Is it about Appearance or Action?
27 May 2013
Matt Andrews, Harvard Kennedy School
A growing governance agenda and post-2015 ambitions
The governance agenda has grown rapidly in the international development community. Words like ‘governance’ are commonplace and widely referenced indicators yield the agenda particularly visible. These include the Worldwide Governance Indicators (WGI), the Country Policy and Institutional Assessment (CPIA), and the Doing Business measures. Such constructs offer implicit definitions of what ‘good governance’ is that inform the substance of governance reforms. These reforms are often sponsored by international organizations and target changes in government institutions, or rules of the game. They account for billions of dollars of annual support to developing countries. With this increasing prominence it is no surprise that there is now talk of adding a governance indicator to the post-2015 development goals. This talk needs to be informed by concerns over the limited achievements of this agenda to date and the incentives it seems to foster for creating governance by appearance and not action.
Governance by appearance
I came upon the idea of ‘governance by appearance’ when researching how many countries recorded better governance indicator scores after years of institutional reform. I thought that most countries would have positive records given that indicators were created by the same agencies supporting reforms; fostering a ‘teach to the test’ scenario. I was surprised. Between 50 and 70 per cent of the countries I looked at (using different indicators) went backwards on scores between 1998 and 2008, even after doing reforms. Cape Verde dropped from 0.3 points on WGI scores even after US$122 million worth of World Bank sponsored reforms. Senegal’s scores dropped by a quarter of a point after 75 World Bank projects with public sector reform content.
There are various problems with this kind of analysis (especially using indicators for inter-temporal research) but it should be noted that the storyline is reinforced by other research. The African and Asian development banks find that many countries score worse on indicators after years of reform, and the World Bank routinely finds that CPIA scores decline or stay pat at low levels even after countries adopt reforms. Such work raises questions: Are indicators capturing the wrong things? Are reforms ineffective? When I construct historical narratives of reform journeys I find a bit of both, and what I call ‘reforms as signals’. This is reflected in an up-and-down pattern of reforms and indicator scores that seems rampant in development and should worry advocates of the governance agenda
The pattern typically starts with governments announcing best practice reforms promoted by the good governance agenda. Benin’s 1990s reforms are a good example. The country emerged from decades on Marxist-Leninist control and needed help from organizations like the IMF. It announced various privatization initiatives and civil service reforms to comply with emerging ideas of ‘good governance’.
These announcements generate positive perceptions of the government and inspire higher scores on governance indicators. Benin’s announcements led to generous support from the IMF and World Bank, and the WGI government effectiveness scores in 1996 were around zero, 7th best in Africa and arguably higher than one might have expected.
Four to six years later, however, observers notice that reforms are not being effectively implemented. Benin’s civil service reforms were questioned, for instance, when studies showed that civil servants who appeared to be ‘fired’ in the mid-1990s had actually been moved onto contract and returned to the payroll in later years. A civil service law that had been successfully presented to the parliament in the mid-1990s was never passed. What observers called ‘partial privatization’ led to government-influenced oligopolies.
Negative perceptions of reform implementation lead to declining scores on indicators. Weak reform implementation is noted as a major reason for dropping governance scores in Benin, which fell to -0.48 in 2001 (15th place in Africa).
Countries are then faced with potential decreases in external support because of weak governance indicators, and start announcing reforms again. They often promise the same reforms they failed to implement the first time around. Benin’s government initiated new waves of civil service, public financial management and even privatization reforms between 2002 and 2004.
The new reforms again generate positive perceptions and short-term increases in governance scores, but these are again short-lived as reform implementation fails. Largely because of the reform commitments, Benin’s WGI government effectiveness scores recovered to -0.35 in 2003, which placed it in Africa’s top ten. Reform efforts failed again in following years, however, because of prevailing political tensions and capacity constraints. Reform failures led to declining WGI scores after 2005, with Benin sitting 17th in Africa in 2008, scoring -0.51 for government effectiveness.
This rendition of Benin’s story is just one of many possible interpretations. It does, however, reflect a key concern about the current good governance agenda that affects many other countries as well (though not all, with middle-income countries having less of an up-and-down pattern). The concern is simply that the agenda creates incentives for governments to use governance reforms as signals, where governments are given ‘good governance’ points based on the forms and structures they commit to adopt, regardless of whether these fit the context or address pressing problems or can ultimately be implemented. When the forms are shown to be a poor fit—or unimplementable—governments are rewarded for rehashing their efforts, committing to adopt the same forms that give indicator scores another short-term boost, but are (again) unlikely to yield better governments in the medium- to long-term. This results in a failure to address the real governance problems and fosters ‘governance by appearance’.
Good governance in action
The up and down, ‘governance by appearance’ effect is a major reason why many countries do not show sustained governance score improvements over time. Given that the governance agenda has this effect in many countries (especially less developed countries), I am particularly concerned about reinforcing the current indicator and reform approach through a post-2015 indicator. I am also sceptical as to whether it is possible to construct a post-2015 governance agenda that fosters good governance that goes beyond appearance, and emphasizes action, but think this is where the discussion needs to focus.
This kind of agenda would not support countries in (or reward countries for) adopting the ‘forms’ of good governance that have dominated the agenda to date. Instead, the agenda—and reforms and indicators in such—would push governments to improve their functionality, regardless of the forms they adopt to do so. Such agenda requires having indicators that would only improve when governments become more functional and capable (of delivering the kinds of services or playing the kinds of roles needed). These could be output or outcome indicators that reflect whether governments use authority to influence society positively. For instance, using infant mortality statistics to show how well a government plays its role in the child health field. These could also be process indicators that map closely with improved functionality. Citizen registration statistics could show the degree to which governments adopt processes needed to foster citizen participation in society, for instance. The indicators could also focus on processes that seem absolutely vital—perhaps core public financial management or transparency measures—but these should always measure process implementation, not just adoption. I like the Global Integrity statistics in this respect. They allow one to see scores countries get for the quality of anti-corruption laws and for the implementation of laws. The gap between these two scores gives one a view on the ‘governance by appearance’ problem.
Indicators and reforms should also allow a higher degree of flexibility than they have in the past. Governance issues differ between countries as do potential opportunities for reform. The contextual nature of this area suggests that any overly-prescriptive global agenda will likely have major limits. This concern often causes me to oppose the idea of a post-2015 governance indicator altogether, because I think these issues need to be decided upon in national policy discussions; not on the international stage. This said, a menu of governance indicator options could allow flexible enough choices to foster contextual fit. Such menu would include various potential ‘indicators’ and give countries a choice over the areas in which they want to pursue reform (and the mode of reform) and face ongoing evaluation (and the performance goals). A middle income country government might decide that it does not need to improve civilian registration but does need to work on the reliability of its public expenditures, for instance. International organizations would help the country on the latter issues and not the former, and assess the country’s governance reform progress by measuring the degree to which it is improving in its chosen areas (given the goals it chooses).
This kind of approach would see country-specific governance problems driving the global agenda, which is vital if one wants to generate a more action-oriented good governance agenda. It would require international organizations to pull back on many normative ideas about what ‘good governance’ looks like, however, and the dream of having comparative statistics of governance across countries. In place of these things there would be more opportunity to focus countries on specific governance problems, facilitate learning about what works and why, and monitor progress on real achievements. Such flexibility will help to overcome the limits of governance and institutional reform and push the governance agenda from its current emphasis on appearance to a new focus on action.
Matt Andrews is an Associate Professor of Public Policy at the Harvard Kennedy School and the author of The Limits of Institutional Reform in Development, Cambridge University Press (2013).