Michael Parry
December 1997
International Management Consultants Limited
Cowdray Centre House
Cowdray Avenue

COLCHESTER CO1 1QB
UK

Tel: 44 1206 549091
Fax: 44 1206 549098
EMail: IMC_Ltd@compuserve.com

TABLE OF CONTENTS

1          Budget structures and budget outcomes

1.1...... The Campos and Pradhan research

1.2...... Aggregate Fiscal Discipline

1.3...... Prioritization, Transaction Costs and Consensus Building

1.4...... Technical Efficiency and Incentive Incompatibilities

1.5...... Impact of Donors

1.6...... Country Experiences

1.7...... Comments on the Campos and Pradhan conclusions

2          Integrated financial management

2.1...... Internal linkages of integrated financial management

2.2...... Importance of classification systems

2.3...... Benefits from improvements in financial management

2.4...... External linkages and interfaces

2.5...... Budget comprehensiveness

2.6...... Dual budgeting - Development and Recurrent Budgets

3          Output oriented budget

3.1...... Accounting as an input-output model

3.2...... Performance budgeting

3.3...... Non-financial performance measures

3.4...... Ex-ante performance measures

3.5...... Ex-post performance measures

3.6...... Applying performance measures and indicators in developing countries

3.7...... Problems in introducing performance measures

3.8...... Conclusions on applying performance measures

4          Information technology

4.1...... Important changes

4.2...... Distributed or centralised architecture

4.3...... Packaged or custom developed software

4.4...... Country experience

4.5...... The millennium bug

TABLE OF EXHIBITS

Exhibit 1: Key institutional arrangements and expenditure outcomes............................... 4

Exhibit 2: Integrated financial management model........................................................ 7

Exhibit 3: Tracking expenditure in public sector financial management systems............... 9

Exhibit 4: Linkage of financial management improvements to benefits........................... 10

Exhibit 5: Recurrent and development expenditure budgets.......................................... 14

Exhibit 6: Public compared to commercial financial management.................................. 16

Exhibit 7: Linkage between economy, efficiency, and effectiveness............................ 18

Exhibit 8: Simplified diagram of centralized or distributed architecture........................... 23

Exhibit 9: Comparison of distributed or networked approaches...................................... 24

This paper will address four major themes:

1.         Current thinking on the link between budget structures and budget outcomes (Section 1 ).

2.         The linkage between these ideas and concepts of integrated financial management at a national government level (Section 2).

3.         Output oriented budgets and performance measures (Section 3).

4.         The application of information technology to the above problems and issues (Section 4).

1         Budget structures and budget outcomes

1.1                                   The Campos and Pradhan research

Prompted by wide variations between countries’ success in achieving intended expenditure outcomes, and especially perceived poor fiscal outcomes in developing countries, there has been important recent World Bank research by Campos and Pradhan [1] .  They sought to apply theories of New Institutional Economics to budget experiences in different countries, and they reached a number of very important conclusions.

Campos and Pradhan identify three basic fiscal objectives for countries:

1.         To instil aggregate fiscal discipline.  In the face of competing demands for limited financial resources, it is essential to be able to maintain controls over aggregate expenditure in the medium term.

2.         To facilitate strategic prioritization of expenditures across programmes and projects.  Within the overall resource envelope how to allocate resources in accordance with policy and social priorities.

3.         To encourage technical efficiency in the use of budget resources.  Ensuring that resources once allocated are used effectively and efficiently.

Within each of these areas, Campos and Pradhan identified three groups of factors which impact on the outcome:

1.         Institutional arrangements,

2.         Accountability, and

3.         Transparency

Their analysis developed the matrix set out in Exhibit 1 , below.  The conclusions are further discussed in the sub-sections 1.2 through 1.4 , after the Exhibit.

Exhibit 1: Key institutional arrangements and expenditure outcomes

The study concludes that aggregate fiscal discipline requires in the first instance a macro economic framework which can identify over the medium term resources that are available - the resource envelope.  This framework will need to embrace both internal and external resources, and needs to be comprehensive to include the whole Public Sector Borrowing Requirement.

Demands for resources will always tend to exceed available resources.  In order to maintain fiscal discipline hard constraints, e.g. UK cash ceilings, and/or legal constraints, e.g. a constitutional requirement for a balanced budget, are needed.  The budget process needs to embrace the maximum portion of public expenditure (problem areas include quasi financial activities of state financial institutions, hidden subsidies of state enterprises, direct use of public receipts).  Also central Ministries need to have control over all expenditure of subsidiary units.

Aggregate fiscal discipline is supported by ex post accountability, formal sanctions and open financial markets (the impact of the latter can currently be seen in South East Asia).  A fully transparent process of budgeting and press discussion is also seen as contributing.

Within a limited resource envelope, it is very difficult to ensure that the resources are allocated to programmes and activities in accordance with public policy priority.  Pressure groups, political interest, strength of groups within government, donors, and outright corruption result in many developing countries allocating resources in a pattern which does not accord with stated policy priorities.  Campos and Pradhan are particularly concerned at countries’ inability to focus expenditure priorities to be coincident with stated social priorities.

It is technically difficult, and expensive, to evaluate the costs and benefits of alternative uses of limited resources.  Objective criteria need to be developed and applied consistently.  Also there is an information asymmetry in that better information is available at a local, rather than central, level.  This favours a flexible approach with consultation and decentralization of detail resource allocation within overall resource envelopes and central budget management.

Prioritization is also supported by ex post reports and evaluations, and also by a Parliament with adequate technical capacity - often a problem in countries with relatively recent democratic processes.  Once again transparency is seen as important.

Technical efficiency in fact covers a whole range of different attributes.  These are often identified by the “Three Es” - Economy, Effectiveness and Efficiency.  These are discussed further under performance measurement.

Three institutional factors are identified as critical - human resource management within the public sector, autonomy of agencies (which is not seen as incompatible with dominance of central Ministries), and predictability in resource flows.  These are linked to important issues of accountability, including accounting, auditing, tenure of chief executives.  Transparency is again identified as important.

The study is critical of the impact of donors on the three target criteria.  Donors are seen as providing yet another set of priorities and conditionalities.  Donor priority setting can undermine national priority setting.  Projects can be fiscally inflationary through either, or both, future debt service and recurrent maintenance costs.  Donor demands for counterpart funds can undermine fiscal discipline. 

Donors also often impose their own systems of accountability, which can undermine attempts to develop a government’s own accounting and accountability systems development.

The study focuses on seven countries.  Australia and New Zealand are seen as being at the cutting edge of reform, though with somewhat different emphasis.  Malaysia and Indonesia are both seen as examples of developing countries which have successfully applied some of these lessons (though this might appear less so in late 1997).  The three African examples - Ghana, Uganda and Malawi - are seen as having been relatively unsuccessful.

The study seeks to establish an objective scoring system for the various aspects of fiscal management within countries, and uses this to score the various countries.

There can be no doubt that this is a very important study which contributes to the development of appropriate country strategies.  It is particularly commendable because it focuses on a limited number of key fiscal outcomes as measures of the success of the systemic changes.

Most of the conclusions are consistent with other writings and experiences of developing countries.  However, a number of points need to be made:

1.         There is evidence that aggregate fiscal discipline has more to do with government commitment than any technical system, and indeed some of the technical systems (e.g. New Zealand, Australia) have been implemented to support a pre-existing commitment to fiscal discipline.

2.         There does appear to be a conflict between the emphasis on a comprehensive centrally controlled budget for fiscal discipline, and the needs of decentralisation for better strategic prioritization.  There are lessons here from commercial financial management, which has moved from central control to a very decentralized approach, but identifying key strategic controls that need to be managed centrally.

3.         As with so many studies, budgeting is emphasized rather than accounting.  Yet budgeting only plans expenditure - actual expenditure is controlled and managed through a fund release and accounting system.  It is noteworthy that commercial financial management places much more emphasis on management through accounting than through budgeting.

4.         There is no consideration of the approach to reforms being led by the UK, involving the separation of policy making from service delivery, and as far as feasible the contracting of the latter to the private sector or quasi independent government agencies.  This can lead to gains in technical efficiency, and also create a better mechanism for prioritization.

5.         The study extends beyond financial management, but has a focus only on expenditure management.  It is not therefore not coincident with the coverage of a study of financial management systems.

6.         Accountability and transparency are not only means of achieving the identified benefits, they are benefits in themselves.

It is concluded that the Campos and Pradhan study provides a focus for future development in financial management.  It should not be seen as the end of the debate on appropriate strategies, many of the conclusions are widely accepted.  What is very important is the emphasis on the need for an integrated series of reforms, rather than improvements in one area in isolation.  This leads to the focus on an integrated financial management approach.

Historically governments have tended to regard the various sub-systems within financial management as being largely separate, within separate Ministries or Departments, and carried out by persons with different skills.  Initially, these subsystems may be posited as follows:

1.         Planning - perspective and project selection.

2.         Medium term financial planning, e.g. Three Year Rolling Budgets.

3.         Annual Budgeting - recurrent, development and revenue.

4.         Fund release and liquidity management.

5.         Accounting for revenue and expenditure.

6.         Reporting and financial statements, and related monitoring and control systems.

7.         Internal and external audit.

8.         Expenditure review.

These sub-systems and their important linkages, are set out in Exhibit 2 , below.  Note that the boundaries of the financial management system are defined, and the important external interfaces.

Exhibit 2: Integrated financial management model

The benefits from treating all subsystems as components within a single system are substantial:

1.         as illustrated by the model in Exhibit 2, the information outflows from one component are inflows to another component

2.         all utilize a common financial valuation and measurement model

3.         recognition of the linkages and commonality can improve performance within each sub-component, and hence of the system as a whole

4.         by treating the system as a whole, more effective procedures, information flows. training, and systems tools can be developed

5.         the imperatives created by modern information technology make it both more feasible, and the gains greater, from treating financial management as a single system

Such integration is also consistent with the conclusions emerging from the Campos and Pradhan study: that improved outcomes results from the combined effect of an integrated series of reforms.

These comments relate to and expand on the model in Exhibit 2 , above.

1.         Planning is linked to budgeting and accounting, especially through the process of selecting and monitoring of projects.  Both of these aspects should be fully integrated within the system to ensure consistent prioritization, and the use of the accounting system to provide financial information on project out-turns.

2.         A medium term financial planning framework, e.g. a Three Year Rolling Budget, is the linkage between a project/perspective planning system and the annual budget.  Experience is that Public Expenditure Plans are both too narrowly focused on expenditure, and too fiscally expansionary, to provide this linkage.  A medium term framework is essential both for aggregate fiscal discipline and for strategic prioritization.

3.         Fund release procedures are unique to public sector accounting, and introduce into the budget/accounting system elements of both liquidity management and also expenditure management.  They reiterate our point that it is at this stage that much de facto prioritization and expenditure control takes place.

4.         Accounting systems embrace all of revenue, expenditure and financing.  They will also include payroll as a sub-sub-system.  Payroll needs to be fully integrated with the accounting system, but will also need to interface with a personnel management system outside the parameters of the financial management system.

These linkages are reflected in the financial management system.  A significant difference of public sector and commercial financial management is the need in the former to track expenditure through its various stages.  This is illustrated in Exhibit 3 , below.  One of the problems in applying commercial accounting software to the public sector is the lack of facilities in most packages for such tracking.

Exhibit 3: Tracking expenditure in public sector financial management systems

The classification system is the common system that integrates the handling and presentation of information through the various sub-systems.  As such its design is of critical importance.  Classification systems are addressed in a separate paper within this workshop.

Seven areas of potential benefits from improved financial management are identified below.

1.         Enhanced resource mobilization

2.         Improved fiscal management through more effective expenditure management, institutions, processes and control mechanisms

3.         More optimal resource allocation decisions to achieve clearly articulated public policy objectives through enhanced identification of the costs and benefits of alternative expenditure decisions

4.         Improved liquidity management of public funds

5.         Improved technical efficiency in managing and utilising resources through improved information flows more relevant to decision responsibilities of managers

6.         Enhanced transparency and accountability of government, providing better historic information as a guide to the future

7.         Good government of public monies and assets, resulting in reduction in the levels of corruption and leakages

These benefits are linked to the key areas of financial management reform in the matrix below.  Note these embrace the Campos and Pradhan expenditure outcomes above.  A tick indicates an important linkage; the text in each cell points to the nature of that linkage.

Exhibit 4: Linkage of financial management improvements to benefits

It is not suggested that this matrix is totally comprehensive.  Nevertheless, it does provide an indication of the areas of focus and their likely impact in a reform programme.

The financial management system has a number of external linkages and interfaces.  Some of these are indicated in Exhibit 2 , above.  These and others are briefly addressed below.

1.         The system continuously interfaces with the national economy, and indeed one of the problems of government financial management is the Government’s dual responsibility for (1) its own operations, and (2) the economy as a whole.

2.         Government policy, particularly in the areas of borrowing and lending, state enterprises, prioritization of expenditure and taxation.  Revenue forecasting depends on government policy decisions.

3.         International donors and organisations.  Such organisations typically have their own requirements for accounting and auditing.  Meeting these within a government system can often cause serious problems.  There is a strong case for better co-ordination of requirements between donors, but this seems unlikely to be achieved.

4.         Various stakeholder groups, e.g. regional, special interest.  The system will often have to provide information relevant to such groups on an ad hoc basis.

5.         State enterprise interact with the Government financial management system through dividends, taxation, investments, loans, subsidies and other transfers.  These need to be tracked and identified by institution.

6.         Similar relationships occur with other state institutions and NGOs.

In addition the system will have a considerable number of other relationships.  Many countries have adopted a debt management system.  This has different objectives to a budget and accounting system, but they should be linked to together.  The widely used Commonwealth DRMS (debt management system) is we understand being re-written to facilitate such linkages.

One of the conclusions reached by Campos and Pradhan is that the budget needs to be comprehensive to be effective.  Effective budgeting requires that the budget process does in fact control all of the expenditures by central government.  Extra budgetary funds weaken the budget both as a resource allocation tool, and as a tool of fiscal management.  Most systems have the potential for large extra budgetary expenditures.  Some examples include:

1.         Funds received by line Agencies which are then available for expenditure, without passing through the consolidated fund.  There may be merit on occasions for linking expenditures to revenues raised, but these need to be controlled through a central budget process.

2.         Quasi fiscal activities of state financial institutions, e.g. loans at low interest rates, and/or without the expectation of repayment, to state enterprises.

3.         Direct access by Projects to donor funds. From a Project management perspective, it may be desirable to by-pass the bureaucracy and have direct access to donor funds, and indeed donors often encourage such a system.  However, this reduces the effectiveness of the budget process to control expenditure.

4.         The existence of multiple funds, outside the Consolidated Fund.  In such cases it is difficult to achieve effective control.  Normally there should at most be three Funds - Consolidated, Development and Emergency.

5.         Direct use by agencies of monies they collect.  In most countries this is against the Constitution, which requires all monies are paid into the Consolidated Fund, but it still happens.  However, from a managerial perspective such linkage may be beneficial, linking expenditure to collection efficiency.  Australia has a system which ensures such funds are paid into a central fund, but makes corresponding additional budget allocations to Agencies.

One of the problems of an inefficient financial management and fund release system is that entities within government will seek to obtain direct access to funding, to meet legitimate desires to deliver outputs.  A more appropriate solution is to address the systemic failures.

Many governments in developing countries divide their budgets into Recurrent and Development Budgets.  Development Budgets are primarily concerned with development projects, which have defined objectives and a finite life.  The Recurrent Budget deals with routine ongoing expenditure, where outputs are more difficult to define, and there is no finite end to the expenditure.  Ideally this budget division should coincide with the distinction between Revenue and Capital expenditure budgets, but in our experience Development Budgets often include expenditure of a revenue nature, and conversely Recurrent Budgets include capital expenditure. Thus a classification matrix emerges with two alternative classifications for the same expenditure:

This is illustrated with fictitious numbers in Exhibit 5 below.

Exhibit 5: Recurrent and development expenditure budgets

Development Budgets were initiated to identify and focus attention on public investment programs.  Development Budgets are frequently linked to development or public investment plans, and are subject to control and prioritization by a planning authority.  One of their objectives is to “ring fence” development expenditure so that socially desirable activities can be continued even if in the face of fiscal restraints.

There are a number of disadvantages of segregating Development and Recurrent Budgets:

1.         The administrative and procedural dichotomy between development and recurrent often reduces the effectiveness of fiscal control and policy prioritization;

2.         Difficulty in identifying and evaluating aggregate resources allocated to sectors, because they are spread over two budgets;

3.         Development Budget tends to receive more attention than the Recurrent Budget, though the latter is often larger in value;

4.         Recurrent maintenance expenditure is discouraged in favour of new projects, which may replace poorly maintained assets;

5.         A flow of development projects must lead to an ever increasing Recurrent Budget but Projects completed are often ignored, and the impact of such projects on the recurrent budgets not properly assessed in planning future aggregate levels of government expenditure;

6.         Leads to confusion on the more important distinction between capital and revenue.

On the other hand a separate development budget can often be administratively convenient. 

1.         A separate “Development Fund” can be established, making it easier to manage development funding from donors.

2.         Development activities can be “ring fenced” and dealt with differently reflecting their financing during period of fiscal restraint.

3.         The problem is not the Development Budget per se, but rather the administrative arrangements and the way information is presented and handled.

It may be difficult to change the existing budget structure, with a separate development and Recurrent Budget.  In such cases, the objective should be to integrate the preparation, format, classification and presentation of the two budgets.  Also the distinction between Development and Recurrent Expenditure Budgets should be made coincident with the distinctions between capital and revenue expenditure.

3         Output oriented budget

3.1                                   Accounting as an input-output model

Financial management in government has not achieved the same status as it has in the private sector.  There are many reasons, but one critical difference between the private and public sector is the absence of money measures of outputs in the public sector. This is explained by contrasting government and commercial entities.

For commercial entities both inputs and outputs, i.e. purchases and sales, are automatically defined in money.  Thus the accounting system provide a comprehensive input-output model which is universally applicable to all businesses, together with a widely understood performance measure - profit.  Accounting has as a result become the dominant method by which business performance is judged.  The primary objective of commercial management is to maximize profit.

Government entities face a different model.  Inputs are also automatically defined in money terms, but outputs, generally service delivery, cannot normally be expressed as money.  Hence there can be no concept of profit.  Furthermore, government revenues are not normally dependent on government expenditure - they are non-requited.  Thus money, and hence accounting, provides an input only model.  Outputs must be defined in non-financial terms.  This illustrated in Exhibit 6 , below.  This lack of a simple input-output model must be seen as a major problem of applying financial management to any public sector organisation, including national governments.

Exhibit 6: Public compared to commercial financial management

In the public sector, if performance is to be measured, then non-financial performance measures need to be introduced.  Performance budgeting provides such an approach.

Whereas program budgeting focuses on identifying related activities as programs, the emphasis of performance budgeting is linking financial inputs to non-financial output targets.  The approach recognizes that most government activities yield service delivery outputs.  Because these are not represented in money terms (see above), the tendency is to judge performance by expenditure according to budget targets - what the expenditure leads is not considered.  Performance budgeting seeks to move the emphasis away from expenditure to a focus on outputs, measured in non-financial units.  This requires the identification of normative output performance indicators linked to the financial inputs.

Performance budgeting therefore requires the identification of goals for budget activities, and the translation of these goals into normative output indicators.  The output indicators become the performance targets against which managers are judged - the benchmarks for management.  Such output indicators should be identified as part of the transparent budget process.

Performance measures may be ex ante or ex post.

1.         ex ante measure - used primarily to decide between decisions alternatives

2.         ex post measures - used to evaluate performance against previously identified criteria

The concept of performance also needs to be further refined.  Performance must be seen in terms of achieving public policy objectives, and is usually measured in terms of economy, efficiency and effectiveness (the “Three Es”).

1        Economy            achieving the policy objective with the minimum required resources, particularly minimum financial resources

2        Efficiency           achieving the policy objective in the most efficient manner (this is close to the concept of economy, but emphasises non-monetary resources)

3        Effectiveness    effectively achieving the policy objective, e.g. if the policy objective was improved education in all geographic regions, large regional schools might be of economic and efficient, but they would not be effective in achieving the policy objective.

Both ex-ante and ex-post measures are properly part of financial management.  Also both need to be integrated with other components of financial management, for reasons already indicated - the needs for common classification and valuation models, and the transfer of information

As indicated above, ex-ante measures are primarily used in deciding between alternatives, and they properly require a paper in their own right.  Essentially the procedures involve three stages.

There are a number of approaches to evaluation:

1.         Cost-benefit analysis - this involves the qualification in monetary terms of costs and benefits, and then discounting them to their present value.  Where costs or benefits are not expressed in money units, then “shadow” prices are created.  This approach has been used extensively but suffers from the fact that shadow prices are based on subjective assumptions

2.         Cost-effective analysis - seeks to identify the minimal cost method of achieving a given policy objective.

3.         Score-card approach - a matrix is constructed which identifies costs and benefits of decision alternatives, expressed in different units, and allows decision makers to use this in making a decision, without formally assigning weights.

4.         Matrix-criteria analysis - the approach takes the score card approach, and assigns weights to the various elements of the score card approach.

As indicated above, ex-ante performance measures, and their application to decision making, is a large subject beyond the scope of this paper.  The brief description above serves to identify the issues and indicate some possible approaches.

Ex-ante performance measures should ideally become the basis of ex-post performance measures.  Thus, for example, when a national planning body uses ex-ante measures in deciding on a project, those measures should become the basis on which ex-post performance measures are established.

As indicated above, the need is to measure economy, efficiency, and effectiveness.  The link between these can be expressed in different ways, but may be represented as in the model in Exhibit 7 , below.

Exhibit 7: Linkage between economy, efficiency, and effectiveness

It should be noted that there is a distinction between performance measures and performance indicators.  Measures are used where performance can be measured on a scale with a degree of precision, e.g. average female participation in primary education.  Indicators on the other hand imply a lower level of measurement ability, e.g. establishment of an effective primary education management structure in 70% of districts.

The following ideal characteristics of performance measures and targets.

Without indicators of achievement, financial measures are of limited value.  Also in many developing countries, both planning bodies and donors are de facto using performance measures for project.  The need is to integrate these into the financial management system but in a flexible manner.  Therefore the following practical guidelines are suggested.

1.         The starting point should be any existing system, e.g. a monitoring system established by the planning body

2.         Ministries should be encouraged to include within the budget a broad statement of their policy objectives

3.         Indicators should ideally be set when a project is initiated, and followed through the project’s life

4.         Wherever feasible, monetary measures should be used.  In a surprising number of cases, monetary measures are available of efficiency (e.g. costs per hospital operation, examination fees compared to costs), but cannot be calculated because of deficiencies in the accounting system

5.         The focus should be on indicators which are simple, and easily measured e.g. primary education participation rates are easy to calculate, whereas effectiveness of the education is much more difficult to evaluate.

6.         Measures can be used which require an element of judgement, e.g. establish an operational and effective health care management unit in X% of districts.

7.         It is necessary to “stand back” from a system of performance indicators and ask how they relate to policy objectives.

8.         Performance measures should be set by the manager responsible for their achievement

The first problem is likely to be organisational resistance, often exacerbated by previous bad experience of attempts to introduce programme budgeting.  Also, the linkages between financial management and performance targets may not be perceived.

Secondly, in many government organisational structures responsibilities for achieving outputs are not clearly defined, and so it is difficult to link performance measures to management responsibility.  Also managers may change, again diluting responsibility for performance.

Thirdly, in many developing countries systems for gathering, recording and reporting the information are inadequate and unreliable.  Indeed, in many cases even basic financial information is difficult to acquire in a useful format.

Finally, there is the need to establish routine procedures for publishing criteria, and then gathering and reporting information on performance as compared to the criteria. This will require the establishment of an appropriate database, lined to the budget and accounting systems.

Reporting of performances should be integrated with routine financial reporting.  In addition, at a more macro level, there is a need to include with government financial statements some indication of performance.  This latter is discussed later in this paper.

If a computer based budget-accounting system is being developed, it should be capable of holding performance indicators, out-turns against them, and including these within routine reports.

Performance criteria cannot be avoided, and always exist, at least subjectively.  The object should be to integrate them with financial measures. Any reforms in financial management should take account of the need for performance criteria, and allow for their eventual incorporation.  Such criteria are within the capacity of most developing countries, and indeed the concept may be easier to systems already geared up to handling development projects and donors.

Ultimately, performance criteria must be integrated within a financial management system.  Without such criteria, the system lacks any good orientation.  Such criteria should be set as part of the planning/budgeting process, and then reported on in