability (T/A) initiative must trigger each element in a specified four-step process: (1) Users are provided with important information; (2) they act on that information by, for example, changing their health care provider or engaging in some modified behavior; (3) providers begin to see these actions as consequential; (4) providers respond constructively.
  • The pathways: Another essential framework, the researchers note, involves the different methods by which results are achieved: “The 2004 World Development Report delineated two general pathways by which users of a public service — citizens, or ‘clients’ in this conception — may press for improvement. In the ‘short route,’ citizens engage with those who provide public services directly. In the ‘long route,’ citizens use their political power — such as voting or advocacy — to press policy makers and politicians to improve the quality of service delivery.”
  • The societal context: The potential success of any T/A initiative greatly depends on surrounding dynamics. There are five basic contextual situations: (1) A situation of competitive services, where users have choice; (2) a situation with little choice, but where service providers are responsive; (3) a situation where there is both little choice and providers are unresponsive because they feel threatened by change, and where users need information to exert direct pressure; (4) a situation with little choice and no provider responsiveness, but with public officials who are willing to act on behalf of the people; and (5) a situation where the general lack of competition, responsiveness and political action potential means that only broad-based social action will be effective.

Based on these frameworks, Kosack and Fung evaluate the 16 studies based on experiments:

  • Experiments in situations where there were competitive services (citizens had multiple options) were successful. These included school performance initiatives in places as disparate as Punjab, Pakistan and North Carolina.
  • Ten of the 16 experiments were conducted in situations where there was little competition and limited provider responsiveness, requiring direct pressure by consumers/citizens. Six of the transparency initiatives conducted in this context were successful. The successes included schools-related initiatives in India and Indonesia. However, four experiments were unsuccessful, including one in India that was found to have “no effect on teacher effort or on learning.” One possible reason for that failure was that the “information provided to communities failed to increase their involvement in education.”
  • Three experiments took place in situations where public officials were willing to help, and these all succeeded. These included an anti-corruption campaign involving newspapers in Uganda and the monitoring of infrastructure spending in Indonesia.
  • The one transparency experiment evaluated that took place in a context where there was little competition, as well as limited responsiveness on behalf of providers and officials, failed. This experiment involved trying to improve nursing practices in India.

Kosack and Fung conclude that the data “suggest that interventions are more likely to be successful when, as implied by the ‘action cycle,’ they provide information that is clearly understandable and salient to citizens (e.g., by showing problems with inputs clearly related to the performance of their providers, and how this performance stacks up against their neighbors’ or against their rights to service); that illuminates problems with the inputs into services, not simply with the performance of the service, which may have myriad causes; and that implies or directly recommends a clear course of action to improve those problems.”

Overall, the evidence furnishes “reason for enthusiasm about the power of information to catalyze meaningful governance reforms,” the scholars write.

Related: For more on general intellectual trends in the area of transparency, see “Open Government and Conflicts with Public Trust and Privacy: Recent Research Ideas.” Harvard’s Edmond J. Safra Center for Ethics is also producing a wide range of research studies on issues of transparency.

Keywords: transparency, good governance

About The Author