cambodia: prsp and budget linkages - world...

30
Draft For comments Please do not quote without permission CAMBODIA: PRSP AND BUDGET LINKAGES Jeni Klugman (PRMPR) and Rob Taliercio (EASPR) October 2003 The authors are grateful to Lindsay Judge for excellent assistance and advice and to Dr Hang Naron, Secretary General, Supreme National Economic Council, Royal Government of Cambodia for invaluable discussions, and Catherine Dom for valuable comments. Thanks to the Government of the Netherlands for financial support. This draft reflects the views of the authors and not those of the World Bank.

Upload: lamdat

Post on 07-Jul-2018

217 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

Draft For comments

Please do not quote without permission

CAMBODIA:

PRSP AND BUDGET LINKAGES

Jeni Klugman (PRMPR) and Rob Taliercio (EASPR)

October 2003

The authors are grateful to Lindsay Judge for excellent assistance and advice and to Dr Hang Naron, Secretary General, Supreme National Economic Council, Royal Government of

Cambodia for invaluable discussions, and Catherine Dom for valuable comments. Thanks to the Government of the Netherlands for financial support. This draft reflects the views of the authors

and not those of the World Bank.

Page 2: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

Acronyms CSD Council for Social Development IFAPER Integrated Financial Assessment and Public Expenditure Review MEF Ministry of Economy and Finance MOEYS Ministry of Education, Youth and Sport MOH Ministry of Health MOP Ministry of Planning MTEF Medium Term Expenditure Framework NPRS National Poverty Reduction Strategy ODI Overseas Development Institute PAP Priority Action Program RGC Royal Government of Cambodia SEDP Socio-Economic Development Plan SEF Sector Expenditure Framework

Page 3: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

1. Introduction and Overview This paper is one of a series of country studies about PRSP-budget linkages in different regions of the world. The Royal Government of Cambodia (RGC) finalized its National Poverty Reduction Strategy (NPRS) in late 2002, and it is an early stage to assess its linkages to the budget and its impact on budget processes. This case study thus focuses mainly on the preparation process, the budget picture and commitments presented in the NPRS, and relevant developments on the fiscal management side. It draws on recent relevant analyses, in particular a study by the Overseas Development Institute (ODI) of results-oriented public expenditure and a review of financial management and public expenditure by the World Bank (IFAPER). Since the budget and, more broadly, fiscal issues, lie at the heart of PRSP implementation, we analyse the existing linkages between the NPRS and the budget with the aim of extracting lessons from Cambodia’s experience and providing suggestions as to how the potential synergies might be better exploited in the future.

Some key initial conditions and contextual factors are highlighted at the outset for readers unfamiliar with Cambodia. First, decades of civil strife ended only in 1993, and macroeconomic stability is even more recent (since 1998). Second, since 1997 Cambodia has been governed by a somewhat tempestuous coalition, which has resulted in coordination problems at both the political and bureaucratic levels. Third, and relatedly, there is not a strong tradition of openness and transparency, and high level decision-making is quite centralized. Fourth, not only is the level of average income very low, revenue collection is low even by developing country standards: fiscal revenue was under 12 percent of GDP in 2001, compared to about 18 percent for low income African countries and 16 percent for low income Asian countries. Fifth, there is substantial reliance on external assistance and an active local aid community. Finally, it is worth noting that the RGC is strongly committed to growth, and increasingly to poverty reduction objectives, as well as to economic and regional integration.

The NPRS was prepared by the Council for Social Development (CSD), an inter-ministerial body chaired by the Ministry of Planning (MOP). It builds on the preceding medium-term development plan (Socio-Economic Development Plan for 2001-2005 (SEDP II) ) and the Government has committed to henceforth merge the SEDP and NPRS processes, both in terms of substantive development and monitoring.

As highlighted in the official assessment by staff of the World Bank and IMF (JSA), the main strengths of the NPRS are: (i) the relatively open and extensive participation in its development, including consultations with the poor, ongoing engagement with NGO groups and interactions with the donor community; (ii) the engagement of virtually all central oversight and line agencies in the process, so that the NPRS is closely linked to strategies owned by the agencies; (iii) the advances on the SEDP II and I-PRSP, including deeper treatment of some key cross-cutting issues, in particular governance and private sector development, and greater operationalization of sector policy in some instances; and (iv) initiation of the development of indicators to monitor progress in poverty reduction.

Page 4: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

Among the major shortcomings of the NPRS identified in the JSA however were the lack of prioritization, and the need to strengthen links to the annual budget and to development of the Medium-Term Expenditure Framework (MTEF). A more general problem was that, at the time of finalisation, the NPRS was overlaid onto existing processes, without rationalizing nor clearly identifying the role of the NPRS in terms of existing budget processes. These and other challenges are explored more fully below.

This paper is structured as follows. The next section reviews the nature of budget processes, highlighting areas where linkages to the NPRS might be strengthened. Section 3 turns more directly to the allocation of public resources in Cambodia, seeking to assess the extent to which the Government’s commitment to poverty reduction is realized in the budget. Section 4 looks at the type of fiscal data that is regularly produced, on the grounds that information is a key building block for analysis and decision-making. Information is also critical to the topic of participation and current mechanisms to ensure accountability and ways in which results and feedback might be better utilized for decision-making, which are addressed in Section 5. Section 6 addresses some key issues related to external financing, before we conclude with suggestions as to how the potential links between the budget and NPRS might be strengthened in the future.

2. The Budget Processes

While there is overall substantive congruence between the priorities set out in the NPRS and the shape of budget allocations, there was an actual disconnect between the two processes with budget decisions preceding and not being directly informed by work and consultations around the NPRS (table 1). Over time, the respective strengths of both processes could be mutually beneficial and improve future quality and transparency of each.

Table 1: Key Steps in the NPRS and Annual Budget Timetables,2002-2003

Month NPRS activities Annual budget/ MTEF steps

March 2002 Provincial consultations MEF preparation of macro-economic and overall budget framework, including agencies’ recurrent budget ceilings

April/May Preparation of line agency submissions

Budget framework submitted for approval by Council of Ministers

June/July First draft of NPRS MEF Budget Circular officially informing agencies about their ceiling

Aug./Early Sept. Feedback on NPRS draft NGO inputs

Timetabled face-to-face MEF/ agency budget hearing meetings

Sept./Oct. Second national NPRS workshop MEF consolidated draft annual budget submitted to COM approval

Page 5: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

NPRS workshop with parliament

Nov./Dec. Further revisions of NPRS based on feedback from line agencies, donors and stakeholders

Senate and Assembly review and debates

Late Dec./Early Jan. NPRS approved by COM New budget approved by National Assembly

Jan./Feb. 2003 NPRS transmitted to Bank and Fund

NPRS presented at CG meeting

MEF appropriation Circular – New budget execution started

The NPRS

A key strength of the NPRS was that the process became more open and consultative through time and has been described by independent observers as “exemplary by Cambodian standards” (ODI, 2002). A number of local level consultations complemented active interest and important contributions at the national level by NGO networks. There was stakeholder involvement from fairly diverse groups, representing women, ethnic minorities, policy research institutes and academics, private sector, trade unions, and different regions of the country. After a slow start, the process successfully engaged virtually all line agencies in a series of iterations in the development of the strategy. One limitation however was that the decentralized commune and village governance structures were not enlisted as part of the NPRS consultative process, though participatory planning processes are being implemented at these levels. Parliamentarians’ involvement was also limited to participation in workshops. However, overall both the NPRS team, and the final paper, demonstrated significant responsiveness to stakeholders’ concerns relative to Cambodia’s technical, political, and budgetary constraints.

The sectoral and structural measures developed by the respective line and core agencies were subject to feedback and iterations during the course of NPRS development. The extent to which these policies and programs present a clear and realistic approach to improving the well-being of the poor varies across sectors however. Agriculture and infrastructure issues arguably receive the weakest treatment in that:

• Many proposed activities in the matrix do not have identified sources of financing and their total cost is beyond the amounts likely to be available;

• Consistent with the pre-existing conventions and the separation of capital and current budgets, line agencies tended to adopt an incremental approach, limiting their review to new and mainly investment project activities rather than undertaking a systematic assessment of the range of existing and proposed activities to determine these with the greatest impact on the poor.

Page 6: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

• The focus is on projects, with relative neglect of policies and strategic issues, which meant that the NPRS failing to present a prioritized set of recommendations.

The NPRS treatment of health and education was much stronger, drawing on existing sector strategies that set out a largely pro-poor policy backed up by policies developed in companion documents. These strategies are reflected in the NPRS, and through SEFs, are better linked with the annual budget and integrated into the MTEF.

The Cambodia experience shows that, thus far, the NPRS process has not managed to promote sector policy and strategy development: where sectors already had policies and strategies, the NPRS built on these but other sectors did not make much progress in developing policies and strategies through the process. However, in less advanced sectors the value-added of the NPRS process has been to motivate line ministries to begin to think about costing and budgeting their programs. That said, most sectors are at an early stage of developing linkages between their policies, strategies, programs, and the budget process. Moreover as highlighted below, the continuing role of the Public Investment Program (PIP) tends to undermine rather than promote a strategic programmatic focus.

Looking ahead, a key challenge is the need to work out how the NPRS participatory process, including the continued engagement of various ministries and agencies and in particular MEF, can be integrated with ongoing Government processes for policy and decision making. An important aspect of this challenge lies in the linkage of the NPRS to the budget, and vice versa. While MEF became substantially involved in the NPRS process, they did not exercise a leadership role (which had been given to the Ministry of Planning (MOP)/GSCSD. As noted above, while there is an NPRS chapter on budget allocations and NPRS priorities are generally being accorded budgetary priority, the process for developing these allocations was not directly linked to the NPRS process. The reasons for the disconnect include:

• The fact that the NPRS timetable was not built around the annual budget process and that the NPRS timetable itself slipped somewhat;

• The nature of the budget process – in particular the traditionally bilateral negotiations between MEF and the respective spending agencies, based on line item budgets – as reviewed further below;

• The nature of the NPRS process – that the MEF became intensively involved relatively late and did not lead the process, and that line agency proposals were only partially costed and not linked by those agencies to their work on the annual budget;

• The reliance on the PIP – in particular in the matrix of the NPRS – which is understandable in terms of the MOP’s traditional focus, and thus on a mechanism which diminishes the role and accountability of line agencies (relative to the MTEF and PAP in selected sectors).

• Capacity limitations at the line agency level, and in the MOP/GSCSD; and

Page 7: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

• Issues of coordination between MOP and MEF, and between the central oversight and line agencies in general, which resulted in disconnects in the budgeting and NPRS processes.

Weaknesses in the NPRS process itself also contributed to the planning-budget disconnect. One issue highlighted by the ODI study was the absence of clear guidance to line agencies on criteria that should be used to prepare proposals. Initially agencies were advised to limit their submission to targeted programs (i.e., only that subset of their activities that they feel benefit the poor), but were subsequently advised that their submission should reflect their entire programs and highlight the poverty relevant aspects The latter is consistent with Bank/Fund expectations that PRSPs should be comprehensive -- see JSA guidelines – although some, in particular for the highly indebted countries that completed several of the early PRSPs, have presented only a subset of activities. The NPRS policy and program matrix lists a large number of activities, of which only about half were costed by the respective line ministries. The budget chapter of the NPRS does recognize that the NPRS activities and projects that are listed in the matrix are not fully resourced and any costing were described as preliminary only. Given analytical limitations and lack of experience, it was not possible for the NPRS to develop full costings of the policies and programs proposed. There was some discussion among Government agencies and donors about the extent to which the NPRS should develop costings. However, given the manifest disconnect with the budget process, it was decided that it would not make sense to cost the NPRS in a budget vacuum. Costing the NPRS would have been a technical exercise devoid of effective linkages to the annual budget which experience elsewhere, for example in the former Soviet Union (Moon 2003) suggests is not at all useful. A disconnect between the PRSP and the budget formulation process is not unusual even in countries where Ministries of Finance do play a leading role in the PRSP process. At the same time international evidence firmly suggests that strong links between policies and programs and the strategic approach set out in the PRSP, and the annual and medium term budget, are critical to the implementation and credibility of national PRSPs.1 We thus turn to review the budget and MTEF process in Cambodia and the scope for synergies with the NPRS, and propose ways to strengthen these linkages over time. 1 See Board papers (2002a; 2002b; 2003), Booth and Lucas (2000); ODI (2003)

Page 8: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

The Budget Process Both the annual budget and MTEF processes reflect a number of “classic” problems which plague many developing countries. These are well-documented for Cambodia (see IFAPER, 2003) and so are only listed here. The key issues relate to: A. Institutional Fragmentation and Donor Dependence

• There is a relatively large number of line agencies many of whom have very limited operating budgets (less than US$1 million annually). Having multiple agencies responsible for broadly similar policy and program areas results in fragmentation and some duplication of programs, exacerbating weak capacity and increasing the risk that relevant agencies will be bypassed. This is especially problematic for rural development and infrastructure.

• Capital and recurrent budgets are prepared separately, with the former being centrally determined (effectively by the Prime Minister in some cases). Thus, for example, the Ministry of Public Works and Transport (MPWT) plays only a marginal role in setting sectoral priorities and managing domestic resources for the sector since MEF controls the bulk of the capital budget and sometimes allocates and disburses funds directly to provincial departments of public works and transport and military-related contractors, bypassing the central ministry.

• The Public Investment Program (PIP) has had little impact on either the expenditure management process or expenditure outcomes. MOP is tasked to appraise line ministries’ proposals to determine whether they meet government’s development priorities and whether costings are realistic, though capacity and other constraints mean that proposals are subject to only cursory review. There is no strategic discussion with the respective line agencies. Allocations for capital expenditures, through the budget’s Chapter 50, are supposed to reflect the domestic financing requirements identified in the PIP. In practice, no such allocations are made. Agencies recognize that the PIP plays no meaningful role in investment decision-making, line agency submissions are tardy, with scant information on proposed projects and virtually none on actual disbursements, while preparation and Council of Ministers approval of the annual PIP tends to slip outside the budget calendar or—as in 2001—is dropped altogether. The result is that the PIP has lost credibility as a planning tool. With the introduction of the MTEF, attention needs to turn to the rationalization and integration of the programming instruments.

• Separate processes for managing civil service reform. As evidenced by the execution rates for the wage bill in 2002 (118% for civil administration and 106% for defense and security), the wage increase programmed in the Strategy to Rationalize the Civil Service (SRCS), approved by the Council of Ministers in October 2001, had not been incorporated into the 2002 budget. Moreover, the projected wage increases for the period 2003-2006, as presented in the SRCS, have not yet been incorporated into the MTEF, though the government is working to improve coordination.

• There is fragmentation and duplication of responsibility for donor tracking inputs vis a vis public investment – between MOP, the Council for the Development of

Page 9: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

Cambodia (CDC) and the Cabinet of Ministers – one result of which is that a significant share of donor resources flow “off-budget”.

B. Weaknesses in Recurrent Budget Formulation

• Historical/ incremental budgeting means that line agency discretion is limited, and also line agencies have little incentive to devote effort to reviewing and justifying their activities.

• Since the information base for budget decisions does not include any review of past performance, there is a limited sense of accountability on the part of spending agencies for results.

• Operations and maintenance is viewed as a residual and in some PAP ministries, there has been a divorce between O&M spending in PAP (Chapter 13) and O&M spending through the regular system (Chapter 11). Moreover, in some cases, regular O&M spending has become, in effect, “non-priority” spending, given that an overly broad definition of priority has been applied across the board in select ministries (e.g., of total MOEYS Chapter 11 spending in 2001, 70 percent was spent on sports).

• Parallel processes of negotiation between MEF and the provinces and the line agencies in some cases leads to a divergence between sector plans and actual provincial spending (see below).

• As highlighted also in the ODI study, the fragmentation of the budget formulation process, together with other, more political factors—including direct intervention from the most senior levels of government in major spending decisions—weakens the potential effectiveness of the budget as a prioritisation mechanism.

C. Weaknesses in Budget Execution so that planned priorities may not be realized.

• The cash-based payments system in operation in Cambodia is a major constraint to effective service delivery. Increasingly, budget execution has suffered from delays and an unpredictable release of funds even for salary payments, due to cash constraints. This undermines operational planning and leads to the build-up of arrears. The system is also plagued by gate-keeping and deficient accounting and reporting systems, resulting in a weak control environment and increasing opportunities for corruption.

• Limited agency discretion, in particular the inability of (most) line agency to reallocate spending across line items in response to emerging needs.

• Provincial discretion regarding execution, which may differ from agreed and approved allocations.

The MTEF There are obvious potential synergies between a well-functioning MTEF and PRSP financing and implementation (see Holmes, 2003) but the foregoing issues around annual budget-making critically affect the development of a medium term budgeting perspective. The MEF began work on introducing a medium term perspective to the budgeting process in 2002 by developing a three-year rolling expenditure framework. A draft of the MTEF for 2003-2005 was developed for the 2003 annual budget process and was

Page 10: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

included in the NPRS. The MTEF calculates an allocable ceiling (as the difference between total expenditures and the aggregate baseline ceiling) and provides indicative budget ceilings by government function as a percentage of GDP. At this stage the calculation of ceilings is largely done “top down” with little “bottom up” costing of sectoral programs except for education and health, which have developed detailed SEFs.

The MTEF should help resolve the inter-related problems of unpredictability of medium term resources and capital-recurrent imbalances, the lack of strategic orientation of budget discussions and the perceived lack of accountability of line agencies for performance. Line agencies should become aware of intended ceilings for the two to three year period ahead. Another benefit is that future maintenance costs of new infrastructure are included in the forward estimates of recurrent spending, since up-front costing of operation and maintenance is required for new facilities. This points to the need, however, to develop capacity in program/project costing at the line agency level, which is a key analytical underpinning of a medium term perspective to budgeting, working to flesh out both “top down” and “bottom up” aspects to the process. The shift to programmatic categories in education and health demonstrate that this is possible, and begin to move Cambodia’s public expenditure management toward tracking and measuring the impact of poverty-related spending.

The current MTEF does suffer from key weaknesses however as follows:

• Exclusion of the medium term wage bill, which reflects lack of coordination in medium term planning and budgeting, undermines the credibility of the MTEF, as well as the NPRS and the reform program.

• At present the MTEF excludes external financing even though this provides more than half the total resource envelope available to the Government. Extension of the MTEF to provide indicative allocations for external financing (see above on information needs), and extension to the remaining priority sectors, would help guide these resources toward priorities defined at a cross-governmental level.

• Relatedly, the next phase of MTEF development would be strengthened by building in and programing capital expenditures by sector (see sections on PIP).

Furthermore, the MTEF is presently produced as a technical paper by MEF, and has no official status. In particular there is a need to reinforce the link between the MTEF and the annual budget. The MTEF needs to make the transition from an internal, technical document to a public, politically backed plan. In this respect, the potential linkages to the NPRS process are important. The objective should be to integrate the processes as far as possible, as has been the case in Albania (see Box 2), so that subsequent updates of the MTEF are carried out in parallel with progress reporting and updates of the full NPRS.

Page 11: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

Box 2: Links between the PRSP and MTEF in Albania In Albania, joint institutional structures for coordination and management of the PRSP and the MTEF ensured close linkages and consistency of policy objectives between the PRSP and the budget, as well as consistency in decision making. The institutional structures for the preparation of the MTEF were a subset of the PRSP institutional structures. The MTEF Steering Committee is constituted of members of the Inter Ministerial Economic Policy Committee, while the NSEED Steering Committee comprises the same plus representatives of the donor community and civil society representatives. Both Committees are chaired by the Prime Minister, who guides the entire activity for the preparation and implementation of the NSSED and MTEF

Source: Kanani, 2002.

The Cambodian context is slightly more complex than Albania’s however in that there is less direct overlap between the GSCSD and the Supreme National Economic Council (SNEC). Given this, one approach might be to give responsibility for the MTEF to the Council of Economic and Financial Policies, an inter-ministerial body with representation from line ministries that coordinates high level policy making, supported by the SNEC and by MTEF-specific units in order to provide the necessary technical and policy advice, with GSCSD responsible for providing key information on NPRS monitoring and results to inform MTEF program decisions. In the context of ongoing NPRS consultations, GSCSD could invite civil society representatives to comment on the MTEF and its links to the NPRS.

This could be complemented by measures to increase the profile of the MTEF, by issuing the MTEF as a chapter in the annual budget and presenting it as an integral part of the budget speech, as well as including it in NPRS progress reports and updates. Tracking MTEF programmed expenditures as a separate table in the TOFE would also raise its profile.

On budget execution, there is forward momentum, and the NPRS emphasises public expenditure and financial management as critical to successful implementation of the poverty reduction agenda. This is also key to improvements in formulation, the credibility of the NPRS, and to the success of MTEF. It is extremely difficult to convince line ministries to take the NPRS and three year forward estimates seriously when they are concerned about whether they will receive any funding for the next three months. The Government’s diagnosis in the NPRS of the problems with existing systems reflects ongoing work in this area (drawing on the IFAPER and previous reports from technical advisers). The NPRS recognizes that slow and unpredictable budget execution resulting from the extensive use of cash, fragmentation of reserve holdings, inadequate coordination between commitment controls and poor cash management, is a major constraint for the delivery of quality public services. The NPRS states the government’s commitment to use of the banking system for an increasing share of payments and inter-governmental transfers; strengthening Treasury controls and reporting; strengthening the internal and external audit functions (including for the Priority Action Program (PAP)); and the devolution of financial controllers from MEF to line ministries.

Page 12: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

3. Pro Poor Budgeting

The broad policy priorities set out in the NPRS gave prominence to: (i) economic growth, especially through promoting export activities, tourism, and agriculture development; (ii) improving governance and public expenditure management; and (iii) infrastructure and human development, especially education and health. In this section we review how these priorities have been reflected in recent and prospective budget allocations, and also highlight concerns about the sub-sectoral breakdown of spending, the need to strengthen underlying sector strategies outside of education and health, and the continuing propensity to under-budget for critical operating expenditures. We argue that it is important to look beyond the evolution of spending that has been labeled as “priority”. To complement the review of sector allocations we also look at available evidence about provincial allocations relative to measured poverty. On the expenditure side, the NPRS recognizes the importance of spending which contributes to broad-based growth and protects vulnerable groups. It notes the desirability of evaluating both the efficiency and equity of government spending. There is a commitment to maintaining the trend, which began in 1999, of increasing budget allocations to education and health, and away from military and security. This is reflected in recent practice as well as MEF projections presented in the NPRS (see Table 2). At the same time, the fiscal path in the NPRS is constrained by prospective debt service obligations and levels of revenue collection that are low by international standards (see IFAPER Chapter 1). The NPRS assumes both improved revenue efforts and strengthened budget management to help meet both fiscal and poverty reduction objectives. Table 2: Rising Share of Priority Ministries, Falling Share of Defence and Security (as percent of current expenditures) 2000 2001 2002 2003 2004 2005 Actual Actual Budget Projected Health 8.4 9.5 10.9 12.0 12.5 12.9 Education 14.0 16.2 18.2 18.5 19.5 20.0 Defence and security 37.1 30.1 24.7 22.0 20.8 19.4 Health and education 22.4 25.7 29.1 30.5 32.0 32.9 Health, education, agriculture, rural development

25.3 30.3 32.9 33.9 35.8 37.3

Source: Ministry of Economy and Finance/NPRS. The pro-poor reorientation of expenditure thus pre-dates the NPRS. Table 4 shows that in the case of education, the annual rates of increase in budget allocation have been high and are projected to maintain this momentum. Interestingly, the Government initiated this significant reallocation to education (and health) without either an NPRS or an MTEF, facilitated by absolute increases in spending derived from revenue growth due to high GDP growth and tax policy changes. In fact, the vast majority of ministries received absolute budget increases over the period 1999-2003 (only 5 of 33 line agencies/government institutions suffered nominal decreases in their budgets). This

Page 13: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

would suggest that“tough” choices about allocations, which would necessarily entail absolute reductions in low priority ministries and programs, have not yet been made. Thus the NPRS and MTEF are timely tools indeed and should help the Government increase spending to priority sectors and programs and decrease spending on lower priorities, while marshalling development partners to rationalize their external assistance strategy and align with the NPRS. In many ways, the strategic approach to education – both in terms of the process, links to the budget, the poverty focus, and the proposed monitoring indicators – illustrates key elements that could usefully be emulated by other line agencies. The NPRS strategy in education is comprehensive, coherent and strongly pro-poor, reflecting the wealth of prior work in a recent sector strategy. The NPRS tackles important constraints in education by seeking to reduce parental formal and informal fees, providing operational budgets to schools, putting in place targeted scholarship programs, increasing the share of government budget allocated to education while allocating 70% of the budget to primary and lower secondary education, and advocating in favor of increased teacher salaries. However, as noted in the JSA, there are continuing practical concerns. First, slow and unreliable disbursements hinder school performance, and result in a continuing dependence on student fees to cover operational costs. Second, low salaries for teachers constrain service provision in rural areas.

Table 3: Government Funding for the Education Sector 2001-2005 (in percent)

2001 Actual Year -1

2002 Actual Year 0

2003 Baseline Ceiling Year 1

2004 Baseline Ceiling Year 2

2005 Baseline Ceiling Year 3

Rate of change 35.2 24.6 16.5 19.4 15.7 Share of GDP 1.71 1.99 2.13 2.33 2.47 Share of current expenditure 16.2 18.2 18.8 19.6 20.2

Source: NPRS, Chapter 5 A similarly detailed breakdown of trends in the health budget through 2005 is presented in the NPRS (Table 5.9), a sector which is also underpinned by a sectoral strategy oriented toward key goals. It is noted that much of public spending on health is donor financed (see below). More generally, outside of education and health, the policy agenda set out in the NPRS does not sufficiently prioritize the various competing measures relative to time and capacity constraints. While Chapter 5 portrays budgetary priorities in the near and medium term, there is a large discrepancy in some sectors, particularly for agriculture, between budgetary priorities and the priorities articulated in Chapter 4 of the NPRS or the action plan matrix in Annex 3. The lack of a specific pro-poor focus outside of education and health reflects several factors, which may also be reinforcing:

Page 14: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

• As noted above, the information and process used for budget allocations – historical, line item based budgeting -- does not entail a strategic nor distribution-sensitive approach.

• “Priority expenditures” defined at the broad sectoral level as is the current case in Cambodia does not, per se, appear to provide sufficient discipline on the designated priority agencies (e.g., sports spending in MOEYS), and may unduly constrain other potentially worthy activities outside the priority sectors. There is a need for a fuller development and definition of the concept of “priority” in the budget context, especially so that the NPRS can address the tough decisions that will need to be made over the medium term about reallocations away from lower priorities at the sectoral and economic classification levels.

• Fragmentation and donor dependence means that line agencies think more in terms of discrete projects rather than coherent overall programs and the need for accompanying structural and sectoral reforms;

• Lack of relevant information about sectoral performance, especially in terms of distributive effects (see below on accountability and monitoring);

• Given the absolute growth in the expenditures of most agencies over the past few years, expenditure rationalization will likely entail both consolidation of government departments and savings from administrative overhead. There are, for example, ten ministries with portfolios in economic services and six with social service portfolios. Eight of the twenty-four ministerial portfolios had budgets of less than US$ 2 million in 2001, a large proportion of which are allocated to administrative functions rather than the provision of services. Small agency and budget size constrain service delivery possibilities and mean that a high share of agency spending goes to overheads rather than on activities that could more directly benefit the poor..

This presents a situation where increased budget allocations to sectors outside of education and health may well be premature relative to the other spending agencies’ strategic focus and capacity to use the resources well. It is clearly important that activities or programs outside of education and health that are pro-poor are appropriately financed within domestic and external financing constraints, and that even for existing allocations for non-priority sectors, existing resource allocations are well used and subject to appropriate scrutiny during the formulation stage as well as accountability for performance.

A final distributive dimension of public expenditure should be considered, namely which regions spending actually benefits. Empirical analysis undertaken by MOP, the World Food Program and other partners, and presented in the NPRS, shows that most of Cambodia’s poor are rural, and that several provinces have much higher than average incidence of poverty. Hence a simple picture of whether spending is pro-poor can be gleaned from a simple comparison of spending against provincial poverty rankings.2 For the poverty estimates we use the poverty gap index which measures the average gap

2 The provincial rankings may be indicative, but are still crude given the likely differentiation within provinces. A set of poverty maps have been developed that match data from the household survey to the census and providing “rankings” down to the village level which should be useful for future analysis.

Page 15: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

below the poverty line of a poor person (so that smaller numbers indicate relatively less poverty), and also indicate the absolute numbers of poor people living in each province.

Table 4 shows that while some poor provinces – including Pailin, Krong Keb, Otdar Mean Chey and Krong Preah Sihanouk - are receiving disproportionately large allocations relative to the national average, ten out of the fourteen provinces that have poverty estimates exceeding the national average are receiving only average or below average allocations for health, education and other expenditures, while some provinces with below average poverty – like Steung Treng and Potanak Kiri – are being above average amounts.

This is not unexpected since recurrent budgeting is largely historical and based on facility and personnel numbers, so that past disadvantage would tend to be perpetuated in current allocations.3 With the current approach, areas with fewer facilities and staff get commensurately less resources, regardless of need. An exception to this general approach has arisen with the system of budget transfers for the new local councils, which are based on a formula which includes population and a measure of need4, although this allocation is limited to about 3 percent of government spending.

3 The Ministry of Health has a developed a formula-based funding method to provincial and district allocations. The formula is clear and easily understandable, and also implicitly recognizes that many of the costs associated with health care provision are fixed across sparsely populated and heavily populated areas. However the formula is not strongly related to health care needs as measured by the size and structure of the population or health status indicators. The IFAPER suggests ways to address this, either through standardized recurrent allocations to facilities; or on the basis of population weights adjusted for other need factors such as cost and health status. 4 This element of the formula is not operationalized, as is the case in many countries utilizing transfer formulae that incorporate poverty indices.

Page 16: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

Table 4: Simple Comparison of Provincial Poverty and Sectoral Spending (per capita) a/ Province Poverty Gap

Index Health spending

Education spending

Other spending

Pailin 62 13.9 6.8 20.2 Siem Reap 19 1.0 0.8 0.8 Prey Vang 16 0.9 1.0 0.8 Krong Keb 18 3.8 3.9 6.3 Kampong Chhang 13 1.1 1.0 1.1 Svay Rieng 12 1.0 1.0 0.9 Banteay Mean Chey 13 1.0 0.8 0.9 Pursat 12 1.1 1.0 1.2 Kracheh 12 1.5 1.1 1.1 Otdar Mean Chey 14 2.0 1.2 2.6 Krong Preah Sihanouk 11 1.5 1.2 1.6 Kampong Thom 8 1.0 0.9 1.0 Preah Vihar 7 1.7 1.0 1.5 Battambang 7 1.1 1.3 1.2 Mondol Kiri 6 3.2 1.9 3.7 Kampot 5 1.1 1.0 1.0 Kampong Speu 4 0.9 0.9 0.9 Kandal 5 0.9 1.0 0.8 Steung Treng 4 3.0 1.4 2.4 Takeo 4 0.8 1.1 0.9 Kampong Cham 3 0.8 0.8 0.7 Pnom Penh 3 0.5 1.2 1.1 Rotanak Kiri 3 2.4 0.9 1.8 Koh Kong 3 1.6 0.7 1.5 National Average 6.9 1.0 1.0 1.0 a/ Normalized to national averages. Sources: Poverty estimates from NPRS, Annex 2, based on MOP/WFP Database on Estimation of Poverty Rates at the Commune Level in Cambodia; provincial spending from the IFAPER.

Some other countries have addressed the need to strengthen the poverty focus of spending by establishing criteria to be followed by groups responsible for preparing budget submissions. These may be published as guidelines for Sector Working Groups (SWGs) in the preparation of the Budget Framework Paper. In Uganda, SWGs must attempt to answer these questions in their Budget Framework Paper (see Box 3). These criteria are set by the Poverty Eradication Working Group, which is located in the Ministry of Finance, and consists of representatives from several line agencies and development partners.

Page 17: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

4. Data Production, Availability and Use

The nature, quality and availability of information has an important bearing on the nature and quality of a PRS as well as on the budget process. In this section we highlight the key features on the data side that would be expected to be relevant to the NPRS, focusing on budget classifications, information on programs and costs, data provided by external partners, as well as its availability and use by key stakeholders. The nature and quality of non-fiscal information that could be utilized to strengthen accountability is addressed in the next section.

The basic budget data used for fiscal decisions and management in Cambodia is based on an economic classification or line items (see Table 5). The line item approach is used in the submissions prepared by ministries and negotiated with the Ministry of Economy and Finance (MEF).

Box 3. Using Poverty Criteria to Guide Budget Allocations in Uganda The Government of Uganda uses seven criteria to assess the poverty focus of Budget Framework Papers submitted by line agencies. These are:

1. Addressing needs of the poorest 20% 2. Addressing gender inequalities 3. Addressing geographical inequalities 4. HIV/AIDS (AIDS affects * percent of the population) 5. Environment 6. Empowerment of communities 7. Addressing intersectoral linkages

At the same time, in order to increase focus and reduce workloads for Sector Working Groups theGuidelines have been amended and simplified over time. For the financial year 2002/2003, the questionswere narrowed down to the following three:

1. Have you addressed the needs of the poor people in your sector? 2. Have you addressed geographical inequalities? 3. Have you designed measures to empower the poor people?

The key objectives of the Poverty Eradication Working Group are to review and make recommendations on (i) the overall allocation of resources and intra-resource allocations within sectors; (ii) other budget policies that have an impact on the poor; (iii) which sectors qualify as “priorities” (to be under the Poverty Action Fund); and (iv) to the Sector Working Groups in applying the PEAP crosscutting principles to the Budget Framework Papers.

Page 18: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

Table 5: Economic Budget Classifications Used in Cambodia Current Capital 10: Salaries and Allowances 11: Operating Costs 12: Subsidies to Provinces and Municipalities 13: Specific Program Agreements (including PAP) 20: Interest on loans 30: Economic Interventions (transfers) 31: Social Interventions (transfers) 32: Contributions to International Organizations 40: Miscellaneous 41: Contingencies

50.1: Construction & Equipment 50.2: Counterpart Funds 50.3: Investment by Foreign Funds

Budget data is also reported according to four broad functional categories—general administration, defense and security, social administration, and economic administration—which play a role in sectoral allocations, as well as by an administrative classification (by ministry, for example).

Moreover, in the Ministries of Health (MOH) and Education, Youth and Sport (MOEYS), a special parallel budget process — the Priority Action Program (PAP) —introduced program classifications for some expenditures (Box 4). In MOEYS, for example, there are twelve priority action programs, including “Primary Education Quality and Efficiency,” “Continuous Teacher Education,” and “Core Instruction Materials,” for which funds are allocated as part of the annual budget process and projected as part of the sector expenditure framework (SEF) that feeds into the MTEF. One benefit of the PAP has been to introduce some meaningful discussions between the respective line agencies and the MEF on sector envelopes and prioritisation.

Box 4. Cambodia’s Parallel Budget Process: The Priority Action Program (PAP)

In the late 1990s, frustration with poor budget performance led MEF, MOH, and MOEYS to look for ways of channeling budget funds to priority sectors that would by-pass the rigidities, delays and uncertainties involved in Cambodia’s over-centralized budget execution process. The resulting Priority Action Plan (PAP) aimed to channel funds directly to health and education spending units from 2000.

PAP sought to ensure that the designated priorities got their full budget allocation by: (a) removing PAP funding from the discretionary monthly cash allotment system, by providing 25 percent of PAP appropriations automatically every quarter; (b) reducing line item detail for PAP allocations, thus giving management greater discretion; (c) introducing budget management centers in spending agencies — to manage their own budgets in accordance with pre-approved plans; (d) replacing pre-audit of spending by post-audit, to be carried out by the ministries’ Provincial Departments of Administration and Finance for sub-national priority action programs on a quarterly basis; and (e) creating dedicated accounts in provincial treasuries that were not subject to discretion over cash release.

Although the impact of the reform has not met expectations, this is due more to deficiencies in the cash management system rather than inherent problems in the PAP, which has given agencies greater flexibility in the use of funds and has managed to ensure flow of funds down to the facility level. Yet it is clearly undesirable to operate multiple budget execution systems. The Government’s aim over the medium term is to integrate PAP with standard budget procedures.

Page 19: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

The main advantage of the line-item approach is its administrative simplicity both for line agencies and for the purpose of financial controls, since the classifications correspond to the administrative procedures which are needed for different types of expenditures. Hence this approach is often followed in low-income countries. However, its drawbacks as a basis for policy making are manifold, and include:

• Absence of information to enable spending allocations to match government priorities beyond a very broad level of aggregation – that is, at the sectoral level (except for PAP);

• Inability to link inputs (money) to either outputs or outcomes at the program level;

• In the event of budget cuts or expansions over the course of the year, line item categories tend to be affected across-the-board, without differentiating to protect priority programs; and

• Decision-makers (such as MEF, Cabinet and the National Assembly) and stakeholders (such as civil society and development partners) concerned about making tradeoffs in the face of very limited domestic resources do not have sufficient or appropriate information about the nature of the proposed spending.

Budget execution reports (the Treasury Operational Functional Expenditures, or TOFE) are produced monthly and aggregated annually. The results are published by MEF in hard copy (and will be made available on the MEF website in the near future). Budget data is also available on both allocations and execution by sector.

A large information gap exists however with respect to external financing5. This arises in part due to fragmentation and duplication of responsibility for tracking donor inputs vis a vis public investment – between MOP, the Council for the Development of Cambodia (CDC) and the Cabinet of Ministers – one result of which is that a significant share of donor resources flow “off-budget”. At present, the most comprehensive source of information on external assistance is the CDC database, consolidating actual and programmed disbursements of external assistance as reported by donors. While reporting is patchy and suffers from double counting the principal weakness, from a strategic planning perspective, is that the database is not compatible with the structure of the national budget, and fails to record the implementing agency for each project.

So how has budget fiscal information been used to inform the NPRS? Despite the limitations of current data, the Government has been able to use available information to good effect. Outturn information by sector is presented in a special budget chapter, covering the pre-NPRS period and also providing information on programmed (three year) commitments through the MTEF. As noted already, the NPRS uses budget data to

5 IFAPER (2003) estimates external finance to be about 130 percent of domestic spending although this figure includes technical assistance and other forms of tied aid.

Page 20: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

reveal several broad trends. These are much more useful than the type of fiscal information found in other PRSPs, including Vietnam’s CPRGS (which has a budget chapter with preliminary costing of some policy actions); and Tajikistan’s PRSP (with costing limited to “poverty reducing” spending, and no reconciliation of needs against the available resource envelope).

Moreover, several years of work supported by key partners, including ADB, DFID, WHO, and others, have led to important progress in program budgeting which finds expression in the education and health sectoral strategies and the MTEF and reflected in the NPRS. However most line agencies were unable to reliably estimate the total (capital and recurrent) costs of proposals submitted for the NPRS, with costings provided by ministries other than education and health limited to capital expenses only. This is in part due to limited technical capacity but also because line agencies had never before been asked to do, as recurrent and capital budgets were conventionally treated independently of each other. There is thus substantial scope to improve the usefulness of fiscal data for decision-making in Cambodia as part of a medium and longer term efforts to improve public expenditure management, and, linked to changes in the budget process and accountability mechanisms, to ensure that the information is actually useful and likely to be used. Some suggestions to this end are made in section 6 below. 5. Accountability and Monitoring

The NPRS places a strong emphasis on results and monitoring, and on continuing transparency and consultations, although the mechanisms to ensure accountability are not clearly defined. At the same time, under existing constitutional arrangements, the key mechanism for accountability for budget allocations lies with the legislature – the National Assembly – and in terms of execution, through a series of internal and external audit procedures and agencies, with the National Audit Authority also reporting to the legislature.6 This section reviews how the potential achievements of the NPRS in terms of social accountability and the future availability of key indicators on an annual and medium term basis might be used to strengthen accountability and monitoring in budget and government decision-making. The aim is to provide recommendations that would avert the risk that the feedback from results will not be used by decision-makers, generally, but also more specifically in the context of budget processes. The NPRS draws on medium-term goals and indicators in a range of areas from the SEDP II. These resonate with most of the areas covered by Millenium Development Goals, including the poverty headcount, and indicators to measure improvements in education, health, and access to drinking water. The government went beyond population averages, to target and track changes in several dimensions of poverty by income group (e.g., education and health disparities between the richest and poorest quintile), and by gender (e.g., health, education, employment and political representation). It is a commendable effort and certainly well conceptualized and clear relative to the PRSPs of

6 See IFAPER, Chapter 4.

Page 21: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

many other countries, which either have too few or far too many indicators, often with unclear relevance to poverty reduction and without clarifying either the institution responsible for monitoring or the frequency of collection. There nonetheless remain some challenges for NPRS monitoring in Cambodia:

• There is limited capacity and weak data at the line agency level, partly reflecting technical skills and inadequate resources. Virtually all the indicators which incorporate a distributive dimension are survey based, and hence only available every three years at best. Most of the information that is available annually is presented as averages and does not capture trends most relevant to the poor.

• In general, there is an over-reliance on survey based information collected by MOP, and no concerted effort to utilize the information being collected by line agencies and service providers. The lack of regular and timely information makes that available less relevant to the annual budget process. Exceptions to this general picture are education, some data collected by MAFF, justice, civil service and MEF.

• There is a need to work out how to utilize information that is being produced, to feed into reporting and decision-making, and in particular, align with ongoing work on poverty mapping which seems to be on a parallel track.

• Partnerships with NGOs, both local and international, need to be developed to verify survey based and administrative results, and to utilize findings from NGO monitoring work.

• Further strengthening of existing small-scale mechanisms for accountability at service provision level (e.g. health management committees, water users’ committees, school community involvement in PAP financial management) could help and, in the future, could become part of the relevant sector’s regulatory framework.

Further development of the NPRS needs to ensure effective integration with line agencies’ information systems, alongside adequate strengthening of those systems. As noted above, under the institutional approach that has evolved in Cambodia over the past year, the GSCSD is responsible for monitoring and consultations, while the Supreme National Economic Council has responsibility for ensuring policy decisions and financing consistent with the NPRS. This demarcation is consistent with the comparative advantages of the institutions, given their respective associations with MOP and MEF, and could work well. However there is a need to clarify how the results obtained from the ongoing NPRS monitoring can be utilized in the annual and medium term budgeting processes. In order to enable this, several things need to happen:

• The timing of NPRS monitoring – or at least the time of the year when results become available – must be compatible with the budget cycle. Table 2 above indicates that the results would have to be available by March in order to inform the setting of sectoral ceilings, and by May/June to be used in intra-sectoral determinations.

Page 22: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

• Relatedly, there is a need to specify which types of results could (or must) be submitted by spending agencies as part of their budget submission. For example, updated information on access or utilization drawn from administrative data and ideally subject to some kind of scrutiny or verification by interested civil society observers. Under current arrangements, for most agencies, this would complement a line-item based proposal.

• There is a need to further open up the budget decision-making process in order to capture some of the gains of the NPRS, while ensuring continued manageability and recognizing the varying technical capacity of potential stakeholders. The most obvious scope currently exists vis a vis the National Assembly, for example through the budget committee becoming more extensively involved in hearings, reviews of past performance and implementation problems.

Cambodia has the framework of a democratic system in place, with active political parties, high electoral turnouts and regular elections since 1998. Legislative powers at the national level are vested in two institutions, the National Assembly and the Senate (the latter is relatively new, its 60 members are appointed and its constitutional powers are limited). There are 122 Assembly members, who are full time.7 Efforts to develop accountability need to build on the nascent democratic institutions, while recognizing that civil society groups do have an important complementary role to play. Although dominated by the two coalition parties these institutions have a sizeable number of opposition party members who use these platforms to regularly challenge the government and withhold approval of critical laws until agreement is reached between the executive and the legislative. TV broadcasts of Assembly debates have helped to open up the legislative process. The Legislature does review the annual Budget Law and is responsible for legislation which serve as policy platforms (e.g. investment, land, forestry). Evidence suggests that the role of the parliament in the budget process is quite weak and that the National Assembly and Senate are rarely given adequate time to review and scrutinize the proposed budget. Usually the National Assembly receives the budget in December and the Government indicates that this is an "Urgent Bill" which means that the Parliament has less time to review and approve it (in the case of the Assembly it means a fortnight and for the Senate, less than five days). Some MPs and Senators have voiced concern about the lack of time and preparation for Parliament to review complex financial bills and particularly the budget law, and that the current process makes it impossible to scrutinize effectively the budget or question the Government on its content. In both the Assembly and Senate there is, a Commission on Banking and Finance which is responsible for reviewing all financial and fiscal legislation including the budget. Again, commission members have complained that they are not effectively engaged in the budget process because the budget is always marked as “urgent” so there is no time for the Commissions to review and comment on the draft.

7 See http://www.cambodian-parliament.org/english/index.htm; http://khmersenate.org/english.htm

Page 23: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

There is even less scrutiny of results than of the proposed budget law, except for the annual Budget Settlement Law – which is now to be accompanied by the independent opinion of the National Audit Authority. While parliamentary involvement in much of the NPRS process was limited, there are recent signs that this could be changing (see Box 5). Box 5: Prospects for Parliament and the NPRS While the level of commitment of Cambodian parliamentarians in the NPRS process was limited for much of the period of development, this appears to be changing. A WBI and Canadian Parliamentary Centre workshop for parliamentarians on PRSPs in November 2002 provided some insights into both the potential role of the legislature and the constraints it faces.

In preparation for the workshop, the Chair of the Senate Finance and Banking Commission took his entire commission and commission staff on a field investigation to the three poorest provinces, apparently for the first time as a team or commission. Subsequent to the PRSP workshop, the Senate Finance and Banking Commission requested assistance to develop a research unit to assist parliamentary commissions with their policy analysis and consultation work to monitor and report on the NPRS and keep the Government accountable for policies to alleviate poverty. The Commission have apparently also indicated they would like to continue parliamentary-civil society relations and exchange that would reinforce the NPRS emphasis on participation. Parliamentarians also identified key training priorities they feel they need to pursue to become more effective in anti-poverty work in the context of the national budget. Source: Activity Report to World Bank Institute (WBI) for “Parliament, Governance and PRSP” Workshop for Selected Commissions of the Parliament of Cambodia, submitted by Parliamentary Centre (Canada), December 10, 2002 Opportunities for civil society to participate in the budget process are limited, as is their capacity to do so. It is not clear in the short term how the budget process itself could be effectively opened up to national civil society, given the type of information used in budget decision-making, and capacity constraints on the part of both civil society to engage and MEF to be able to manage such a process. On the other hand, the voices of civil society expressed through NPRS monitoring, and more directly through the media, can be influential. One constraint vis a vis the press and the web sites is its urban bias and limited reach. The education ministry is probably at the forefront of budget transparency, with budget information being widely available to all managers in the MoEYS and the donor community (although less so at the district and school levels)8. Consideration may be given to emulating this approach in other line agencies. Another important factor in Cambodia is the external accountability mechanisms that have evolved over time through Consultative Group meetings. “CG benchmarks” have been developed, especially in areas and for measures of special concern to major partners. There are four working groups, and each has developed monitoring benchmarks agreed upon with their government counterparts and at the last two CG meetings there has been agreement on priority areas for joint monitoring. The areas to be covered are largely limited to policy actions, as opposed to the actual outputs or outcomes that are

8 ODI study, 2003.

Page 24: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

also being tracked under the NPRS. Specific numerical targets have been set for disbursements (budget execution) in the priority sectors. The areas are summarized in Box 6. Box 6. 2002 Consultative Group Meeting—Joint Priority Monitoring Areas A. Legal and Judicial Reform and Governance -Key components for enacting legal and judicial reform in place. -Visible actions taken to fight corruption. -Coverage of implementation of Procurement Sub-decree (1995) extended. B. Natural Resources Management -Forest concession system rationalized on basis of Sustainable Forestry Management Plans. -Legal framework for community-based natural resources management strengthened. C. Social Sectors -Budgetary disbursements to social sectors improved. -Declared actions to support HIV/AIDs prevention implemented. D. Fiscal Management and Public Administration Reform -Civil service reform plan updated consistent with MTEF -Revenue mobilization strengthened through broadening collection of taxes and fees.

However the role of the CG benchmarking is unclear. In particular, the Government has routinely not met targets but this appears of little consequence as targets are then rolled over to the next meeting. However, this may change over time if unsatisfactory patterns develop. Additionally, it may not be ideal to separate domestic and external accountability mechanisms, particularly in the face of low national capacity and in the context of a poverty reduction strategy which has been developed nationally. Consequently, the Government and partners should work towards an agreed set of indicators agree that are consistent with, and can be used to monitor the NPRS, and avoid having additional “external” indicators with a separate monitoring process divorced from the NPRS. Streamlining NPRS and CG monitoring is a shared responsibility between both the Government and the donor agencies and needs to take account of processes which may be established at a later stage such as the “localizing” of the MDGs. It is notable that health and education have proposed innovative mechanisms for joint monitoring of progress of their sectoral result frameworks including annual joint reviews of sector performance reports and of revised/rolled forward program and financing frameworks linked to the MTEF (see Box 6). There are plans for this to be informed by provincial performance reports in the near future.

Page 25: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

Box 6: Encouraging Experience with Education and Results Monitoring

In the education sector the requirements arising from the joint MOEYS/donor monitoring process have shifted incentives towards improved information management, in a direction which facilitates result-based management. At central level especially, there is a growing demand for information and an increasing awareness of the effectiveness of good information and analysis for enhancing policy making and program implementation. The Ministry has started developing a capacity to design and implement targeted surveys to help monitor the links between program outputs and strategic outcomes. It has also started linking education information with available cross-sectoral data to monitor the extent to which the ESSP is really pro-poor. Appropriate support and carefully designed plans for further decentralisation can help further to change incentive frameworks at sub-national levels and to address critical needs for improved management capacities at these levels. Source: ODI Study, 2003 6. Donor Financing and External Alignment A key goal of the PRSP approach is to improve the alignment of external financing with national priorities and to reduce the transaction costs of accessing assistance. The benefits of alignment in countries like Cambodia, where external financing exceeds domestically mobilized public resources by around 30 percent9, are obvious. This is clearly an ambitious aim although tangible progress toward these goals in countries such as Uganda, Tanzania and Ethiopia provides ground for optimism. Whether external assistance becomes aligned with national priorities and the domestic budget in a particular country case will depend on a range of factors, some country specific, others generalisable. In this section we review the extent to which current flows might said to be aligned, whether the pre-conditions for better alignment are in place, or at least in train, and possible ways to accelerate progress.

As shown by the IFAPER, the composition of external flows has shifted gradually over time toward the social sectors, and away from core government functions and economic development. Favoured activities appear to be community and social services, and water and roads (see IFAPER Table 2.16). Total spending in several key sectors – including health, agriculture, transport and, especially, rural development - is dominated by external flows. Caution is, however, needed in interpreting these figures, not least because the external assistance itself is dominated by technical assistance – indeed spending on foreign experts is triple the civil administration’s wage bill. The share of technical assistance is especially large in the priority sectors of agriculture and area/rural development (amounting to half the assistance provided).10 Recent analysis suggests a progressive geographical distribution of external financing in the case of health, for example, but no further conclusions could be drawn since information was only available for a minority (25%) of spending (IFAPER, 2003).

9 Over the period 1996-2001, reported external flows amounted to $2,122 compared to $1,852 worth of domestically mobilized resources (IFAPER, Chapter 2). 10 Interestingly the technical assistance share is relatively low – 20% -- in education despite the fact that this sector is a leader in policy and program development and strategic and innovative thinking.

Page 26: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

Sectoral diagnoses reveal clear shortcomings in the extent and quality of service provision – with respect to health care, for example, only just over half the population (54%) live within ten kilometers of basic public health services; only 14 % of rural women in the bottom quintile had a medically trained person attending birth of a child, and almost two thirds (63%) of MPWT’s road network is in poor or bad condition11. In such a situation, arguments to shift external financing to enable these shortcomings to be addressed would seem persuasive. Indeed, while knowledge transfers and capacity building is critical, it seems appropriate, given the need for investments in facilities and maintenance, to direct external financing better to meet those priorities. Moreover, where sound sectoral policies are in place, and as improvements in public expenditure management are implemented, a shift toward more general budget support might be expected for those partners – like the EC, the World Bank, the ADB and DFID – whose management is supportive of this type of assistance. In mid 2002 a Government-Donor Partnership Working Group was established under the CG mechanism, chaired by the Government, and there is on-going work on harmonization/joint portfolio management led by MEF with ADB and WB and more recently Japan. However, better alignment of assistance to Cambodia’s priorities, including through budget support, demands certain preconditions. While individual donor preferences obviously vary, recent international experience allows some generalizations about the aspects of country context which facilitate alignment.12 These include (i) strong country ownership of a national poverty reduction strategy with (ii) clearly articulated structural and sectoral policies and programs and a sustainable macroeconomic framework (iii) evidence of momentum toward implementation; (iv) adequate diagnosis of public expenditure management, and concurrence of government and stakeholders in the key conclusions, and agreement on critical steps; and finally but not least, (v) confidence in the governance arrangements, and evidence of steps to redress diagnosed problems and to improve transparency and accountability.13 Each of these broad dimensions has more specific applications at the sectoral and sub-sectoral level – for example in the case of roads, donor support for operations and maintenance might be contingent on a more transparent approach to procurement and tendering of such activities. Cambodia has made recent progress in some important key respects that should facilitate alignment. In particular, the recently completed IFAPER (joint WB/ADB with IMF and DFID participation) provides a matrix of PEFM reform priorities, all of which have been incorporated into the Prime Minister's draft Economic Action Plan, 2003-2008, demonstrating the emergence of a Government-donor consensus around both the problems and solutions. On the other hand, while there is broader acknowledgement and growing consensus about the ways to address broader governance and civil service

11 IFAPER, Chapter 3 contains extensive sectoral diagnoses linked to evidence on outcomes. 12 See World Bank/ IMF (2003) PRSPs – Progress in Implementation – Detailed Analysis, available at http://www.worldbank.org/poverty/strategies/ 13 See, for example the donor survey commissioned as part of the work of the OECD-DAC Task Force on Donor-Practices. See OECD-DAC, “Harmonizing Donor Practices for Effective Aid Delivery-Good Practices Papers,” DCD/DAC/TFDP (2002)12, 2002, OECD, Paris, France

Page 27: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

reform problems, the pace and scope of reform will likely have to quicken in order to facilitate donor alignment and support. Recent discussions and official papers14 suggest that while the Government concurs with the focus of African countries on national poverty reduction strategies as the framework for harmonization at the country level, there is a sense that caution is required in moving toward a focus almost exclusively on a budget support modality linked to NPRS given the risk of delays in aid delivery. The Government’s approach for the present time is to continue to combine projects, programmatic approaches and NPRS-based budget support. As highlighted by Senior Minister Keat Chhon at a recent high level forum on harmonization,15 the Government believes that the NPRS provides the strategic framework for reducing poverty, addressing both sectoral and cross-sectoral issues, and that the challenge now is to clearly define entry points for programming and financing the agreed action plan. Based on international experience, the Government is looking to well-defined external assistance programs that fully incorporate the action plan of the NPRS as the most effective entry point. 7. Emerging Lessons and Recommendations

The potential NPRS – budget links in Cambodia could be substantially strengthened in ways that would be mutually beneficial to the quality and nature of the NPRS and the budget processes. Each has elements which can be used to open up and discipline the other – the budget perspective brings the link to fiscal and absorptive constraints, the need for sequencing and the prospects of rigorous application of technical criteria. The NPRS is a tool which enjoys substantial line agency buy-in, reflects a process which was unprecedented in its openness and transparency, has a strong focus on results, and can be used to align the substantial external resources which Cambodia receives to national programs to a much more greater extent than in the past. However for successful implementation of the NPRS, MEF’s full engagement is critical. The proposed arrangements, whereby policy development and financing of the NPRS falls to SNEC, and monitoring and consultations to GSCSD is a very promising avenue to address a number of the key weaknesses that have been identified in this paper. Recommendations that aim to increase the likelihood that this potential be fully exploited include:16

• Timetables need to be aligned. Results from NPRS monitoring need to become available at a time which is relevant to the annual budget process (for both inter and intra- sectoral allocations) and medium term adjustments. This could begin in 2004. Over time the progress reports on NPRS implementation should be timed so as to inform the budget process and could become background papers to

14 See Cambodia’s Approach to Tackling the Harmonisation Issues, Prepared for the High Level Forum in Rom, Italy, February 24-25, 2003 by Council for the Development of Cambodia (CDC) 15 Mr Keat Chhon, Senior Minister, Minister of Economy and Finance, at High Level Forum, Rome February 24, 2003. 16 A comprehensive agenda to improve public expenditure management was agreed with government in context of recent IFAPER. Details are not related here.

Page 28: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

national budget documents. For this to happen, MOP and MEF would need to work closely together and integration of the NPRS process with the annual budget process would need to have official support from a body such as the inter-ministerial Committee for Economic and Financial Policies, a high level policy making body.

• Coordination with the MTEF and PIP processes would also have to be ensured. In order to integrate the NPRS with the budget, the PIP, as the capital budget component, will also have to be integrated with the MTEF. The time is right for redefining the role of the PIP and making it an effective public management tool in the context of the MTEF.

• Prioritization of spending, based on the NPRS and SEDP II, will need to be carried out as part of the MTEF, PIP, and annual budget processes. Confining the task of prioritization to the PIP risks undermining the poverty reduction agenda.

• Institutionalize the MTEF. Development of the MTEF will have to be integrated into the MEF’s institutional framework and sector budget/planning cycle routines. The policy orientation of forward expenditure planning requires a range of new skills—in policy analysis, costing and programming. Experience from Africa suggest that MTEF’s may be more effective where existing departments are complemented by new units responsible for both high level coordination and specialized technical analysis in order to provide the support necessary for implementation.17 Across government, the MTEF support structure could be integrated with existing NPRS units and committees. Development of sector expenditure frameworks in the non-priority sectors is likely to be a lengthy process. SEF’s will have to be built on agreed sector policy frameworks and development strategies, and aligned with well-defined and costed programs.

• While broad policy directions like the shift from defence and security spending to the social sectors are internalised18, there is a need for transparent, technical criteria for the prioritization of major spending proposals across sectors (which could be complemented by sector specific guidelines) for use in the annual and medium term budgeting processes (as in Uganda). Since the extent to which effective prioritization can take place is severely circumscribed by the line item based budgeting, the use of such criteria would initially be applied only to those agencies which have adopted some program budgeting – viz. education and health – and hopefully extending to agriculture19, rural development and possibly roads in the near term.

a. Fuller development and definition of the concept of priority spending is needed. This should be informed by the messages emerging from consultations on the NPRS, and should include agreement on the criteria for assessing whether a particular program should be regarded as a priority. This work should be led by MEF but would involve substantive

17 Le Houerou and Taliercio (2002). 18 Though salaries in the defense and security sector increased in 2002, along with those of the civil administration. 19 As MAFF is organized by programmatic departments (e.g., forestry, fisheries, extension, etc.), some progress could be made here in the short term as well.

Page 29: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

line agency engagement as well as feedback from civil society, development partners, and possibly parliamentary review and approval.

b. Once the concept of priority has been more fully defined, proceed to re-examine existing programs in terms of consistency between identified priorities for poverty reduction and resource allocations, including in the rural and agriculture sectors (This should be complemented by other kinds of analysis, including cost-effectiveness analysis).

c. The priority criteria could also be used to guide capital spending – this would increase the transparency and accountability of public resource allocations.

d. Use the results of the poverty mapping work to check proposed allocations and executed expenditures against provincial and district measures of deprivation.

• Expansion of programmatic budgeting will be critical to strengthening the

strategic and pro-poor focus of resource allocations. MEF’s plans to assist other priority sectors – beyond education and health -- in developing their SEFs in 2003, most likely the other “priority” sectors: agriculture (MAFF) and rural development (MRD). An additional candidate for consideration should be the MPWT, given the importance of the road program to growth and poverty reduction.

• It will also be important to improve budget execution at both the central and

provincial levels. In addition to improving cash management and reforming Treasury operations, this would likely entail on sector earmarking at the provincial level, on an interim basis, which would also allow line agencies to improve targeting of spending, by applying allocational formulae based on indicators of need and/or performance.

• While the broad concept of priority has been used to guide broad sectoral

allocations, at the execution stage, when shortfalls arise, prioritization takes place on a line item basis(e.g., operations and maintenance spending becomes a residual). This means that priority programs may be deprived of critical operating resources. MEF could investigate and clarify how priority expenditures will be protected during execution in the face of revenue shortfalls.

• Rationalization of spending will also have to take into account civil service

reform issues. A functional analysis planned to begin in late 2003 will provide useful inputs for reconfiguring the public sector to focus better on its core objectives. (A longer term issue is the need to consolidate agencies with similar responsibilities, with clearer demarcations needed in the short term --for example, the relations between the Ministries of Public Works and Transport, and Rural Development, on rural roads.

• Compensated retrenchment will likely form part of any comprehensive

expenditure rationalization program. Moreover, pay reform will have to assess

Page 30: CAMBODIA: PRSP AND BUDGET LINKAGES - World Banksiteresources.worldbank.org/INTPRS1/Resources/Attacking-Poverty... · CAMBODIA: PRSP AND BUDGET LINKAGES ... more directly to the allocation

the tradeoff between increasing salaries, which are very low and compressed, and lower priority spending.

• The quality of reporting and programming of external financing commitments

needs to be improved. If indicative allocations are to be useful for planning purposes, this would involve:

a. Partners providing upstream information in a form which enables the authorities to record project and program financing against the responsible government agency.

b. Closer coordination between the Council on Development Cooperation and MEF on external resource management.

• There is scope for opening up and clarifying the budget formulation process –

both in terms of making the information base for MEF decisions available ((i) above), and more formally taking into account feedback from users and service providers about performance and problems. The NPRS monitoring results should be used to inform allocation decisions. In particular:

a. MEF could ask line agencies to include key performance data and qualitative feedback as part of their budget submission;

b. Budget hearings could be opened up beyond bilateral (MEF-line ministry) discussions, to include key stakeholders outside of government. Alternatively, or in addition, the minutes of these meetings could be published (eg on the MEF website).

c. The involvement of parliament in the budget process could be strengthened. Given current constraints, a possible way forward is to provide and ensure additional lead time to parliament in reviewing the laws, as well as an upstream review. This could enable the Finance and Banking Commission to conduct their own pre-budget consultations and analyze government policies in order to prepare for the budgetary review. Training and skills upgrading, as well as staff with technical expertise, is needed if parliament is to fulfill its potential oversight role. (Canadian technical assistance is being provided in these areas.)

• There is need to build capacity within line agencies to undertake relevant

analytical work, interpret monitoring results and estimate costs. This is needed to circumvent the “vicious circle” whereby sectors like roads and infrastructure do not receive adequate attention (as in for example the initial phase of the MTEF), in part due to capacity problems in the line agencies. Having clear criteria related to poverty reduction and sectoral objectives to evaluate line agencies’ annual budget submissions (see (2) above) should assist in this.

More generally, in order for the budget formulation process to become more meaningful, there are further well-recognised areas for improvement, including the need for a more predictable flow of funds, so as to allow the various institutions to plan and schedule activities, and to ensure that operating and maintenance requirements are estimated and allocated together with capital funding decisions.