Developing a single concept memorandum for the work.

Strengthening peer review.

Ensuring the fullest possible participation and leadership of recipient governments.

Promoting the full exchange of information and analysis between project teams throughout the work, and coordinating the missions and work programs of these teams as much as possible.

Sequencing and prioritizing recommendations and action plans for implementing agreed reforms that reflect the capacity of the recipient country.

One approach would be to combine PERs-or at least their public expenditure management components-and CFAAs into a single product. But while this idea has theoretical attractions, it also raises practical difficulties. Combining the two instruments would increase the scope and length of an already lengthy report (unless PERs were split into components based on a sequence of activities spread out over time by use, for example, of the government-led approach described above for Tanzania). Moreover, separating the public expenditure management components of PERs from the expenditure analysis and policy components would effectively revert to a long-discarded approach to PERs, and would lose the essential connections between assessing policy options, preparing budgets, and delivering effective public services.

Still, the idea of integrating PER and CFAA processes has merit. It would ensure that economists, public management specialists, and financial management specialists work together more closely and that PERs and CFAAs are seen more as a single process of assessment work, reaching toward a common goal, than as separate products with few connections. Thus the recommended approach is to foster such coordination and collaboration by preparing consistent (or perhaps consolidated) guidelines and questionnaires, and changing Bank rules and procedures to facilitate the joint planning of PER and CFAA activities, joint missions, and, where appropriate, joint products.

A first (minimal) step in the right direction appears to be feasible. The terms of reference for PERs and CFAAs, with sufficient detail on key issues and proposed approaches, should require formal reciprocal clearance by the relevant sector directors in the Financial Management network and Poverty Reduction and Economic Management network, respectively. (Subse-

quently, as proposed earlier, these terms of reference should be subject to formal comments by the IMF-which should reciprocate with Fiscal ROSCs-and then be submitted to recipient governments for consideration and approval.) The goal should be to make the two exercises start from a common basis, and not attempt ex post reconciliation, amalgamation, or dilution of their findings.

Informal consultations on the two instruments, though common, have often been insufficient. In some cases a PER and a CFAA have been carried out in the same country, sometimes even during the same period, with no substantive interaction between the two and only perfunctory recognition of the others results in the eventual reports. Formal reciprocal clearance, with its concomitant responsibility, could do much to resolve coordination problems upfront without requiring major organizational changes or preempting more comprehensive long-term solutions.

For CPARs, which have a relatively well-defined and narrow focus, coordination requirements are less demanding. The review, however, should be conducted with full awareness of a countrys corruption and governance problems rather than as an exercise that considers government procurement an isolated function. But this approach is already implicit in CPAR guidelines and can be achieved by adding a governance expert to CPAR teams, by engaging in consultations on terms of reference with the appropriate unit in the Poverty Reduction and Economic Management network (without formal clearance by sector directors), or by doing both. It is also important that PERs reflect the findings of previous CFAAs and CPARs. Thus the PER concept note should always include reference to procurement issues, and the PER process should incorporate CPAR findings and recommendations. These practices, already common, should be made explicit in CPAR and (eventual) PER guidelines.

A recently launched review of the Banks public expenditure work should help draw together the issues and concerns discussed above and propose practical solutions. The central concepts underpinning the review-that public expenditure work should be country-focused and country-driven, linked to Poverty Reduction Strategy Papers (where they exist) and donors aid and technical assistance programs, and cover all relevant components of the public expenditure management cycle-are the same as those advocated in this study (for further details, see the section below on the proposed programmatic approach).

A final minor but important point involves Bank terminology for the initial documents for fiduciary economic and sector work. There are initiat-

ing memorandums, concept papers, concept notes, issues papers, issues notes, and initiating concept memorandums referring to the same basic documents. The use of these different terms sometimes causes confusion inside and outside the Bank. The Bank might consider reserving initiating memorandums for a structural adjustment loan (as has been done since the late 1980s) and settling on one of the remaining terms for PERs, CFAAs, and CPARs.

INSTITUTIONAL AND GOVERNANCE CONSIDERATIONS

All the assessment guidelines should include a brief definition of institutional and governance issues and explain their relevance to analyses of public expenditure management. The generally accepted definition of governance is found in World Bank (1992, p. 1):

The exercise of authority, control, management, power of government. A more relevant definition for Bank purposes is the manner in which power is exercised in the management of a countrys economic and social resources for development. The Banks main concern with sound development management thus extends beyond building the capacity of public sector management to encouraging the formation of the rules and institutions which provide a predictable and transparent framework for the conduct of public and private business and to promoting accountability for economic and financial performance.

Governance is essentially about decisionmaking and implementation processes and the capacity underpinning these processes. What makes governance good or bad is the level of transparency, accountability, predictability, and participation. But as noted, it is difficult to use these four characteristics to assess the overall governance of public expenditure management, and much easier to do so for a specific area such as procurement, audit, or internal control. The extent of transparency, accountability, predictability, and participation is determined by the underlying rules of the game-that is, institutional arrangements. Thus one can assess these dimensions by observing the rules in action.

When analyzing a specific area, staff could be encouraged to ask whether information and decisionmaking are transparent, how individuals are held accountable for their actions, if regulations and rules are enforced effectively and uniformly, and the extent to which appropriate participation is