The emerging global architecture to support poverty reduction rests on the principle of mutual accountability. Donor countries are to be accountable for providing aid in ways that support country development strategies. Developing countries are to be accountable for using aid and other resources effectively. But, as this Global Monitoring Report (GMR) will detail, accountability with regard to aid resources is only a small part of the governance agenda. Both donor and developing countries are to be accountable more broadly for enhancing the checks and balances fundamental for development.
Part II of this GMR will spotlight these governance facets of the new architecture. The objective is to consider how to approach a satisfactory monitoring framework in the field of governance that is relevant for the scaling up of aid. This requires some analysis of relevant aspects of governance, a review of available data, and the use of some simple typology to allow for the wide differences in country situations.
There is strong evidence of a link between the quality of a country’s governance system and its development performance: Empirical evidence links governance to growth, over time—as, for example, in the thousand-year evolution of governance systems that underpin today’s developed countries—and across countries, though some debate persists as to causality.
Statistical evidence suggests that the causality between growth and governance is two-way—implying that gains in either can give momentum to a virtuous spiral of development improvement. Figure II.1 provides evidence of the causality from governance to growth: it illustrates the statistically robust partial relationship (controlling for initial income and schooling levels) between the quality of governance across developing countries in 1982 and income growth over the subsequent two decades.1
Governance and growth, 1982–2002
Source: Steve Knack, 2005.Note: Governance measure is an index from the International Country Risk guide (ICRG). The growth estimates are the unexplained residuals after controlling for the impact of initial levels of income and education.Governance and growth, 1982–2002
Source: Steve Knack, 2005.Note: Governance measure is an index from the International Country Risk guide (ICRG). The growth estimates are the unexplained residuals after controlling for the impact of initial levels of income and education.Governance and growth, 1982–2002
Source: Steve Knack, 2005.Note: Governance measure is an index from the International Country Risk guide (ICRG). The growth estimates are the unexplained residuals after controlling for the impact of initial levels of income and education.Econometric studies show that the benefits of public health spending on child and infant mortality rates are greater in countries with better governance—and that, as countries improve their governance, public spending on primary education becomes more effective in increasing primary education attainment (Swaroop and Rajkumar 2002).
The scale of corruption has also posed extraordinary costs on some countries. A conservative estimate is that the former president of Zaire looted the treasury of some US$5 billion—an amount equal to the country’s entire external debt at the time he was ousted in 1997. The funds allegedly embezzled by former presidents of Indonesia and Philippines are estimated to be two and seven times as high, respectively.2 Micro-level studies reveal the ubiquitous daily impact of corruption—and the benefits of scaling it back. In health care, for example, during the first nine months of a 1996-7 crackdown on corruption in Buenos Aires, Argentina, the prices paid for basic inputs at public hospitals fell by 15 percent. In customs, the use of private international firms to conduct preshipment inspection of imports has been associated with increases in the growth rate of import duties of 6 to 8 points annually.3
Many elements of the “development checklist” are governance related. Are there mechanisms in place to ensure that public resources reach their intended purpose with little leakage? Is the investment climate supportive of growth and reductions in income poverty? Can countries develop plans and do they have the institutional capacity to execute them? Is there adequate information and transparency in government to foster the active civil society to build greater accountability? Are the incentives and accountabilities of teachers and health care workers adequate to ensure low absenteeism and shirking? Does the rule of law protect the rights of citizens? The answer to these questions depends on the quality of national governance systems.
Getting governance of a quality needed to meet the Millennium Development Goals (MDGs) is not simply a matter for aid recipients. The global milieu has powerful influences on the governance system in developing countries. Global markets can be the source of virulent, corrosive corruption—or a powerful disciplining device, helping to strengthen developing-country governance. Donors and international institutions can provide aid in ways that can impose practices and reporting requirements that fragment and overwhelm already fragile governance systems, or in ways that help strengthen governance. Many of the areas noted in chapter 3 and addressed in the Paris Declaration are relevant for the governance agenda. Beyond aid, global check and balance mechanisms can provide new governance instruments for helping poor countries meet the MDGs. Recent work on standards and codes provides sources of good practice for all countries, and is increasingly being used to benchmark performance. Part II of the GMR therefore considers both national governance systems and the emerging global framework to support good governance.
Notes
The relationship remains robust with and without the inclusion of developed countries in the sample. It remains statistically significant, though somewhat weakened, when nations in the East Asia and Pacific region also are excluded from the sample.
Svensson (2005), quoting Transparency International’s 2004 report.
Svensson (2005), reporting results from Di Tella and Schargrodsky (2003); Yang (2005).
