The bureaucratic arena refers to all state organizations engaged in implementing policy as well as in regulating and delivering services. Governance issues in the bureaucratic arena take on special significance given the massive pressures that have been placed on public agencies in recent years to become leaner, more efficient and bring services closer to the people. As part of a project to undertake comprehensive governance assessments, we focus here on the nature of the rules (formal and informal) that affect the bureaucracy. While issues of bureaucratic governance are not constitutive of development per se, they are seen as crucial determinants of the degree to which a country makes social and economic progress - or fails to do so. The rules that determine procedures in the bureaucracy, whether formal or informal, are especially important for public perceptions of how the state operates.

This paper presents our findings on the bureaucracy arena in 16 developing countries. Four observations stand out as especially important. The first is that bureaucracy is one of the more problematic arenas of governance in the countries in our study. Hiring is rarely on merit, bureaucrats are seldom seen to be accountable, and the operations of the civil service often lack real transparency. The second point is that the relationship between rules and structures, on the one hand, and performance, on the other, is difficult to establish. It is not that the link is unimportant. It is, but even disaggregated, it is hard to know what rules have what kind of effect. In our study, transparency and accountability proved to be the indicators where there is some evidence of a correlation with performance.

Our third observation is that reforming the bureaucratic arena is really difficult. One of our country coordinators summarized it well when he said, "This arena has proved hardest to reform in the years of transition and the average scores are just another evidence of these difficulties." It is clear that reforms take time to implement and even longer to have an impact on development outcomes. The fourth observation we want to make is that reforming the bureaucracy requires sensitivity to regime-specific issues. Context matters, but it is encouraging that level of development, according to our study, is not a critical issue. It is possible for poor countries to improve bureaucratic governance. Tanzania and Thailand are two very different countries where reforms have been undertaken and progress has been registered.

These are obviously interesting and important finding for all those who are in the business of improving public administration or public sector performance. Our study suggests that these are issues that should be given primary attention. This study also provides encouragement for organizations like Transparency International that is in the forefront of combating corruption by working practically on issues such as accountability and transparency. However, our study also suggests that more work is needed in order to throw more light on the issues where there is a deviation from the 'ideal model' advocated by agencies in the international development community. Above all, it must be recognized that reform is not merely a 'technical' issue. The bureaucratic arena cannot be treated in isolation from other governance arenas, although public sector reform efforts have tended to do so. Our findings indicate that a pure public administration perspective may be too narrow and technical in many cases and fail to focus on the real problem.