8 Experience of OECD Countries with Performance Budgeting

Marc Robinson
Published Date:
October 2007
  • ShareShare
Show Summary Details
Teresa Curristine1

Over the past two decades, the majority of OECD governments have sought to shift the emphasis of budgeting from inputs towards results. Across OECD countries, this introduction of a performance focus to budgeting has proven to be a powerful idea with applications to different levels and sectors of government. There is no single model of performance budgeting. OECD governments have adopted different models and approaches depending on their particular national capacities and priorities.


The introduction of performance information (PI) into budgeting by OECD governments is widespread and well established (nearly 75 percent of all OECD countries include non-finance performance data in their budget documents) albeit varied in the approaches taken to it. Seventy-seven percent of countries introduced their first government-wide initiative on outputs measures at least five to ten years ago, with over 40 percent working on outputs measures for more than ten years. Countries’ approaches to performance budgeting are not static, but rather evolving, with 75 percent of countries having introduced new initiatives within the last five years (Curristine, 2005a, pp. 92-3).

Countries use a variety of different methods to assess non-financial performance, including performance measures, evaluations, and benchmarking. In the 2005 OECD Survey on PI, approximately 90 percent of countries reported developing both performance measures and evaluations to assess performance. Of those countries that have developed performance measures, over 50 percent produce a combination of outputs and outcomes for most of their programs (Curristine, 2005a, p. 90).


Despite the length of time OECD countries have been engaged in performance budgeting, two of the major issues they face are, first, the quality and relevance of the data used as the basis for performance budgeting, and, second, ensuring such PI is used in the budget process.

As regards the first issue, not only does the quality of PI vary widely across OECD countries but it also varies widely within countries themselves, from ministry to ministry and from sector to sector. On a positive note, in the 2005 OECD Survey on PI, countries reported that over the past five years the quality of performance data available to them has improved. However, despite this improvement, more progress is needed. For example 50 percent of Ministries of Finance (MoFs) who responded to the 2005 OECD Survey reported that the cost-effectiveness and value for money reviews provided by their spending ministries were insufficient, incomplete and, in many cases, inaccurate.

As regards the second issue, the extensive increase in the volume of PI available to OECD governments has not been matched by a corresponding increase in use. For performance budgeting to succeed, it is important that PI be used in budgetary decision-making. One of the first issues for countries to address is ensuring that there is a process in place for such information to be integrated into the budget process. While many countries have adopted such a process, how it has been instituted varies from country to country. Furthermore, countries’ experiences show that even having a procedure in place to integrate performance information into the budget process does not guarantee its use in budgetary decision-making. Other factors influencing the use of PI include the quality, relevance, and credibility of the data, the institutional capacity of the relevant MoF and spending ministries, and the political and economic context.

For OECD countries, performance budgeting has proven a challenging task given the technical complexity of implementing it, as described above, and also given the behavioral changes it requires of decision-makers. Civil servants and politicians must learn to change their behavior and use PI in the decision-making process. In order to achieve this change, countries need to introduce reforms to create incentives to encourage civil servants and politicians to use this information in a manner whereby governments may reap positive benefits and avoid perverse incentives.

This chapter concentrates on performance budgeting (PB) at the central government level, and does not address the efforts of state or local governments in this area. Furthermore, it focuses on attempts to introduce a government-wide approach to PB. Throughout this chapter the term “performance information” is used to refer to both evaluations and performance measures.

This chapter is divided into three parts. The first examines attempts to integrate PI into the annual budget process. The second discusses the use of PI expenditure prioritization exercises, concentrating on spending reviews. The third examines other factors that influence the use of PI and the challenges OECD countries continue to encounter to improve its use in budgetary decision-making.

1. Integrating PI into the annual budget process

PI can be used at different stages and levels of the budget process. This section concentrates on the use of PI at the budget formulation stage in budget negotiations between the MoF and spending ministries and briefly addresses its use in negotiations between spending ministries and agencies.

PI in budget negotiations between the MoF and spending ministries

OECD countries have taken a variety of approaches to including PI in budget negotiations. These can be split broadly into formal and non-formal approaches to including PI in budget negotiations. Even within such categories, countries have adopted a wide variety of methods, some of which are described below.

Some countries have followed a formal comprehensive approach, in which the MoF requires all ministries to present performance plans and/or performance results along with their spending proposals. Other countries have a formal but non-comprehensive approach that requires performance plans or results either (1) from some ministries and sectors only, or (2) only when ministries present proposals bidding for additional funding (for either new or existing programs).

As regards non-formal approaches, some countries leave to individual ministries the decision as to whether to present PI to the MoF in the budget context. Certain other countries make systematic use of PI in negotiations between ministries and their subordinate agencies over performance contracts or agreements, while excluding PI from any significant role in budget between the MoF and ministries.

Even having formal requirements to present PI to the MoF does not mean it will be used in budget negotiations in all cases. As can be seen in Figure 8.1, in over three-quarters of countries that responded to the OECD 2005 Survey, performance measures are part of the budget discussions between the MoF and either all spending ministries (29 percent) or some spending ministries (50 percent). There is less use of evaluations: 51 percent of countries use them in budget negotiations, and less than 10 percent use them for budget negotiations in respect to all spending ministries.

Figure 8.1.Is PI used as part of budget discussion between the MoF and spending ministries?

Source: OECD 2005 Survey on PI in the Budget Process.

When PI is part of the budget negotiation process, a key question is how it is used. One may distinguish between the use of PI by the MoF for planning purposes and for accountability purposes. In both these cases there is an issue about how PI should be linked to funding. In the first case, the question is the degree of linkage between funding and expected results. In the second case, the question is how far funding is affected by actual results achieved in the past.

Planning purposes

Most budget negotiations have traditionally included some element of discussion on planning—that is, about expected performance—as budgetary estimates often provide some indication of what a ministry aims to achieve with funding (for example, to build a certain number of roads or hospitals). The introduction of performance budgeting has formalized this process and placed a greater emphasis on setting explicit expectations or targets for the results to be delivered with the funding provided.

The majority of OECD countries do not have a systematic government-wide approach to linking expenditure to targets. As can be seen in Figure 8.2, over a third of OECD countries reported that expenditure is linked to some output or outcome targets, while over 46 percent of countries only link expenditure to a few or no outputs or outcome targets.

Figure 8.2.Are expenditures specifically linked to each output or outcome target?

Source: OECD/World Bank Survey on Budget Practices and Procedures 2003.

Of course, performance plans and targets are not necessarily discussed or approved during the budget process, but may take place outside of the budget process either with the MoF or with other central agencies. Indeed, the development of these plans can be part of an unconnected reform initiative. In some countries, planning is completely separated from budgeting, and strategic and performance plans are primarily presented and approved by the Prime Minister’s/President’s office, the Ministry of Planning, or the legislature.

Accountability purposes

The MoF can use PI to hold ministries and agencies accountable for their performance, including by taking the past performance record into account when deciding on budget allocations. A key issue in performance budgeting is how tightly funding can be linked to performance results. In practice in most OECD countries, there is, at the level of the budget as a whole, at best only a loose or indirect link between funding and performance results. The majority of MoFs engage in what might be referred to as “performance informed” budgeting—that is, performance data is used along with other information on fiscal policy, policy priorities, and political factors, to inform but not to determine budget allocations. According to the 2005 OECD Survey, 76 percent of MoFs frequently use performance measures, and 66 percent frequently use evaluations, to inform budget decisions (Curristine, 2005a). In these cases, there is no automatic or mechanical link between performance results and funding. PI is only one consideration in the decision-making process on the level of funding to be provided, and how much weight is given to it will be influenced by the wider economic and political context and specific characteristics of the relevant policy area.

The exception to this is performance budgeting based on formula budgeting. This approach is not used in OECD countries at the level of the budget as a whole but, rather, when it is used it is applied in specific sectors. Two-thirds of respondents to the 2005 OECD Survey on PI stated that they do not directly link performance results to appropriations. Formula funding is mainly applied to sectors such as health and education.

In the majority of cases, the MoF does not often use performance results to financially reward or punish agencies. Table 8.1 shows the percentage of MoFs in OECD countries that often use evaluations or performance measures to eliminate programs, to cut expenditure, or to determine pay.

Table 8.1.Percentage of OECD MoFs that often use PI for the listed courses of action
Performance measuresEvaluations
To eliminate programs:4%To eliminate programs:11%
To cut expenditure:10%To cut expenditure:15%
To determine pay:11%To determine pay:5%

As can been seen from these numbers, it is rare that PI is used by the MoF when deciding on these courses of action. The difficulty in linking funding to results reflects the fact that the issues and context surrounding budget decisions are complex. The capacity of the MoF to eliminate or even cut back programs may be restricted by lack of institutional capacity and power or lack of political support.

In addition, the actual decision-making power can rest with the Prime Minister or be the result of political negotiations within the cabinet. Budgetary decision-making takes place in a political context and proposals to cut back or eliminate programs can encounter political resistance, especially if these programs bring benefits to important political groups or are high political priorities.

It is questionable whether financially rewarding good performance and penalizing bad performance on a government-wide level is likely to lead to improved agency performance. To automatically cut funding without understanding the causes of poor performance could make the situation worse and condemn badly performing agencies to continue to underperform. Performance measures do not explain the underlying causes of poor performance. Performance in any given year can be influenced by a variety of factors both internal and external, which may or may not be within the control of an agency. In addition, in some OECD countries, it is uncertain if the PI is of high enough quality to be used in budgetary decision-making in this manner.

In addition, such an approach can generate perverse incentives and encourage agencies to manipulate data. Incentives to provide accurate information are influenced by the expectations of how it will be used in decision-making. If funding is tightly and automatically linked to results, there can be incentives to engage in gaming and to manipulate data in order to receive more money or to avoid receiving less. As Wildavsky (1974) pointed out over 30 years ago, it is politically irrational to expect agencies to provide objective information if it will be used to cut back their programs.

Automatically linking performance to results on a systematic government-wide scale is not the best approach to promoting the production of credible information and encouraging agencies to focus on maximizing their use of PI. No OECD country has done this, perhaps realizing the difficulties of adopting such a systematic approach.

While MoFs do not financially penalize or reward agencies for their performance, they do, however, still use PI to hold ministries to account. PI acts as a potentially useful signal of problems within programs and agencies and serves as a trigger for the MoF to more closely monitor or review agencies and programs. The most common course of action taken by MoFs against poorly performing agencies is that resources are not increased and the program is reviewed during the course of the year. Other actions include making the continuation of programs conditional on future performance.

PI in budget negotiations between spending ministries and their agencies

OECD research indicates that PI is more often used by spending ministries than by the MoF (OECD, 2005b). PI can be used at the budget formulation stage in negotiations between spending ministries and their agencies. In this context, a common approach to integrating PI into the budget process is through discussions on agencies’ performance agreements and contracts. This is an approach widely used in countries with executive agencies, such as New Zealand, the UK, the Netherlands, and the Nordic countries.3 These discussions may concentrate either on future targets or on past performance, or may involve a combination of both. With the exception of the purchaser-provider model used in New Zealand, in many cases there is only a loose link between funding and targets.

Another common use of PI by spending ministries is to redistribute resources (see Figure 8.3). The capacity of spending ministries to reallocate resources is influenced by the wider budgetary framework and rules. In most countries, ministries at the budget formulation stage have the capacity to propose changes in the distribution of resources between subordinate agencies and programs. Existing rules and procedures can sometimes force spending ministries to reallocate. The MoF in some countries—for example, New Zealand and Sweden—seeks to control expenditure and limit proposals for increased spending through the use of fixed spending caps. This requires ministries to remain within the set expenditure limits and forces them to cover any increase in spending through internal reallocations.

Figure 8.3.How are the results of performance measures generally used by spending ministries/departments? (Multiple responses possible)

Source: OECD 2005 Survey on PI in the Budget Process.

Across OECD countries there is a wide variation in the quality and use of PI by spending ministries in the budget process. Even within the same country, there can be wide variations among different ministries in terms of the quality, the extent of use, and the weight given to PI in budget discussions. For example, in Norway, PI is requested by the different ministries in their letter of instruction to agencies. Despite formal requirements and processes, many agencies have made only limited progress in developing meaningful performance measures. The character of PI changes from agency to agency. Some have defined objectives and developed a few performance measures; others do not develop any performance measures. These problems make it difficult for spending ministries to use this information in budget negotiations with agencies. Like many other OECD countries, Norway struggles with the problem of developing good-quality performance measures and data (OECD, 2006, p. 21).

As already indicated in this volume, PI is not just used for budgeting purposes; it also aims through the “managing-for-results” approach to improve the management of programs. As Figure 8.3 indicates, the most common use of performance measures by ministries is to manage programs.

PI has the potential, if it is of good quality, relevant, and timely, to be used in such a way as to improve programs. It can indicate the degree to which programs are achieving their objectives and help to identify which policies and processes work and why These factors can help to improve program management and thereby boost efficiency and effectiveness. There are no comparative cross-country studies showing the extent to which the use of PI in this manner has in practice improved performance. These types of studies are difficult to conduct due to comparability issues and also to the difficulty of distinguishing the impact of the use of PI from that of other relevant initiatives and factors. However, in a national country context there are many anecdotal examples of how PI has improved operational efficiency, some of which are discussed elsewhere in this volume.

2. PI and expenditure prioritization

Expenditure prioritization is concerned with clarifying governments’ key objectives and priorities and tunneling funds towards high-priority areas and away from lower-level priorities. Expenditure prioritization seeks to improve allocative efficiency and to reallocate resources. The motives for governments to engage in central expenditure reallocations across government are diverse and include new political priorities, fiscal stress or abundance, program overspending, and substitutions of inputs such as changes in technology (OECD, 2005a, pp. 10-11). OECD countries have used a variety of methods and budget institutions to approach reallocation, which include medium-term frameworks, different rules for budgetary discipline, and spending reviews.

This section focuses on the approaches that potentially make use of PI in expenditure prioritization decisions, and therefore concentrates on spending reviews.

Expenditure prioritization and spending reviews

Here we examine the use of spending reviews as an approach to increasing allocation efficiency Spending reviews can be ad hoc or systematic and part of the annual budget cycle or separate from it.

Program evaluation is a particularly useful input for spending reviews. As Figure 8.4 highlights, the majority of program evaluations are conducted by spending ministries. OECD MoFs have, however, found the quality of the evaluations produced by spending ministries to be poor, particularly in respect to evaluations of cost-effectiveness and “value for money.” MoFs therefore in many cases conduct their own evaluations to provide them with information to be used in the budget process.

Figure 8.4.Who is responsible for conducting evaluations in your country?

Source: OECD 2005 Survey on PI in the Budget Process.

Spending reviews differ from ordinary program or policy evaluations in that they are centrally driven exercises and they concentrate on issues of allocative efficiency across government and examine the consequences of alterative funding levels.4 These reviews may be conducted either on an ad hoc basis or as a systematic part of the budget process. The motives for engaging in spending reviews vary, although they are influenced by the underlying economic context and are more likely to occur under conditions of either fiscal stress or fiscal abundance.

Ad hoc spending reviews

An example of expenditure prioritization and reallocation that took place under conditions of fiscal stress is the 1994 Canadian Program Review. This was a one-time exercise driven by a combination of factors, including high annual deficits, pressure from the financial markets, and the election of a new government. The review was driven from the center, which established the guidelines, set departmental spending, cut targets, and specified how departments were to go about reviewing their programs (including assessment of program objectives and purposes, the impact on society, whether the program should be provided by government, and means of improving program efficiency). Using these centrally established qualitative criteria, all departments provided action plans setting out their strategic priorities and alternative packages of programs and activities. Final plans were submitted to a special committee of ministers set up by the Prime Minister. The committee made specific recommendations on departmental programs to the MoF which incorporated them into the budget. The results were substantial cuts (averaging 21.5 percent) to departments’ budgets (OECD, 2005a, pp. 14-15).

Given that this exercise took place under conditions of fiscal stress, while performance criteria on program effectiveness and efficiency were included it is not clear how much weight was given to this information during the decision-making process, especially since the decisions took place in a highly charged political context and involved discussions and negotiations between ministers.

Systematic spending reviews

Other countries have sought to design systematic cross-government expenditure reallocation mechanisms. These include the UK with its Spending Reviews (starting in 1998), the Netherlands interdepartmental policy reviews (1981), and value for money reviews in New Zealand (2001). All these reallocation initiatives were designed during times of fiscal abundance, and the primary objective in many cases was to improve the performance of the public sector and allocative efficiency (OECD, 2005a).

The UK process is discussed in greater detail in Chapter 12 by Peter Smith. The system was initiated with a Comprehensive Spending Review in 1998, followed by Spending Reviews in 2000, 2002, and 2004. A further Comprehensive Spending Review was underway at the time of writing, and was scheduled for completion in 2007. This biannual approach aims to reallocate money to key priorities and to improve the efficiency and delivery of public services. After a review of existing departmental spending, each department develops three-year spending plans and Public Service Agreements (PSAs). The Treasury negotiates with ministries their key performance targets for the next three-year period, which are included in their PSAs.

The Dutch approach differs from the UK approach in that it is an annual exercise which is not designed to be comprehensive. The Dutch policy reviews focus on a limited number of priority areas every year (approximately ten) and were originally conducted with the idea of developing alternative funding scenarios for these policies and recommendations on how to improve efficiency The reviews are carried out by small working parties, which are often chaired by an external member and include participants from the relevant spending ministry, the MoF and the Ministry for General Affairs. The results of these reviews are published and submitted to parliament. These reviews have been used by both the spending ministries and the MoF during the budget process, and have also been used by cabinet and opposition parties (OECD, 2005a, p. 43).

Challenges with spending reviews and reallocation

The experiences of the UK and the Netherlands are seen as successful examples of spending reviews. By contrast, the New Zealand experience with value for money reviews underlines the difficulties in conducting these exercises. For the 2002 budget year, the New Zealand cabinet agreed to 17 reviews, but, despite the recommendations in these reviews, only two significant reallocations occurred. These instances were in ministries that had participated in the exercise voluntarily. The New Zealand reviews highlighted the need for obtaining the cooperation of spending ministries and political support in the recommendations if reviews are to be implemented (OECD, 2005a, p.19).

While spending reviews can highlight the need for cross-departmental reallocations, in order to actually implement change, high-level political support from the cabinet and/or the Prime Minister may be required and, depending on the nature of the political system, the legislature. This depends in part on the nature of the political and administrative systems and relevant power of the MoF. In only a few countries is the MoF powerful enough to carry out an exercise in cross-departmental reallocation without high-level political support. In the US system, the support of the legislature is also needed for most reallocations across departments.

Notwithstanding the examples discussed in this section, significant central reallocation across government is not common. In OECD governments’ budgets, except in conditions of fiscal abundance, the funds available for reallocations are generally marginal because of the extent of mandatory spending, entitlement programs, and prior commitments. Politicians are often reluctant to place certain spending, such as entitlement programs, in the spotlight. In this sense, in many OECD countries much of the annual budget process remains incremental, and inputs still play a significant role. Even in circumstances when reallocation occurs, the reallocation is not necessarily the result of systematic spending reviews. This is just one of many possible reallocation institutions and mechanisms.

Although program performance can be a valuable input in spending reviews, one should not expect a mechanical linkage between performance and spending cuts or increases; particularly because budget decision-making takes place in a political context, so poorly performing programs and those with high-priority objectives may receive funding increases.

PI must compete for attention with other sources of information, priorities, and tools in the budget process. While the MoF and the budget office have the objective of improving allocative efficiency, their primary role is to maintain aggregate fiscal disciple. Although this volume argues that over the long term PB can help achieve this, other more traditional and perhaps more blunt tools such as centrally imposed spending cuts for ministries can help to achieve this in the short term, although there may be an issue with long-term sustainability. Even if PI is integrated into the process, timing issues can work against the use of PI. With the tight deadlines in the budget process, there is a danger that serious discussions of performance or even of targets are left out in favor of discussions on inputs.

3. General factors influencing the use of PI in the budget process

Sections 1 and 2 highlighted that, while having procedures to integrate PI into the budget process is important, it does not necessarily mean that PI will be used in decision-making. Key actors in the process must have the incentive to change their behavior and break from the traditional approach to budgeting.

Other general conditions and factors which influence the use of PI in the budget process include the quality of PI, the institutional capacity of MoFs and spending ministries, and the wider political and institutional context. This section concentrates on the latter two factors. OECD countries continue to face challenges regarding many of the technical aspects of these reforms and with providing incentives for key actors to use this information in budgetary decision-making.

Institutional capacity of the MoFs and spending ministries

The institutional capacity of the MoF influences the role it is able to play in promoting the use of PI in the budget process at all levels. The capacity to steer reform initiatives and expenditure prioritization exercises is influenced by its position in the wider institutional framework and its power vis-à-vis other central agencies and spending ministries. This affects the extent to which the MoF can impose its views, as opposed to being forced to rely on persuasion.

Across OECD countries, the role played by MoFs in developing PB initiatives, and their subsequent implementation, varies widely. It ranges from developing performance measures and setting and/or negotiating performance targets, to a completely decentralized approach in which the MoF has no involvement. In some countries, such as the UK and Chile, the MoF has a high degree of involvement and plays a strong and active role in developing and implementing these reforms. This more centralized approach contrasts with the much more decentralized one in, for example, Iceland, where the MoF has a low level of involvement and it is solely the responsibility of individual ministries to develop and implement these reforms.

The degree of involvement of the MoF is a function not only of the PB approach and the level of interest in these initiatives, but also of the relative position and power of the MoF in the wider institutional and political system. For example, in the UK, the Treasury has taken a PB approach, which involves expenditure prioritization through two-year Spending Reviews. The strong formal role for the MoF in this initiative has been promoted by a high level of political interest and the centralized administrative and institutional structure in which the Treasury has a powerful position.

There are risks with having too much or too little central involvement from the MoF or other central ministries in these reforms. Previous incarnations, such as the Planning, Programming, and Budgeting System (PPBS) and zero-base budgeting (ZBB) have been accused of being overcentralized. Too centralized an approach can result in those responsible for delivering the target lacking the flexibility and managerial freedom to get the job done and the motivation to deliver results. Furthermore, centrally set and driven performance targets can distort behavior at lower levels, creating incentives to cheat and twist information in order to meet requirements at the center. If PI is not used in the manner intended, given the extensive costs in generating it, it risks becoming an expensive paper exercise.

On the other hand, too little central involvement can result in no pressure to change. The risk is that reforms are announced and not implemented at lower levels. If there is too much decentralization, the center can lack the information necessary to make decisions. There can also be a lack of uniformity in the development and presentation of PI. This makes it difficult to compare performance across government and hinders expenditure prioritization. In addition, central coordination can promote the development of joint goals and initiatives which cut across ministries.

For each country, the relative risk of having too much or too little central involvement is influenced by the existing institutional structure. Those countries that operate in a decentralized system of public administration, in which the MoF has relative weak formal power, face a different set of challenges to those that operate in a more centralized system.

Resources in terms of staff and expertise also influence institutional capacity. PI is different from financial information. In order to make judgments and compare performance, the MoF needs the relevant expertise to be able to analyze and evaluate the information received from different spending ministries.

Since spending ministries are dependent on agencies for information, they, like the MoF, need the capacity to understand and evaluate the information they receive. This is necessary if they are to make judgments about the appropriateness and quality of performance measures, and about the realism of proposed targets. It is mostly the agencies that decide on detailed goals, priorities, performance measures, and targets. Even if the interest is there, ministries may not have the expertise or knowledge to develop performance measures or even to effectively monitor performance. The danger is that this can lead to the passive provision of data without it having any real weight in the decision-making process.

While this section has concentrated on the role of the MoF, it is not always the MoF that plays the central role in initiating or monitoring PB reforms. For example, in Brazil the Ministry of Planning is responsible for developing and monitoring PB reforms. In some countries the responsibilities are split across departments. For example, in Korea, the Ministry of Planning and Budgeting is responsible for performance budgeting and operates a system similar to the US Program Assessment Rating Tool (PART) system. The performance management side is divided between the Ministry of Government Administration and Home Affairs, which is in charge of general management, and the Civil Service Commission, which is responsible for personnel policy. The office of the Prime Minister coordinates the overall reform effort and other ministries work in this area.

Importance of politicians and the wider political and institutional context

Politicians have an important role to play in promoting the development and use of PI in the budget process. Their role potentially involves applying pressure on other actors to implement PB, playing an active role in setting objectives, and using PI in budgetary decision-making. The role of politicians in the legislature and executive will vary depending on the nature of the legislative-executive relationships in the budget process, which in turn is influenced by the type of political system—presidential, semi-presidential, or parliamentary.

Most models of PB and “managing-for-results” more generally assign to politicians the role of setting clear objectives for agencies and monitoring performance against those goals. Most OECD governments now present their objectives and performance results to the public either in government-wide annual performance plans and reports (as, for example, in the UK), or in ministerial-level plans and reports (as, for example, in Canada and the US) (OECD 2005b). This approach, if successfully implemented, can help to clarify what the government is trying to achieve and how programs contribute to governments’ wider policy objectives or outcomes, and it can also help to improve transparency.

However, politicians have not always availed themselves of this opportunity. Setting clear objectives is one of the challenges that OECD countries continue to encounter. In any system with multiple principals, or lack of agreement on the role of an agency, there can be competing and even conflicting goals and demands for agencies. This problem is more pronounced in separation-of-powers systems in which control of the bureaucracy is divided between the legislative and executive arms of government, as in the US.

It has proven particularly difficult to get politicians in their capacity as legislators to pay attention to and to use PI.5 Only 19 percent of OECD legislatures use PI in decision-making, and the figure is only 8 percent for members of legislative budget committees (OECD, 2005b, p. 72). This is despite the fact that in some cases these reforms were the initiatives of the legislature and not the executive; for example, the Government Performance and Results Act (GPRA) in the US.

A key challenge is to create good-quality and relevant information that takes account of the constraints of timing and the capacity under which political decision-makers operate. There has been a large increase in the volume of PI provided to the legislatures in OECD countries. This information has improved transparency, and it enables politicians in the legislature, should they wish to do so, to use this information to hold the government to account. There are examples of this both in the UK and Holland, where the government opposition has used the government’s performance results to highlight problems.

In many cases, however, politicians complain about receiving too much information of variable quality and relevance. Not only is there a problem of information overload, but it is also often the case that performance information is not presented in a clear or comprehensible manner. Politicians in the legislature and the executive have different informational needs. To be useful, information needs to be tailored in such a way as to recognize these differing requirements. It should also be provided at the right time for the relevant decision.

In most countries these reforms have not resulted in any changes to procedures in the parliamentary budget process. The process has not been altered to integrate PI into budgetary decision-making. The exception to this is France, where the Loi Organique relative aux Lois de Finances (LOLF) reform has given a greater role to the legislature.

The support of politicians in the cabinet or that of the Prime Minister is often needed to implement the recommendations of spending review and expenditure prioritization exercises. However, their willingness to do so can be influenced by many factors in the wider political and economic environment. For example, coming close to an election, or during times of fiscal abundance, it can be difficult to obtain political support for recommendations to reduce spending even for programs that are inefficient. Furthermore, politicians want to be responsive to their constituents and special interests. In some political contexts, programs and agencies are continued even though their existence is questionable on efficiency and effectiveness grounds. Meeting these political needs is not necessarily conducive to using PI in budgeting decision-making.

This approach can be more of a challenge in systems where the legislature is powerful and influential in the budget process. For the reform to succeed it requires not only that the politicians in legislature and the executive change their behavior and support this approach, but also close cooperation between them. Whether the political system is consensual or adversarial will impact on their ability to cooperate and the use politicians make of PI.

It needs to be recognized that politicians are slow at changing approaches and methods, and they are faced with a different set of incentives and structures than civil servants. Many OECD countries continue to struggle with providing the right incentives to motivate politicians, especially those in the legislature, to use PL6


There is no one model of PB, and OECD countries have adopted a variety of methods and approaches to it. OECD countries now have a mechanism which enables politicians and managers to set and clarify objectives and to monitor progress toward achieving these goals. PB has encouraged OECD countries to put greater emphasis on planning and to move towards an outcomes focus in policy design and delivery, and has also improved transparency by increasing the amount of information provided to the legislature and the public on the performance and results of the public sectors.

Despite these benefits, OECD countries continue to struggle with a number of challenges and issues. It has proven easier to increase the quantity of PI produced than to ensure that this information is actually used in budgetary decision-making. OECD countries have adopted a variety of processes to integrate PI into the budget. PI is part of budget negotiations in most OECD countries and in some cases it has been part of expenditure prioritization through spending reviews. Nevertheless, making sure that these processes work remains a challenge. Integrating PI into the budget process is a necessary but not sufficient condition to assure its use in decision-making. Most importantly, it is not just a matter of process, but also of having the right incentives to motivate decision-makers to use this information in the budget process.

OECD countries continue to face the question of how tightly to link funding to performance. The current practice on a government-wide scale in OECD countries is not automatically or mechanically to link funding to results. It is seriously questionable whether such an approach on a government-wide scale is even desirable given the quality, technical, behavioral, and incentive issues involved. PI is used in budget negotiations along with other information on fiscal factors and political priorities. It is used to inform but not to determine budget allocations. PI plays a greater role at the decentralized level. Spending ministries and agencies make the most use of PI, and do so principally to manage programs.

In most OECD governments’ budgets there is little room to maneuver, and reallocation decisions are based on marginal funds. In making decisions on marginal funding, performance is only one of many factors that can be taken into consideration. The amount of weight given to this information depends on the information itself, the relevant sector or policy area, and the wider political and economic factors. PI must compete with other sources of information and budgetary tools to gain the attention of decision-makers.

Other factors influencing its use include the quality of the PI itself, the capacity of the MoF and spending ministries, and the wider institutional and political context. Although there are some exceptions, most OECD countries continue to struggle with providing good-quality and relevant information to decision-makers in a timely manner. They also continue to face the challenge of changing the behavior and culture of civil servants and especially politicians. What incentives to provide and the capacity to provide them are decisions influenced by the wider political and institutional structures.

As PB is intuitively appealing, it is easy to oversell and exaggerate its potential. In some OECD countries, reformers have failed to learn from earlier incarnations of PB, have initially been overambitious, and have oversold the reform potential to administrative and political leaders. This has resulted in unrealistic expectations about the capacity of PB to transform budgeting. It is important to bear in mind that the central budget process remains a political process, that inputs in most countries still play a central role and that, with the possible exception of times of fiscal constraints or abundance, budgeting remains in many cases largely incremental, with PI impacting on the margins.

There is no one approach to PB that can succeed in all countries. Rather, each model of PB needs to be adapted to the relevant context and be seen as part of a learning process. OECD countries’ approaches are continuously evolving as they learn from their experiences and take account of PB’s potential and its limitations.


The views expressed in this chapter are those of the author and do not represent the opinions of the OECD.

The data presented in this chapter are from two sources. First is the OECD 2005 Survey on the Development and Use of Performance Information in the Budget Process. This questionnaire was sent to the Ministries of Finance in all OECD countries and two observer countries—Chile and Israel. There was a high response rate: 26 out of 30 OECD countries and the two observers completed the questionnaire. All answers are self-reported. The second source is the 2003 OECD/World Bank Survey on Budget Practices and Procedures, for which 27 out of the 30 OECD countries completed this survey.

For more details on the definition and principles of executive agencies see Schick (2002).

In some of the literature the term “program review” is used refer to spending review.

The UK is an exception to this with its Special Cabinet Committee on public services and expenditure chaired by the Chancellor.

For a more detailed description of the possible incentives motivating politicians to change behavior see Curristine (2005b).


    Other Resources Citing This Publication