How can governance be measured to promote constructive change? This article from the World Bank summarises progress made in its initiative to test and develop policy-relevant, politically acceptable, quantitative indicators of governance. It suggests that, if consensus is key to developing governance indicators that promote institutional change, then sensitivity to the often-justified concerns of governments is key to that consensus.
Qualitative and case study analyses from the late 1980s and early 1990s began to focus attention on the importance of political institutions and governance for economic and social development. These studies generated a demand for work based on quantitative analysis. To date, analyses linking governance to a variety of development outcomes have relied on subjective and broad indicators of corruption or the rule of law. This work, based on a first-generation of governance indicators, was instrumental in drawing further attention to the crucial role of governance in development. However, the indicators can be of limited value in identifying high-payoff reforms and in building ownership for reform by developing-country governments.
There are two major principles involved in the process of generating indicators that are useful for practical reforms. Political acceptability is key in developing neutral quantitative benchmarks of good governance that can be embraced by reformers. Measures should also be institutionally specific so that reformers know which institutions to reform and how to do so. To ameliorate political and specificity problems to the greatest extent possible, the second-generation of governance indicators has embraced specific criteria that indicators must meet:
- Proposed indicators should be replicable through a well-documented process. The data should come from sources that are politically acceptable.
- Broad country coverage is necessary for testing relationships between indicators and valued outcomes. Ideally an institutionalised procedure should be in place or could be set up to collect data on the proposed indicator in the future.
- Indicators should be measured in a consistent manner across countries and values should reflect what the indicator claims to measure.
- Indicators should measure either a particular set of governance institutions or a defined output; and should not be unduly affected by forces exogenous to the aspect of government it is trying to capture.
In December 2001, the World Bank’s Governance and Public Sector Reform Group of the Poverty Reduction and Economic Management Network hosted a workshop concerning progress made with the second-generation governance indicators. Discussions at the workshop showed that:
- Support for the cautious approach adopted by the Bank in testing indicators for inclusion in a second-generation dataset.
- The tenor of discussions on governance indicators has changed since the March 2000 International Development Forum and there is now considerable interest from developing countries in cooperating with governance indicators.
- Some indicators could be added to the Millennium Development Goals.
- Other international organisations with broader mandates, which include promoting democratisation and human rights, are undertaking new measurement initiatives.
- These initiatives are expected to draw on the methodology and output of the second-generation indicators project.
