Brieng Paper

June 2005

LINKING POLICIES AND BUDGETS: IMPLEMENTING MEDIUM

TERM EXPENDITURE FRAMEWORKS IN A PRSP CONTEXT

MTEFs and PRSPs: The Issues

Strong linkages between policy, planning and budgeting are

necessary for developing country governments to use limited
resources efciently and effectively.A medium term perspective
to budgeting can be useful in this process and specific
instruments such as MTEFs are important in improving budget
management in developing countries. The appeal of MTEFs
lies in their potential to link the often competing short-term
imperatives of macroeconomic stabilisation with medium and
longer term demands on budget resources, thereby contributing
to improved policy making and planning and more efcient
and effective service delivery.
Box 1: Characteristics of MTEFs
Key technical characteristics of a fully functioning MTEF
can be dened. These include a medium term scal
framework, estimates of the future costs of existing
policies, and sector strategies setting out sector and
sub-sector priorities for future spending. While these
elements are unlikely to all be in place before an MTEF
is introduced, making progress towards them is a critical
part of early MTEF development.
An MTEF is an integral part of the annual budget cycle
and typically consists of:
i) A top-down resource envelope consistent with
macroeconomic stability and broad policy priorities;
ii) A bottom-up estimate of the current and medium
term cost of existing national programmes and
activities;
iii) An iterative process of decision-making, matching
costs and new policy ideas with available resources
over a rolling 3-5 year period.

The characteristics listed in Box 1 point to obvious linkages

between MTEFs and PRSPs (see Figure 1 overleaf). A PRSP
lays out in detail the poverty reduction priorities of a country,
identifying specic policies and strategies, and ways to monitor
progress in each sector. Therefore, while PRSPs identify the
long-term objectives and priorities for poverty reduction,

the MTEF provides the framework for allocating resources,

within which explicit policy choices and trade-offs are made
depending on resource availability.
Given the increasing importance placed on the development
of an MTEF to support PRSP implementation, it is useful to
assess progress made so far in the development and use of MTEFs
in different countries, and to investigate the nature of linkages
with PRSPs. In general, the case studies on which this brieng
paper is based suggest that MTEF implementation appears to
be a signicant driver of wider reform. At a broad level, and
based on a qualitative assessment, the countries can be divided
in terms of the direction of the development of their MTEFs
in the following categories:
i) Maturing (South Africa, Uganda)
ii) Getting it together (Albania, Benin, Rwanda, Tanzania)
iii) Struggling (Burkina Faso, Cameroon, Ghana)2

Key Findings from the Case Studies

Initial Conditions

The general message from the case studies is that initial

conditions do matter when introducing an MTEF, but
that the MTEF itself can be a valuable catalyst for building
basic budget conditions (e.g. improved scal stability and
expenditure control).The Tanzanian case in particular illustrates
how initial conditions can help explain progress with the
MTEF. Strong leadership from the Ministry of Finance,
progress in macroeconomic and scal stabilisation, concurrent
development of a basic but integrated nancial management
system and strong stakeholder involvement helped the process
of MTEF implementation and the MTEF itself stimulated
further improvements in budget basics.
Similarly in South Africa and Uganda, the introduction of an
MTEF promoted the development of budget basics and also
increased internal demands for greater budget comprehensiveness
(i.e. capturing a greater amount of resources on-budget).
In the absence of favourable initial conditions, some countries
have tried a partial introduction of an MTEF, covering only a
sub-set of sectors. On balance, the evidence from the case studies
suggests that to be effective, the MTEF needs to cover the whole
of government, while recognising that sectors will develop at
different speeds. A key point is that the weaker the basics the less
onerous MTEF design should be.

Integration

Integration of the various phases of the budget cycle, including

the MTEF, and with other key policy and planning processes,
including the PRSP, is central. Evidence from the case studies
suggests that an integrative MTEF process is the single most
signicant factor distinguishing the maturing case countries
from the others. In an integrative process, there is a seamless
progression from the setting of the aggregate resource envelope
to the allocation of resources and then to the detailed annual
budget estimates submitted to Parliament. Such integration

ODI Brieng Paper June 2005

A common criticism of Poverty Reduction Strategy Papers

(PRSPs) is that they represent a wish-list of policy measures
which are poorly prioritised and too often de-linked from
availability of nancial resources in the budget. Medium Term
Expenditure Frameworks (MTEFs) have been hailed as the
solution to this problem, and introduced with donor support
in a number of PRSP countries. However, the evidence so far
is mixed, and highlights a number of factors which need to
be taken into account to effectively link policy-making and
budget processes. This brieng paper is based on nine country
case studies which investigated the experience of implementing
MTEFs in a PRSP context.The countries were Albania, Benin,
Burkina Faso, Cameroon, Ghana, Rwanda, South Africa,
Tanzania and Uganda.1

Figure 1: MTEF-PRSP annual cycle and links with donor support

Donor
Activities

PRS cycle

Ministry
of Finance

PRGF Mission
Indicative donor
financial commitments

Participation

Sector
Reviews

Expenditure
Reviews

PRS
Report

Macro Framework &

preliminary ceilings
prepared

PRGF Mission
Confirmation of
donor commitments

National PRS
Annual Review

Macro Framework revised,

Budget Framework Paper
prepared & final ceilings
approved by Cabinet

Submission of
Budget to
National
Assembly

Debate &
approval by
National
Assembly

Sector
Consultations
Ministries/
Provinces/
Agencies

Finalise MTEFs
within approval
ceilings

Strategy & Issues

Papers
(+preliminary
MTEFs) prepared

can ensure that the various actors at each stage of the process
have an understanding of how their actions affect the process,
and can use this understanding to modify their behaviour in
order to support the common goal of a more performanceoriented public sector.
A key characteristic of the implementation of the MTEF in the
maturing countries (and in Albania) is a clear budget timetable
that allows sufcient time for a strategic planning phase and for
the preparation of detailed budget estimates. This reects the
efforts of these countries to achieve integration, but also their
recognition of the importance of a predictable budget process.
Evidence from other country cases (e.g. Burkina Faso, Cameroon,
Ghana) is that frequent changes to the budget timetable can create
uncertainty and undermine credibility of the process.
On another level, integrated institutional arrangements are also
conducive to effective budget management. Separate budget and
planning agencies can undermine the comprehensive approach
to budget making. However, where there is no immediate
prospect of creating a single budget agency, clear institutional
arrangements are necessary to ensure a joined-up planning and
budgeting process. Linkages to the PRSP process are also aided
by ensuring that the PRSP process is coordinated by the same
budget agency.

overoptimistic assumptions about revenue etc.), for most of the

study countries donor funding is even more volatile. Donors
need to place a much higher value on providing predictable
funding ows, particularly in the case where donor assistance takes
the form of budget support. As a response to this volatility, for
example, Uganda adopts discounted forecasts of donor inows
to ensure greater predictability in the budget process.

Priority Setting & Resource Allocation

The case study analyses suggest that the MTEF has its greatest
impact at the stage of priority setting and resource allocation.
This is also the most obvious point of intersection between
the MTEF and the PRSP.
Through the provision of a realistic, hard budget constraint,
and effective political engagement (see below), the MTEF/budget
process legitimises policy choices, leading to greater credibility of
resource ceilings, budget allocations and predictability of funding.
Translating the aggregate resource constraint into budgetary
allocations is increasingly happening via top-down ceilings.
Evidence from Albania, South Africa and Uganda indicate that
these ceilings increasingly reect sector priorities as well as the
aggregate constraint. PRSPs are clearly providing an impetus to
injecting priorities into the determination of top-down ceilings.
Signicant progress is also being made in separating collective
(Cabinet) decisions from those appropriately left to individual
ministers.

Fiscal Stability & Sustainability

ODI Brieng Paper June 2005

Source: SPA Action Learning Mission

Report Rwanda 2002

Early attention to scal stability is of fundamental importance to

the implementation of an MTEF.Aside from the more obvious
reasons linked to macroeconomic stability, scal stability plays a
critical role for the MTEF and the budget process, improving
the predictability of funding and its subsequent positive
implications for rational and efcient budget implementation.
Evidence from the case studies indicates that honesty and
realism in setting the aggregate resource constraint are key
factors for achieving predictability.
Aggregate estimates are prepared in the rst stage of the budget
cycle and it is at this point where greater honesty and realism can
lead to improved predictability in expenditure.While problems in
dening a realistic aggregate resource constraint can stem from a
number of domestic factors (e.g. incentives to overstate revenue;

Political Engagement

Early engagement with the political decision making process

is crucial. The case studies show that where politicians are not
involved in the early stages of budget formulation, an MTEF is
unlikely to be effective. Engagement at the political level, with
Cabinet involvement in the early stages of budget formulation, is
the major factor that has begun to give substance to the MTEF
process in various case studies, in particular to the credibility
of the expenditure ceilings. If politicians are unwilling to be
constrained either by a hard budget constraint or by agreed
priorities, then the whole budget process will be undermined,
leading to poor budget implementation. This also applies to

Strategic Management at Sector Level

PRSPs, where strong political engagement in key decisionmaking stages can help move both the MTEF and PRSP
forward (e.g. Uganda,Albania). See Box 2 for further discussion
on political involvement.

In addition to strategic planning of the overall budget, a

more strategic approach to management at the sector level
is very important for enhancing the credibility of budgets
and underpinning improved budget execution. For example,
the Albanian Ministry of Finance issues guidelines for the
preparation of sector expenditure strategies as an integral part
of the MTEF/budget process. These guidelines recommend
setting indicative resource guidelines and showing how existing
resources are allocated across key programme areas within the
sector, summarising overall policy objectives of the sector, and
providing suggestions for how allocations should change over
the coming three-year period.
Similar strategic approaches at the sector level are being
implemented in a number of the other case countries (Ghana,
Rwanda, South Africa,Tanzania, Uganda), with the help of sector
working groups or in the case of Rwanda and Tanzania, through
the annual Public Expenditure Review process. A key message
from the case country experiences is that the sector strategic
management approach should be fundamentally concerned
with ensuring that ministers, ministries and agencies challenge
existing policy and themselves to ensure the best use of whatever
resources they receive.

Box 2: Budgeting and Political Realities

The MTEF concept is underpinned by certain values, such
as the values of scal discipline and improving budgetary
outcomes. It is acknowledged that politicians are often
driven by concerns, including re-election, that are not
necessarily in line with better budgetary outcomes. It is
quite possible that the rational, rule-based approach of
the MTEF is quite antithetical to the way in which politics
operates in a country; and without political backing,
technocratic efforts to construct an MTEF can be left
high and dry.
But there are also many instances where it is evident
to the political elite, as well as to key technocrats, that if
they want to deliver effectively on their political choices
and build the condence of their citizens, they need a
better budget process to do it. Viewed from this angle,
the MTEF (and budget process more generally) becomes
not merely a technical instrument for consolidating and
optimising the use of resources but a potentially crucial
instrument for building and reinforcing political legitimacy
and hence domestic accountability.

Budget Execution & Performance Focus3

One of the ultimate objectives of an MTEF is to improve

the quality of budget execution. A key message is that by
contributing to greater predictability of policy and funding, an
MTEF can provide the basis for quality budget implementation.
Both the MTEF and the strategic sector processes can provide
platforms for the introduction and use of performance measures
to support expenditure allocations. For example, Tanzania
and Uganda require sectors, as part of the budget process,
to describe past performance and state on a rolling basis
the strategies, objectives and targets they intend to achieve.
However, developing and sequencing a performance focus
to the budget is not an easy task, and evidence from the case
studies indicates that it this is difcult to achieve in practice.
Moreover, countries tend to be better at setting performance
targets in their budget strategies than establishing systems for
monitoring and measuring performance, and holding those
with responsibility to account.

Strategic Phase to Budget Management

A key lesson from the case studies is the need to create space at
the outset of the preparation of the budget for a strategy phase.
A tight link between this phase and the preparation of the
detailed budget estimates is evident in South Africa. In Uganda
and Albania, there are two distinct phases to the budget process.
The rst focuses on strategy and culminates in the decisions
on inter-sectoral budget ceilings while the second involves
preparation of detailed estimates within the agreed ceilings, and
culminates in the reading of the budget. One important message
is that detailed costings at the strategic phase can distract from
a focus on priorities and strategies.
A common development has been the introduction of some
form of strategic document to help focus discussion and identify
options available to decision makers. For example, in South Africa
there is the Budget Policy Statement, in Uganda the Budget
Framework Paper, in Albania the MTEF document and in Burkina
Faso the Circulaire Budgetaire. The value of these documents
is that they provide a common framework around which
stakeholders can engage in a debate about priorities, options and
choices. Such a document has great potential to become the key
mechanism for integrating the MTEF and the PRSP.
Measures need to be in place for dealing with existing and new
policies. A number of the cases show that more explicit processes
for dealing with new policies are being built into the systems, and
that the PRSP has the potential to focus attention on this issue.
Sector working groups can also be encouraged to develop or
consider new policy initiatives and this can help cement the links
between policy, planning and budgeting. In contrast, approaches to
assess existing policies are less-focused. Indeed it is arguable that
one of the great gaps in work on public expenditure management
over the last ten years has been the failure to focus attention on
the relevance and effectiveness of ongoing policies. The result is
that new spending is treated as additional, and spread thinly over
too many programmes and activities, rather than prioritised on
the basis of an assessment of the effectiveness of existing policies
and of their impact on poverty reduction. This highlights a clear
gap in MTEF/PRSP integration.

One concern about MTEFs is that they are often perceived as

a centralising force, even in contexts where decentralisation is
ofcial policy.As already noted, the case studies have shown how
the MTEF is associated with a shift to a more strategic approach
to decision making and, associated with this, a clearer distinction
between collective decisions at the centre and individual sector
decisions. A decentralised system of government increases the
complexity of resource allocation decisions, but an MTEF can
help clarify expenditure assignments and therefore support
decentralisation efforts. While every solution to this problem
will be country specic, more attention needs to be paid
to matching authority and responsibility between levels of
government.

MTEFs and PRSPs: The Linkages

The key point of intersection between an MTEF and the

PRSP, as already mentioned, is at the stage of priority setting
and resource allocation.The country studies suggest that where
the MTEF has focused on providing a disciplined framework
for decision making about strategic resource allocation, it has
forced attention onto policy priorities. With the explicit focus

ODI Brieng Paper June 2005

MTEFs and Decentralisation

on poverty reduction generated by the PRSPs, this appears

to be putting some backbone into recent and/or faltering
MTEFs.4
However, it is also clear from the case studies that there are
risks of focusing too much on PRSP priority programmes and
expenditures, leading to unbalanced allocations. For example, in
Tanzania, non-wage recurrent expenditures are prioritised to the
PRSP priority sectors and programmes. This leads to increased
uncertainty over whether and when non-priority programmes
will receive funding, with resultant risks to efciency and
effectiveness in these latter programmes. Measures need to be
in place to reduce the uncertainty faced by non-priority sectors
in the release of funds.
The macroeconomic framework is expected to be the same
for the MTEF and the PRSP and, similarly, the aggregate
resource constraint will be the same for the MTEF and for
poverty reducing expenditures. Yet the PRSP raises the issue
of a desire to support an expanded resource envelope to
accommodate increased aid ows to contribute to poverty
reduction. This is a complex issue and a major concern is that
multiple sources of funds will risk softening the budget constraint.
Thus a key objective of the MTEF process must be budget
comprehensiveness, capturing the full range of resources available
to government on-budget.
Both the PRSP and budget processes require wide participation
in order to ensure ownership and credibility.A key concern is that
whatever institutional arrangements are built for participation in
the PRSP, they should at least link with those built for budget
participation. Consultations around the PRSP priorities need to
be timed so as to feed into the MTEF/budget process.
Parliamentary involvement in the budget process is essential
to ensure credibility and predictable budget implementation.
Evidence from the case studies shows that better costing and
target setting as part of the PRSP process can provide the basis for
more vigorous engagement of Parliament in the appropriateness
and realism of the detailed budget estimates.

ODI Brieng Paper June 2005

Recommendations

One of the truly signicant developments coming out of

effective MTEF/PRSP linkages is the much sharper focus on
the strategic phase of budget preparation and the accompanying
political engagement in resource allocation.What the PRSP is
also bringing to the budget table is a concern to follow through
on the results being achieved.
The central recommendation that arises from the case studies
is the need to continue to focus on developing a coherent and
integrated process of policy making, planning and budgeting.
A second broad recommendation is the need to manage
expectations. In the case of the MTEF, this includes managing
expectations about additional funding as well as expectations as
to the rate of progress in improving budgetary outcomes. In the
case of the PRSP, expectations need to be managed around how
quickly the PRSP will become the overarching policy framework
for all ministries and agencies and how quickly the linkage
between performance and resources will develop. A number of
more specic recommendations can also be made and listing
these also serves as a useful summary of the issues raised in this
brieng paper. These are presented in the table below.
Endnotes
1. Malcolm Holmes and Alison Evans (2003):A Review of Experience
in Implementing MTEFs in a PRSP Context: A Synthesis of Eight
Country Studies. The Albanian case study. The Albanian case study

Issue

Recommendation

Budget Basics

Build the Basics to support MTEF and PRSP implementation, particularly in the area of basic budget discipline

Comprehensive
Budget

Work to make the budget/MTEF as comprehensive as

possible

Institutional Set-up

A single central agency is preferable for coordinating

all aspects of the expenditure side of the budget. PRSP
implementation is also facilitated where it is coordinated
by the same budget agency

Credible Resource
Constraint

Realistic and honest estimates for both likely revenues

and expenditures are vital if the aggregate resource
constraint is to be credible

Political
Engagement

Involve Cabinet early on in decision making about the

budget

Top-Down Ceilings

Top-down ceilings, reecting not only the aggregate

resource constraint but also broad strategic sector priorities, should be employed as a key MTEF mechanism and
to facilitate the linking of the MTEF and the PRSP

Strategic Phase to
Budgeting

Support the strategic phase of budget making with a

document such as a Budget Framework Paper

Sector Working
Groups

Sector working groups, mandated to coordinate the

development of strategic approaches to budgeting and
to engage stakeholders, should be established where
an MTEF and PRSP exist

Performance
Orientation

Developing a more performance oriented budgeting system should begin with the centre focusing on improving
funding and policy predictability through the MTEF and
PRSP, and line ministries focusing on developing clearer
objectives and strategies, programme and strategic management capabilities and performance information

Resource Use

The efciency and effectiveness of resource use must

receive more attention

Decentralisation
Context

MTEFs and PRSPs being implemented in a country pursuing decentralisation should be managed in such a way
as to support decentralisation

was prepared separately and is included by kind permission of the

World Bank author. The other studies were undertaken in 2002/03,
with nancial assistance from DFID and the EC.
2. As about two years have passed since the case studies were carried
out, the situation in some of the countries may have changed.
3. On this issue, see also CAPEs work on Results-Based Management,
available at: http://www.odi.org.uk/PPPG/cape/publications.html
4.Although exceptions to this are Uganda where the poverty focus of
the MTEF pre-dated the donor-driven PRSP, and South Africa where
there has been a less explicit poverty focus.
The synthesis paper and all case studies can be downloaded
from: www.odi.org.uk/PPPG/cape/publications.html
For bibliography see next page.

The principal authors are Paolo de Renzio and Samantha Smith.

For further information, please contact p.derenzio@odi.org.uk
Overseas Development Institute 2005 ISSN 0140-8682
This and other ODI Brieng Papers are on ODIs website at:
www.odi.org.uk
Brieng Papers present objective information on important
development issues. Readers are encouraged to quote or
reproduce material from them for their own publications, but
as copyright holder, ODI requests due acknowledgement and
a copy of the publication.

Overseas Development Institute, 111 Westminster Bridge Road, London SE1 7JD
Telephone: +44 (0)20 7922 0300 Fax: +44 (0)20 7922 0399 Email: publications@odi.org.uk

Bibliography

ODI Brieng Paper June 2005

Bevan, D and Palomba, G. (2000) Uganda: the Budget and Medium Term Expenditure
Framework Set in a Wider Context, paper prepared as background for the World
Banks PRSPC, nanced by DFID.
Bevan, D and Adams, C (2001) Guidance Note: Poverty Reduction Strategies and
the Macroeconomic Policy Framework , paper prepared for DFID.
Booth, David and Lucas, Henry (2002) Good Practice in the Development of PRSP
Indicators and Monitoring Systems, ODI Working Paper 172.
Byaruhanga, Charles (2002) Poverty Reduction and Public Sector Reform in Uganda.
The Roles of Institutions in Past Experience and Challenges for the Way Forward.
Paper presented to seminar on Modernising Government: Integrating Structural
and Budget Reforms for a Better Performing Public Sector Adam Smith Institute,
London. April.
Campos, Ed and Pradhan, Sanjay (1996) Budgetary Institutions and Expenditure
Outcomes. Policy Research Working Paper, No. 1646. World Bank.
Folscher, Alta (2002) eds. Budget Transparency and Participation: Five African Case
Studies, IDASA.
Foster, Mick; Conway, T. et al (2002) How, When, Why Does Poverty Get Budget
Priority. Poverty Reduction Strategy and Public Expenditure in Five African
Countries. ODI Working Paper 168.
Moon, Allister (1998) Aid, MTEFs and Budget Process, Manuscript, World Bank
Schick,Allen (1998) Why Most Countries Should Not Try New Zealands Reforms,
World Bank Research Observer,Vol. 13, No. 1 p. 130.
World Bank (1998) Public Expenditure Management Handbook, PREM Network,
Washington D.C.
World Bank (2001) PRSP Sourcebook, Chapter on Public Spending and the Poor.
World Bank (2002) Medium Term Expenditure Frameworks: From Concept to
Practice. Preliminary Lessons from Africa, Washington DC.
World Bank/IMF (2002) Good Practices for PRSP Design and Implementation: A
Summary for Practitioners, Washington DC.

Overseas Development Institute, 111 Westminster Bridge Road, London SE1 7JD
Telephone: +44 (0)20 7922 0300 Fax: +44 (0)20 7922 0399 Email: publications@odi.org.uk