Copenhagen Consensus Center
Home Menu

Post-2015 Consensus: Governance and Institutions Assessment, Hilderbrand

Summary of the targets from the paper

Governance and Institutions
Target Benefit for every dollar spent
Provide legal identity for all, including registering all births More than $1
Develop effective, accountable and transparent institutions at all levels Uncertain
Ensure responsive, inclusive, participatory and representative decision-making at all levels Uncertain
Ensure public access to information and protect fundamental freedoms, in accordance with national legislation and international agreements Uncertain
Substantially reduce corruption and bribery in all its forms Uncertain

Summary

Governance and institutions have been among the focal themes of planning for the post-2015 global agenda, to follow the soon-to-expire Millennium Development Goals (MDGs). It is also important to people more generally: in a global survey conducted by the UN (“The World We Want”) respondents named “an honest and responsive government” as third priority, just after a good education and better health care, and rated it as a higher priority than 13 others.

Some assessments of performance on the current MDGs have blamed poor governance when results were disappointing. Another perspective is that “good governance”, especially components such as participation, transparency, inclusiveness, and access to justice, is a part of development itself; this latter view is particularly prominent in UN statements on the post-2015 agenda. Nevertheless, there are many problems and challenges associated with setting appropriate goals and targets.

Although the papers in this series focus on benefit-cost analysis, governance targets raise a set of fundamental questions that need to be addressed first. Also, the costs and benefits are less amenable to quantitative measurement and only general estimations are possible, at best.

Goal 16 proposed by the Open Working Group – “promote peaceful and inclusive societies for sustainable development, provide access to justice for all and build effective, accountable and inclusive institutions at all levels” – is the most relevant to governance. The range of proposed targets is very broad, and this paper focuses on the key subset of developing effective, accountable, transparent and inclusive institutions. Of the 12 targets proposed, the following are considered:

16.5   substantially reduce corruption and bribery in all its forms

16.6   develop effective, accountable and transparent institutions at all levels

16.7   ensure responsive, inclusive, participatory and representative decision-making at all levels

16.9   by 2030 provide legal identity for all including birth registration

16.10  ensure public access to information and protect fundamental freedoms, in accordance with national legislation and international agreements

Some definitional matters

The terms “governance”, “institutions” and “institutional reform” are widely used, but there is little agreement on definitions. One set of definitions refers to how public authority is exercised and decisions made.  A second definition makes a distinction between “government” and “governance” and uses the latter to refer to the specific idea that contemporary policy problems are addressed by government, the private sector, and civil society work together to get things done. I use governance and institutions, and governance reform and institutional reform, interchangeably while dealing with dimensions of what is generally considered as “good governance”, including effectiveness, transparency, and accountability.

Governance on the Development Agenda

The emphasis on governance in the post-2015 proposals is an expression of an existing governance agenda; despite the lack of a target in the MDGs, governance and institutions have been central to the work of development agencies over the past couple of decades. Thinking about development began to change in the mid- to late-1990s, moving from a primary focus on economic policy and putting a new emphasis on institutions. Effective states began to be seen as essential for properly functioning markets, as well as for social and human development.

Other trends also contributed to the governance agenda. In the context of globalization, work on international competitiveness also began to call attention to the quality of institutions. There had also been public sector reforms in developed countries, with an emphasis on setting goals and holding managers and organizations accountable. Similar concerns about improving performance developed in the development assistance arena. IFIs and regional development banks began to ask borrowing countries not just for policy reforms but for efforts on governance, including in particular strengthening financial institutions and accountability. Evidence that poor governance—including corruption—hinders development became the basis for an active and quickly expanding governance agenda for developing countries.  In addition to donor pressure, some political leaders undertook public sector and other institutional reforms out of a desire to make things work better, with Singapore becoming a model of how strong institutions can make a difference.

Issues in Measurement

A major difficulty for both analysis and policy relating to governance and institutions is the problem of measurement. Among the most widely used measurements are the World Bank’s Worldwide Governance Indicators (WGIs), introduced in 1996, which are based on statistical aggregation of multiple surveys, and are mostly perception-based. Another widely-used metric is the Quality of Government from the University of Gothenburg, which focusses on the structure and behavior of public administrations via expert surveys. There has been an explosion of indicators in recent years, but in all cases the major concern is the weakness of the underlying data, which are subjective and not very reliable.

Indices that add up several different measures and put them into one summary measure are very effective in getting attention for issues, as they are visible and seemingly easy to understand.  But they are not very useful as a measure of what is actually being done.  On the other hand, single, narrowly defined indicators do not capture the multidimensionality of governance. “Dashboards” have been advocated to provide a wider range of information. However, they are not easily comparable between countries, and still risk repeating the problem of the unmanageable number of expectations, which has characterized the governance agenda.

Governance and Development

Governance and institutions or institutional reform are considered to affect development through a variety of pathways, including their effects on:

  • investor confidence and willingness to invest;
  • facilitation of the working of the market;
  • effective functioning of government (in managing the economy, delivering services, etc.);
  • effective and efficient allocation and utilization of public resources for the public good;
  • citizens’  opportunity and ability to participate and hold government accountable.

Most of the arguments relate to economic growth, in particular that certain institutions, including property rights and the rule of law, are essential to allow the free market to operate and increase investor confidence. There are also those who are more concerned about poverty reduction, and emphasise other government functions. The level of corruption is of concern to both groups, and is argued to hurt the poor disproportionately.

The global governance agenda has relied substantially on research findings from around the turn of the century, which concluded that institutional factors explained differences in levels of development, and particularly that security of property rights was the most important single factor behind variation in wealth of countries. However, the central place of these institutions has also been criticized. The claims that the findings indicate causality rather than simply correlation have been questioned, and any causality which does exist may actually be in the opposite direction, with good governance itself being a result of economic development. Also, the argument that governance changes are required as a prerequisite for development is seen as ignoring historical evidence to the contrary. A final criticism is that the governance agenda as currently practiced is enormous and unmanageable for countries that are asked to make improvements across the board, and there is not strong enough evidence about the effect of particular reforms to be able to easily prioritize and narrow the agenda. 

More recently, a series of studies have analysed governance reform efforts. One showed that, of 145 countries that had introduced institutional reforms between 1998 and 2008, government effectiveness improved in half the countries and worsened in half. Analyses of transparency and accountability initiatives (TAIs), defined to emphasize citizen-led initiatives, find similarly mixed results. Similarly, most analysts conclude that, despite the enthusiasm for and large investments in participatory efforts, the evidence in support of them is weak. Community engagement in health and education service delivery was found to be somewhat positive, but the causality was described as “vague”.

Overall, then, there is little clear evidence for direct causal links between governance improvement and broad development outcomes. The causal relationship seems to be more nuanced than cross-national studies suggest, and we can conclude that context is very important in shaping outcomes and must be taken into account when designing and implementing initiatives. It seems that it is difficult to come up with interventions or targets that would be generally applicable.

What do we know about how to improve institutions/governance?

The political will of top leadership and lack of country ownership of donor-sponsored programs are often cited as key factors in disappointing results. However, these are both necessary but not sufficient for success. Despite the tendency to talk about reforms as if they are technical, changes in institutions and organizations are deeply political; they require broad participation, negotiation, and adjustments if they are to work.

There is also, in the view of some researchers, too much emphasis on the “form” of institutions. Donors and other international actors ask countries to adopt particular institutional forms that are considered international “best practice”.  This is generally not successful and may be harmful, for a variety of reasons. In particular, institutional reform efforts often try to transplant a formal structure into a setting where it does not have the more organically developed, informal underpinning that is required to make it work. In particular, this work implies that those individuals and organizations supporting reforms internationally should take care not to be setting requirements, or targets, that push best-practice solutions on countries and that create incentives for reforms on the surface that do not actually mean better governance in the country.

Issues in setting governance targets

Establishing goals, setting targets, and measuring progress toward the targets come primarily out of the performance management approach and are integral to it. However, success requires skilled managers who are committed to the idea and to the goals, who communicate all the way through the organization, and who organize the people and effort around the goals. Also, if the targets or how they are measured do not accurately get at what is intended, they can create incentives that lead to effort that does not help reach the goals. There is indeed disagreement about whether setting international development targets is sensible in general, and there is more intense debate about measuring and setting targets with regard to governance performance.

There are undoubtedly daunting challenges for using targets for governance, and many pitfalls to be avoided. There are many difficulties in setting measurable targets and real risks that government officials will find ways to meet goals which are window-dressing rather than real reform. Setting the same targets for all countries assumes that those targets are relevant for all countries, which is certainly not the case for governance.

Targets can be categorized in various ways. The standard approach in public management is to distinguish between outcome and output or process. The MDGs were explicitly outcome measures, and post-2015 proposals are couched in the same terms. However, a criticism of this approach is that it can lead to a short-term perspective that finds a way to get the results but neglects the longer-term and more difficult institution or capacity building. In the governance area, outcome measures are not always a good fit, and it can be difficult to agree on meaningful targets.

A more recent categorization is by form versus function. By adopting international standards for particular processes (the “form” approach), there is a high risk of engaging in “rituals of reform” rather than real improvement. Function targets, on the other hand, focus on improving performance of key functions are quite specifically defined. Examples across a range of areas include birth registration, keeping citizens safe (for instance, in terms of road safety or crime rates) and food security.

It is also possible to categorize targets by breadth. Broad governance targets such as greater transparency make intuitive sense, but it is not always clear what they include and it is very difficult to measure them.

Assessing governance targets

There are challenges to the use of cost-benefit analysis for governance, since neither benefits nor costs can be easily quantified or monetised for targets such as increased transparency or participation. One of the most significant limitations is that costs, as well as benefits, of implementing particular reforms or approaches will vary in different contexts. Assessment of proposed governance targets can therefore only be done at a very general level and should include not just economic costs but factors such as the likelihood of success, the quality of the data and the administrative and political challenges.

This assessment of five of the proposed OWG targets, plus some additional ones compiled by Foresti and Wild, is less to conclude that a particular target is better than another, but more to indicate that particular kinds of targets are likely to be more helpful and more productive of good results than others, and to suggest considerations that need to be taken into account.

With the very broadly defined OWG targets, it is impossible to estimate even in a general way what costs would be, because there is no way to know what measures would or would not be taken toward the target. Estimation of benefits has to rely on possible (but unproven) linkage with broad development outcomes or some subjective valuation of the governance dimension itself. Target 16.9 is different from the others:  providing legal identity to all, including registering all children at birth.  In particular, the second part is a very specific function as a target. It is clear, concrete, and could serve as an incentive for action. It has several advantages:  it requires considerable organizational and administrative capacity to accomplish, including reaching throughout the entire country.  That capacity, once built, could have spin-off effects for other work that the government needs to be able to accomplish that required similar capacity.

There are a number of other potential targets in the area of effectiveness. For example, there are several potential targets or indicators relating to public financial management that seem promising in terms of institutional effectiveness:  improved bill-paying by government, relationships between budget plans and expenditure, better revenue generation. Cost-benefit assessments for these and other similar targets seem likely to be generally better than average, although expenditure tracking surveys involve substantial cost.

There are several transparency-related targets which are really about adopting particular forms, such as Freedom of Information laws.  If the target involves only adoption, costs are low but benefits also low.  If the target is about actually implementing the form or standard, then benefits may potentially be higher, but costs will certainly be higher.

Foresti and Wild list a number of proposals in the freedoms and participation category. There are several indicators of broad governance outcomes, but the issue here is less cost than political sensitivity and the degree of willingness of powerful political leaders to be willing to open up the political system. The one proposed target that is defined in narrower, more functional terms, is about citizen monitoring of public services, but there is no clear evidence that this is actually generally effective at improving outcomes. Trying to set targets in this complex and very political dimension seem likely to lead to efforts that may be more window-dressing than real.

Reduction in bribery does seem a valuable target, especially if more broken down into more specific types. Transparency approaches for addressing corruption would be somewhat less costly and sometimes (but not necessarily) less difficult politically than monitoring and enforcement approaches. Getting rid of red tape and streamlining regulations to minimize opportunities for rent-seeking would generally have a lower cost.

In more general terms, then, broadly stated governance goals are aspirational but not very helpful as real targets that affect behavior and performance. Outcome targets are generally preferred over outcome and process types of targets, but they are quite difficult and not always possible with regard to governance and institutions. Overall, “form” targets may be less costly than other types, to the extent that they may not involve real implementation of the institutional change, but they also fail to provide benefits. Potential benefits also have to be weighed against potential risks and problems.   Some of these have to do with the nature of the target and the data/measurement problems.  Others have to do with the difficult of implementing reforms to meet the target, including both political and administrative challenges. 

Function targets that are narrowly defined and focus on outcomes come out as the most promising, especially where they are concrete enough that there is something that can be measured. As an indicator of government effectiveness, and with other significant benefits, the proposed target for birth registrations is very strong.  Other similar targets in other areas where government effort and capacity is important for citizens and for the economy might be similarly useful.  Targets for increasing transparency and access to information in key areas, perhaps especially budget formulation and implementation, may also be promising.  Anti-corruption targets that sort out specific areas of corruption and set targets for important, measurable ones, would be likely to offer benefits, but may be difficult to implement.

Conclusions

It is important to avoid the post-2015 goals just being an extension of the global governance agenda of the past two decades, with its plethora of goals, large investments, and disappointing results. We can draw some general conclusions from the analysis in this paper.

  1. Government effectiveness and anti-corruption are two areas in which targets are more likely to be possible.
  2. In general, it is neither appropriate nor effective to set goals that involve the adoption of particular institutional forms, or solutions.  Targeting important functions and letting countries decide how to address them is more likely to produce good results.
  3. Single composite measures tend not to be terribly helpful; limited dashboards of more disaggregated targets would provide better guidance for policy and assessment.
  4. People and countries will disagree about how to prioritize institutional reform. Countries at different levels of institutional development and with different configurations of strengths and weaknesses will have different challenges, so flexibility must be built into a system of goals and targets.
  5. Based on past experience with the governance agenda, it is important to avoid the perception (or reality) that governance targets are being imposed on low-income countries by rich and more institutionally developed countries. 
  6. Although the evidence linking broad governance dimensions and broad development outcomes such as growth and poverty reduction is not clear, within sectors such as education and health the improvement of institutions such as delivery systems and accountability are more closely connected to potential improvements in sectoral outcomes.  Relevant, sectorally defined governance targets could be usefully included as part of other development goals, rather than separately.
  7. It is important to invest (both globally and within countries) in data collection for relevant indicators and in building the capacity and prominence of statistical agencies.