accompanied by the specification of relevant targets or benchmarks, with program managers facing consequences for their success or failure in achieving these targets.

Because effective accountability usually requires appropriate benchmarks, it raises (directly or indirectly) the often misunderstood concept of conditionality. In Joseph Golds (1981) classic definition, conditionality is nothing more than a means for the appropriate use of resources. Because resources are never provided without reason or purpose, the relevant question is not whether conditions should exist. Rather, it is whether the conditions specified are relevant and sufficient-or unnecessarily burdensome, intrusive, or even counterproductive-and whether they are derived through genuine dialogue (essential for effective implementation of reforms with a significant institutional content)-or imposed from the outside. Even when an external accountability assessment is intended to be purely advisory, it is generally understood by the aid agency and the recipient government that the assessments conclusions will be taken into account in the future provision of assistance.

So, in developing countries the operational issue is not whether public financial accountability is fully adequate. In most cases it is not. The operational issue is whether accountability is being strengthened at a sufficient pace to warrant representing to aid providers and recipients that fiduciary and development objectives are more likely to be achieved, partly through the use of the aid. This in turn calls for measures to improve the legal and organizational framework for public expenditure management, systems and processes for expenditure programming and budget preparation and execution, accounting, reporting, and audit, and administrative and financial management capacity. (As discussed below, these are the five broad headings under which components of public expenditure management are grouped in this study for the technical mapping of the various assessment instruments.)

OPERATIONAL IMPLICATIONS

The preceding discussion has four main practical implications. First, it can be misleading to assess overall risk by adding up the individual weaknesses in a public expenditure management system. The three levels of public expenditure management are interrelated, as are the various parts of the administrative apparatus. Just as the balance of payments or fiscal accounts must be analyzed as an interrelated whole, the assessment of the risks asso-

ciated with untied budget support should be based on the entire system for public expenditure management. In some cases the overall risk may be smaller than the sum of individual risks; in other cases it may be larger. What is needed is judgment, not a mechanical exercise.

Second, because development agencies have differing goals and objectives as regards their aid programs and development assistance, their views of risk differ as well-as do the objectives, coverage, and methodologies of their financial accountability instruments. This is a major reason for the duplication between the instruments used by different agencies.

Third, if an instrument covers a broad range of issues, it is bound to be more diffuse than an instrument focused on a specific component. It is tautological to say, for example, that a procurement review is more narrowly targeted than a public expenditure review.

Finally, a given instrument may cover many components of public expenditure management but some only superficially, while another instrument may have narrower coverage but greater depth. With these considerations in mind, the next section describes the key features of the main assessment instruments.

The Main Instruments for

Assessments

Donors use six main instruments to assess public expenditure, procurement, and financial accountability: World Bank Public Expenditure Review (PERs), Country Financial Accountability Assessments (CFAAs), and Country Procurement Assessment Reports (CPARs),1 IMF Reports on the Observance of Standards and Codes of Fiscal Transparency (Fiscal ROSCs), World Bank-IMF Public Expenditure Tracking Assessments and Action Plans for Heavily Indebted Poor Countries (HIPC AAPs), and EC audits. In addition, the U.K. Department for International Development (DFID) is developing a new approach to assessing fiduciary risk based on a methodology similar to that of the HIPC AAPs.2

All these assessment instruments share common approaches and objectives, including describing and analyzing systems for public expenditure, procurement, and financial accountability, recommending improvements, increasing financial integrity, and improving budget outcomes. Assessments also foster another crucial outcome: increased interaction between development agencies and recipient governments (see box 1). Yet the instruments emphasis, coverage, and approach vary considerably, reflecting the fact that they were developed by different agencies or different parts of the same agency. The instruments are summarized below and described in more detail in annex 1. (Because the DFID approach to assessing fiduciary risk is still being developed, it is not part of the mapping exercise-though it is discussed below and in annex 1.)