Top-down expenditure ceilings are the bedrock of the budget process. They operationalise the government’s high-level fiscal objectives in concrete terms and provide a focus for budget allocation and control. This is done through a multi-stage process where the government decides on aggregate spending allocations before negotiating detailed appropriations. Top-down expenditure ceilings should cover a multi-year period and allocated to ministers to support accountability. Flexibility to amend ceilings should follow well-defined rules, including for managing cyclical expenditure and accommodating unforeseen spending pressures. The central budget authority should have the capacity to assess such reallocations and ensure that decisions by line ministers are within the overall policy and budget framework.
Quality Budget Institutions
Abstract
Top-down budgeting refers to setting limits on total spending within which more detailed allocations are then agreed. One development that solidified the move to top-down budgeting in OECD countries was the decision to organise the budget process in a manner where the government first agrees a fixed envelope (a ceiling) for aggregate expenditure, before deciding how to allocate the resources into appropriations across line ministries and other government entities. This established a budget process where negotiations over how to allocate resources across ministries are within the limits of that ceiling.
Top-down expenditure ceilings have become the bedrock of a modern budget process. They operationalise the government’s high-level fiscal objectives in concrete terms and reflect allocations and reallocations to fund new policy initiatives (Box 4.1). They are complemented by other reforms that extend budget planning over a medium-term horizon and give ministers more flexibility to allocate resources within their portfolio. Together, these practices support greater aggregate control over public expenditures and help to improve the efficiency of spending (Moretti, Keller and Majercak, 2023[1]).
Expenditure ceilings are a bridge between the government’s high-level fiscal objectives and the detailed appropriations authorised by the legislature as part of the annual budget process.
Fiscal objectives and fiscal rules
Fiscal rules are numerical parameters set by the government to permanently constrain budgetary aggregates, usually based in legislation. This is distinct from a “fiscal objective”, which is a target that is not legally binding but mandated through political decision or established custom and practice. Fiscal objectives and fiscal rules are usually set as a share of GDP.
Expenditure ceilings
Expenditure ceilings are an upper limit on the level of government expenditure, established for a multi-annual period and/or the annual budget. Ceilings can be set for different aggregates in the budget including totals for all expenditures, for each ministry portfolio or for important spending categories. Expenditure ceilings are commonly issued in nominal terms, though some countries set multi-year ceilings in real terms and translate these into nominal terms for the purposes of annual budget planning.
Budget estimates and appropriations
Budget estimates are the government’s detailed budget proposal to the legislature. Appropriation refers to an authorisation made by law or legislative enactment directing payment out of government funds, under specified conditions or for specific purposes. They usually present information on which administrative unit will spend the resources (e.g. ministry of education), for what programme (e.g. early child development) and using what economic type of spending (e.g. grants to schools).
Most OECD countries establish the budget ceiling before settling detailed allocations for specific programmes. A smaller set of OECD countries has successfully established medium-term expenditure ceilings as the basis of forward-looking spending controls – creating what is commonly referred to as a medium-term budget framework. These countries have found different ways to maintain the credibility of medium-term budget ceilings while allowing sufficient flexibility to accommodate changes in political priorities and the economy.
4.2.1. Early examples of successful top-down budget reforms
A number of OECD countries established top-down expenditure ceilings in the 1980s and 1990. Netherlands, Sweden and the United Kingdom were among the early adopters. Each established a top-down approach to budgeting in the wake of the recession of the early 1990s, and each moved to extend those ceilings over a medium-term horizon.
The Netherlands (1994) established top-down expenditure ceilings linked to the coalition agreements of incoming governments, setting binding spending limits for the full parliamentary term. These followed the adoption of fiscal rules as part of the country’s entry to the Economic and Monetary Union.
Sweden (1996) introduced a two-stage budget process and three-year budget ceilings for aggregate spending that frame the estimates of detailed appropriations. In 2000, this was supplemented by a fiscal rule requiring the government to run a budget surplus over the business cycle.
The United Kingdom (1992) switched from starting the budget process with bidding letters from departments to a cabinet decision on budget aggregates. Initially, the government set a binding three-year ceiling for aggregate expenditure, before changing in 1998 to a system of Spending Reviews, which allocated line ministry budgets for a multi-year period.
4.2.2. Top-down budgeting is a widespread practice among OECD countries
Since the 1990s, nearly all OECD countries have introduced top-down expenditure ceilings. In the 2023 OECD SBO Survey on Budget Frameworks, 33 out of 36 countries reported issuing top-down expenditure ceilings to guide budget formulation; and 26 did so in the context of a multi-year budget framework (Figure 4.1). Among these countries there are numerous examples of effective top-down budgeting practices:
Ceilings for the annual budget negotiations: This is the practice in Chile which has held strong central control over fiscal policy and the budget since the early 2000s, but has maintained an annual budget without official forward estimates.
Indicative ceilings for outer years: This is how the budget controls work in New Zealand where each annual budget cycle agrees figures for the annual budget and for the following four years. When a new annual cycle commences, the baseline figures for that year are carried over from the previous cycle provide a starting point for the negotiations over the next budget.
Binding medium-term expenditure ceilings: Budget ceilings are set for a multi-year period, as they are in Finland and in the cases of the Netherlands, Sweden and the United Kingdom noted above. The duration and flexibility of expenditure ceilings varies, but the expectation is that budget ceilings will hold firm for more than one year at a time.
Number of OECD countries that issue expenditure ceilings covering different time periods, 2023
Note: Australia, Belgium and Italy do not use top-down expenditure ceilings as part of their budget systems. Australia and Belgium publish multi-annual expenditure forecasts that are not ceilings in nature. In Italy, legislation authorises the use of expenditure ceilings, but they have not yet been used in practice. Chile, Norway and Spain use binding top-down expenditure ceilings for the upcoming budget year, while for Costa Rica, Japan, Korea and Luxembourg they are solely indicative. Data for Lithuania and Mexico are not available.
Source: OECD (2023), Senior Budget Officials Survey on Budget Frameworks, Questions 16 and 17.
4.2.3. Strengthening top-down ceilings in the face of rising spending pressures
Despite continued improvements, many OECD countries face challenges embedding effective top-down expenditure controls, particularly where these are expected to limit spending over a medium-term horizon. Low economic growth and rising spending pressures are requiring difficult decisions to constrain the growth of public expenditure and maintain fiscal sustainability.
Stable fiscal objectives and rules are a precondition for firm expenditure ceilings, but a strong system of top-down expenditure ceilings will be necessary to translate fiscal policy into limits for the budget. It also allows countries to reduce the number of fiscal rules and simplify the fiscal framework by allowing more decisions to be made and communicated through the medium-term budget (Moretti, Keller and Majercak, 2023[1]).
Countries must embed expenditure ceilings in a form that will work in their specific context, but reforms to strengthen top-down expenditure ceilings can be guided by some common considerations:
Decisions on the overall level of expenditure for the next budget should be made early in the budget process, as part of a top-down budget process that guides the allocation of spending from overall limits to organisational and programme level budgets.
Ceilings should be established for the medium-term and credibly balance the need to:
Support accountability by aligning medium-term expenditure ceilings to the government’s term in office and allocating ceilings for both aggregate expenditure and to line ministries.
Provide flexibility, guided by clear rules, to allow the government to respond to changing policy priorities and to manage spending that cannot be controlled through appropriations.
Ceilings should be aligned with expenditure baselines, allowing both to be monitored regularly throughout the financial year in partnership with line ministries.
An effective top-down budget process guides the government’s decisions from aggregate fiscal constraints to the detailed appropriations that are approved by the legislature. By setting expenditure ceilings that are consistent with its clear fiscal objectives, the government agrees on how much it is willing to spend, from the “top down”. Budget negotiations determine how to allocate resources within the expenditure ceilings to different priorities, drawing on updated baselines and new spending proposals developed by line ministries from the “bottom-up”.
In principle, this means that expenditure ceilings should be set at the start of the annual budget preparation process prior to the consideration of “bottom-up” spending requests from line ministries. Once set, the ceiling should not be varied during the budget preparation process to ensure the credibility of the overall framework and to ensure that expenditure is in line with the government’s fiscal objectives.
This approach to top-down budgeting requires a multi-stage decision-making process (Figure 4.2). The first stage updates the economic assumptions and expenditure baselines to assess the sustainability of current policy commitments. The second stage is used to agree the budget strategy with Cabinet, including broad goals for tax, spending and borrowing. The third stage is used to negotiate the allocation for each vote and translate this into the detailed budget estimates.
For most countries this process runs each year as part of the annual cycle of budgeting. In a few countries, such as Finland, France, the Netherlands and the United Kingdom, there are separate processes that determine medium-term expenditure ceilings and annual appropriations. There can be overlaps between the stages, as in Germany, where tight fiscal rules allow for a distinct form of top-down budgeting with political discussions over the budget priorities starting from the outset (OECD, 2015[2]). Nonetheless, multi-stage budget negotiations are used in a wide range of settings and generally follow these three stages.
Source: OECD.
4.3.1. Stage 1: Updating the budget framework
The first stage of the budget process updates the planning assumptions for the budget using revised macroeconomic forecasts, tax forecasts and expenditure baselines. This offers an assessment of how the government is expected to perform against its fiscal objectives, under the assumption that there are no policy changes. It identifies room for new tax and spending measures, or if there is a need for fiscal consolidation.
During this period the CBA will gather information on emerging spending pressures and develop options for making budget savings. Ministerial discussions will start to shape the budget strategy. Denmark and Germany for instance use this period to identify topics for spending reviews.
4.3.2. Stage 2: Setting the envelope
The second stage starts the process of top-down budgeting. It sets the government’s budget strategy, including targets for revenue, expenditure, the budget balance and borrowing. This translates the government’s fiscal objectives into expenditure ceilings that can be compared to figures in the budget.
These ceilings are set for total spending and sometimes a few important categories of expenditure, like wages and salaries or capital investment. This provides the overall envelope for budget negotiations that is consistent with the government’s strategy.
The budget strategy and aggregate expenditure ceilings are usually agreed by the Cabinet. In some countries ceilings are also presented to parliament. This establishes collective agreement over the goals for the budget and gives the finance ministry a clear mandate for negotiations.
4.3.3. Stage 3: Negotiating vote-level ceilings and estimates
The third stage is used to negotiate spending priorities, firstly to define expenditure ceilings for line ministries and then to prepare detailed budget estimates.
Once aggregate expenditure ceilings are decided, the CBA will show how the ceilings relate to expenditure baselines to identify the space between the two figures. It will also articulate the process and requirements to consider new spending proposals. These processes and requirements are refined each year to reflect the government’s objectives for the budget and to respond to feedback from previous budget cycles.
Negotiations are usually held between ministers, who are supported by the CBA and the relevant line ministries as they consider the proposals. Ministry or programme level expenditure ceilings will be approved by Cabinet and may be submitted to parliament. The final stages of budget preparation are used to agree detailed budget estimates and prepare the appropriation or budget bill for submission to parliament. In some OECD countries, parliament will propose further amendments to the appropriations which will be factored into the final appropriation or budget act.
A well-designed top-down budget process reinforces aggregate spending control and helps control budget deficits. However, governments also need to look further ahead in order to tackle emerging spending pressures and help line ministries plan for future policy changes and investments. This means that expenditure ceilings should:
be established for the medium-term.
be set for aggregate expenditure and allocated to ministers.
provide in-built flexibility to accommodate changes in economic assumptions and policy priorities.
As with all areas of spending control, effective medium term expenditure ceilings will require trade-offs between these features. On the one hand, they must specify clear and stable spending limits, while on the other hand, they must not become so rigid that spending policy becomes unresponsive to changing conditions or priorities. Each country has its own approach to managing these tensions (Figure 4.3).
Note: T1 is the upcoming budget year. The table only refers to the 26 OECD countries with medium-term top-down expenditure ceilings and shows how long binding or indicative ceilings apply and at what level (granularity). Data for Portugal on the granularity level of the medium-term indicative expenditure ceilings are not available. Data for Lithuania and Mexico are not available.
Explanations on country-specific details (letters a to q):
(a) Other for Colombia: ceilings by sector (e.g. education, health, defence). (b) Other for Denmark: ceilings by sector (state, municipality, regions) and type (expenditure, transfers). (c) Other for Estonia: top-down expenditure ceilings are set at ministerial level (on current expenditure and investments) and performance area level (on current expenditure). Ministerial level expenditure ceilings include one or several performance areas. (d) Indicative ceilings for Finland: indicative spending limits are also set for T1+3 years on a rolling basis for each administrative branch (ministries). (e) France did not have a multi-year programmatic law at the time of the survey. The most recent law used fixed ceilings over a 5-year time period. (f) Other for Germany: additional restrictions for line ministries in top-down decision are possible. (g) Programme level for Iceland: ceilings are set out for 35 expenditure areas. The ceilings are divided into appropriations by expenditure and economic functions and shown for each line ministry. Those appropriations are then further divided by each minister into agencies and projects. (h) New Zealand uses expenditure ceilings at the appropriation level, complemented by multi-annual indicative allowances at the total expenditure level. These allowances are in nature a fiscal management tool, so at each Budget are either allocated to appropriations or adjusted to ensure fiscal targets are achieved. (i) Other for New Zealand: Appropriations are bound by type (departmental or non-departmental expenses, capital or output expenses). Allowances are operating allowances (for decisions impacting operating revenues and operating expenses) and capital allowances (for decisions impacting capital expenditure). (j) Portugal: data on the granularity level of the medium-term indicative expenditure ceilings are not available. (k) Other for Slovenia: ceilings by group of budget items. (l) Other for Sweden: the framework includes an aggregate expenditure ceiling, combined with frames for the 27 expenditure areas that make up the budget. (m) Total expenditure level for Türkiye: Ceiling at the overall/total expenditure level for all budgetary entities, excluding Regulatory and Supervisory Agencies, based on current expenditure and investments. (n) United Kingdom: top-down expenditure ceilings are set at Spending Reviews (SRs). (o) Programme level for the United Kingdom: not set except in limited circumstances; for example, the 2020 Spending Review provided multi-year funding to build 40 new hospitals. (p) United States: the legislature can provide top-down expenditure ceilings for T1 for Appropriations subcommittee level (called 302(b) allocations). Appropriations enacted by the legislature provide top-down ceilings. The President's budget request provides indicative top-down expenditure ceilings based on the President's policies over the 10-year period. (q) Programme level for the United States: for the legislative branch (binding ceiling for T1): Programs in the jurisdiction of the Appropriations Committees receive top-down ceilings in permanent law, typically on a yearly basis.
Source: OECD (2023), Senior Budget Officials Survey on Budget Frameworks, Questions 16, 17 and 19.
4.4.1. Expenditure ceilings should be established for the medium-term
Expenditure ceilings should be established for the medium-term, on a rolling basis or consistent with the government’s term of office. If the former, decisions by a new government should show changes from the previous one.
The majority of OECD countries (22 out of 36) set top-down expenditure ceilings for a multi-year period.
Rolling ceilings: Most countries establish “rolling” ceilings for the budget year and a further two or three years. Each year, the first year of the ceilings becomes the basis for preparing the new budget, and a new ceiling is agreed for the outermost year. Rolling ceilings can be indicative, meaning that the government intends to revise the ceilings each budget as is the practice in Canada and New Zealand. But they can also be binding, as they are in Sweden where the main focus of budget discussions is establishing the detailed appropriations for the next fiscal year and extending the ceilings for the final year of the budget framework.
Fixed term ceilings: The practice of setting ceilings for a “fixed” period is less common, with only six OECD countries choosing to set limits for a defined period of time. Finland and the Netherlands use ceilings for a fixed period that is aligned with the parliamentary term. In both cases, expenditure ceilings operationalise the plans agreed in the government’s coalition agreement. Fixed-term ceilings in the United Kingdom can be set for varying periods and can be changed by the government through a new Spending Review.
Any system of multi-year top-down expenditure ceilings needs to strike a balance between continuity and accountability in the context of parliamentary terms and changes in government. For this, rolling and fixed ceilings have different advantages. Rolling ceilings that extend beyond the parliamentary term provide greater transparency over the incumbent government’s fiscal plans and can provide continuity over spending in the first year after elections. Fixed-term ceilings can be aligned to the parliamentary term so that the government can be held accountable for them but may not show the full costs of policy decisions taken towards the end of a government’s term in office.
Countries have found ways to manage these tensions. In the Netherlands, expenditure ceilings are established as part of the coalition agreement at the start of a new government and last for the duration of the parliamentary term. When the ceilings have expired at the end of a parliament, negotiations over a new coalition agreement use expenditure baselines as a starting point for agreeing on ceilings for the next parliament, with projections updated by the Bureau for Economic Policy Analysis, an IFI. In Sweden, rolling expenditure ceilings are set for three years. Each year, parliament approves a new ceiling for the outermost year of the budget framework. However, these ceilings can be changed when there is a new government subject to approval by parliament.
Since 2004, Finland has controlled spending through multi-year expenditure ceilings that are fixed for the duration of the parliamentary term, typically four years. Upon forming a new government, coalition partners agree expenditure ceilings and rules governing these limits as part of the coalition agreement. While not legislated, the ceilings are detailed in the Government Programme and guide budget decisions throughout the government’s term.
Only the ceiling for total expenditure is binding for the full parliamentary term. This is set in real terms and covers approximately 80% of central government budgetary expenditure, but excludes items such as unemployment benefits and other expenditures that are sensitive to economic fluctuations along with debt interest payments and some other areas of expenditure and off-budget funds.
The real spending limit is set at the start of the government term in the first annual General Government Fiscal Plan. It is translated into a nominal ceiling for budgeting purposes using economic assumptions prepared by the Economics Department of the Ministry of Finance, which has independent status. These assumptions are updated throughout the budget preparation process.
The aggregate ceiling is distributed to administrative branches, but reallocations between these ceilings and across years is permitted provided they do not increase spending in real terms. The ceilings also include buffers that provide additional flexibility. The total expenditure ceiling is larger than the combined ceilings for the administrative branches. As well as this “margin” the government includes a supplementary budget provision. However, these allocations are usually small relative to the overall expenditure ceiling, with a provision of EUR 0.4 billion provided for supplementary budgets in 2024 compared to a total expenditure ceiling of around EUR 65 billion.
Changes to the expenditure ceilings are usually made through the budget process. These are reported together with the relevant economic assumptions in the general strategy and outlook of the budget proposal. They are also annexed in the General Government Fiscal Plan, showing changes from the original plans agreed at the start of government term along with the latest economic assumptions.
The system of spending rules has evolved with every change of government. This development has been guided by spending limits system working groups that typically prepare a report for each parliament. In 2023, for example, a new dispensation mechanism was created to allow the government to react to significant unforeseen crises.
Source: Ministry of Finance, Finland.
4.4.2. Expenditure ceilings should be allocated to ministers
In most OECD countries medium-term expenditure ceilings are binding on budget totals. However for ceilings to be enforced by the CBA, the total level of annual expenditure should be disaggregated into separate ceilings and allocated to ministers who can be held accountable for spending decisions. Ceilings can also be used to limit or protect important spending categories.
Most OECD countries set expenditure ceilings for expenditure aggregates and by organisation. Some have medium-term ceilings only for aggregate spending (e.g. Switzerland) or organisational-level spending (e.g. Estonia). Slovenia combines annual ministry-level ceilings with two-year aggregate ceilings and three-year programme ceilings (mostly for capital investment). However, the most common approach in OECD countries is to have both aggregate and ministry ceilings running over the same period.
In addition, many countries disaggregate expenditure ceilings to guide allocations for important spending categories (Table 4.1). These sub-ceilings may be used to limit certain expenditures, as is often the case for administration or staffing costs. Alternatively, ceilings can help protect spending or even allocate more resources to a particular activity, such as public investment. Other categories of expenditure, such as debt interest, will be chosen because of how they are controlled.
Governments tend to allow some flexibility to adjust organisational ceilings and category-specific ceilings, provided these changes remain consistent with ceilings for total expenditure. A decision to increase the ceiling for one ministry will typically require a corresponding decrease in another, which reinforces the collective commitment in government to keep spending within ministry-level ceilings.
Note: Australia, Belgium and Italy do not use top-down expenditure ceilings as part of their budget systems. Data for Lithuania and Mexico are not available. “Other” includes, but is not limited to, sub-ceilings by type (e.g. transfers), sectors (state, region, municipality), topic (e.g. social security, Covid recovery, climate response) and specific limitations per line ministry.
Source: OECD (2023), Senior Budget Officials Survey on Budget Frameworks, Question 22.
4.4.3. Ceilings should provide a degree of in-built flexibility
As expenditure ceilings are intended to enhance the control of aggregate expenditure they should, in principle, have broad coverage of the government’s fiscal activities. Nearly all countries set expenditure ceilings in nominal terms, which offers a more transparent link to the annual budget than ceilings which are established in real terms and then converted each year into a nominal limit for budget purposes.
However, some types of expenditure cannot be easily controlled through a firm nominal limit on spending. Equally, budget ceilings set in nominal terms can become disconnected from the real value of expenditure if inflation is high or volatile. A medium-term perspective also increases the likelihood that there will be changes to the government’s policy agenda and other planning assumptions. These factors require consideration for how spending ceilings can be credible within a country context. More specifically, how can they bind expenditure plans within the term of a government, while allowing some flexibility to adjust to changing economic assumptions or spending pressures.
Certain expenditure ceilings may need to be deemed flexible in nature or controlled separately. This is especially applicable to cyclical areas of expenditure, such as unemployment benefits, but it can also be true for non-cyclical benefits. Interest expenditure may be similarly treated. These are examples of expenditures that are difficult to forecast accurately and are harder to control through the budget. If interest payments on government debt increase, for example, they still need to be paid. Similarly, setting a cash limit social security does not take away the obligation to pay entitlements once the spending limit is exceeded. Asking ministries to cover an overspend created by those entitlements may also undermine other programmes in the core operating budget.
The implication is that countries will either need to:
Exclude some categories of spending from expenditure ceilings;
Provide flexibility to revise ceilings when there are changes to specific budget lines; or
Use a margin or contingency reserve to accommodate forecast errors.
Some OECD countries have created specific categories of spending that are treated differently with respect to expenditure ceilings. Examples include non-vote expenditure in Ireland, non-departmental expenditure in New Zealand and annually managed expenditure in the United Kingdom. Health and social security funds may also be controlled separately, as is the case in France (Box 4.3).
In France, budget ceilings are guided by multi-year expenditure rules and controlled through separate budgets for State expenditures and compulsory social security.
The Law on the Programming of Public Finances (Loi programmation des finances publiques, LPFP) is prepared every 2-3 years and sets the budget balance objective for general government. The LPFP also establishes ceilings for the State budget missions for three years, and spending targets in the health and social sphere. These guide the annual allocation of public expenditures through two separate budgets: the state budget (the projet de loi de finances, or PLF) and the social security budget (project de loi de finances de la sécurité sociale, or PLFSS).
The social security budget is different from the State budget in two main ways. Firstly, the social security budget is limited mainly to the basic compulsory social schemes (ROBSS) and the old-age solidarity fund (FSV), while other social security expenditure, including unemployment insurance and supplementary pensions are included in the State budget. Secondly, while the State budget sets limits for expenditures (“limitative” authorisations) with a few important exceptions such as debt interest, the social security budget has the status of a forecast or objective (“evaluative authorisations”).
Successive governments have sought to align the content and timing of these laws to strengthen control over social security expenditures and support parliamentary oversight. As a result of these changes, the draft PLF and PLFSS are submitted to parliament together in October and approved in December. These are followed, after the close of the financial year, by two laws for approving the State and social security accounts. Since 2022, social security policy evaluation reports have been annexed to the social security accounts to inform parliamentary debate.
Source: Moretti and Kraan (2018[3]).
Another well-used approach to build flexibility into expenditure ceilings is to establish an unallocated reserve, or “margin”, to meet unforeseen and exceptional contingencies while ensuring the integrity of the high-level fiscal objectives. Budget margins and unallocated reserves are used in many countries.
Budget margins create a buffer between planned expenditures and the ceiling. This may be created by using conservative estimates or by setting organisational ceilings that add to less than the limit for aggregate expenditure. A budget margin is not approved separately from the ceilings and is not appropriated. Ireland, Finland and Sweden employ budget margins to guard against overspends from cyclical and demand-led spending or to provide space in outer years of the budget framework for budgetary decisions in relation to new priorities at that time (Box 4.4).
Contingency reserves establish an unallocated provision in the budget, usually intended to meet unforeseen (or unforeseeable) expenditures. Once a claim on the reserve is authorised, the expenditure will be allocated to the corresponding votes and budget lines through a supplementary budget. In 2023, 16 OECD countries reported having a reserve for the budget year as part of their top-down expenditure ceilings, including Iceland, Norway and Spain.
The design and application of a budget margin and a contingency reserve should be transparent and guided by clear criteria in order to prevent misuse or overspending. The main challenge with using a budget margin or contingency reserve within expenditure ceilings is that it should be large enough to absorb potential additional spending needs or the realisation of some fiscal risks, but small enough to avoid diluting fiscal discipline and unduly limiting spending in priority areas.
The purpose of budget margins or contingency reserves can be defined in a way that limits new spending requests from budgetary entities. Another approach is to link reserve to specific areas of funding like social assistance, as is the practice in Denmark where the ceiling for transfers has a contingency margin of 1%. Under any of these approaches, clear criteria for a reserve claim and effective controls by the finance ministry are usually required.
Fiscal policy in Sweden is guided by a surplus target for general government and a debt anchor for consolidated gross debt. These fiscal rules are translated into a ceiling for central government expenditure covering the next budget and a further two financial years (utgiftstak).
Once approved by parliament (the Riksdag) the ceiling will only be reviewed if there is a change in government or a significant economic shock. That means that in most years the government will maintain the ceilings agreed in the previous budgets, and will propose a new ceiling only for the third year of the budget framework.
The central government ceiling includes most expenditure. All 27 Expenditure Areas that make up the budget are included, except for debt interest payments. The ceiling also includes expenditure on pensions which are financed from a pension fund instead of an annual appropriation. Together these constitute “ceiling-restricted expenditure”.
The ceiling is set in nominal terms. To provide some flexibility for the government to accommodate changes in economic assumptions and unforeseen events, the overall ceiling includes a “margin” between the total expenditure ceiling and the total ceiling-restricted expenditure. This provides a buffer that allows budget changes without reopening the approved expenditure ceiling. The overall ceiling is disaggregated into a frame for each Expenditure Area to guide the preparation and approval of the next budget. These frames provide an indicative allocation for line ministries for the following two years , but unlike the overall ceiling, these allocations are indicative and may be revised through subsequent budget bills.
Source: Ministry of Finance, Sweden.
Ministers should have some flexibility and autonomy to reallocate resources within their expenditure ceilings. The CBA should prepare guidance on how the flexibility is to apply, for instance to reallocate resources between different categories of expenditure (e.g. personnel and other operating expenditure). The CBA should have the capacity to assess such reallocations and ensure that decisions by line ministers are within the overall policy and budget framework.
4.5.1. Delegating budget flexibility within clear rules
Governments are increasingly delegating greater authority for budget management to line ministries. This has been observable through the reduction in the number of line items included in the budget along with the spread of programme budgeting. In the 2023 OECD Senior Budget Officials Survey on Budget Frameworks, 13 out of 36 countries reported using “lumpsum” appropriations to give line ministries more flexibility over their budget allocations (OECD, 2023[4]). The survey illustrates some of the flexibilities that governments have to amend the budget during the financial year.
Two further types of flexibility for line ministries to change their budgets include (Figure 4.4):
Changes between budget lines. Budget reallocations are often needed to support the implementation of the government’s programmes. 32 out of 36 OECD countries permit reallocations between some or all appropriations, with 31 setting a limit on changes or a requirement for approval by the CBA.
Changes across fiscal years. 31 out of 36 OECD countries also allow line ministries to carry-over all or some unspent appropriations from one year to the next, subject to agreed limits and approvals for most countries. Such flexibilities are usually offered for spending on multi-year commitments.
In some countries these flexibilities can be applied without returning to parliament. In other countries, the changes need to be authorised by parliament.
Number of OECD countries, 2023
Note: Formulation of ‘Utilised (with limit and/or approval of CBA)” was done by accounting survey responses which had at least one indication of either a ‘threshold’ or an authorising body requiring approval (e.g. CBA) on the usage of Flexibility Measures. This includes the survey response "Other". Data for Lithuania and Mexico are not available. Hungry and Italy practice carry-overs, but no data is available for possible authorisation requirements.
Source: OECD (2023), OECD Senior Budget Officials Survey on Budget Frameworks, Questions 39.1 and 39.2.
These flexibilities can be managed through a system of delegated limits for budget amendments. These may be set by the CBA for line ministries, but also for budget holders within the line ministry. Changes for small amounts within a programme can be authorised by the programme manager or the finance department in the line ministry. Larger changes or virements between spending categories require authorisation from the ministers responsible and the finance minister. Under some systems, changes made by line ministries under delegated authorities are aggregated and reported to the cabinet and included in the supplementary estimates for approval by parliament.
An effective system of delegated limits avoids inefficient decision-making from a constant flow of minor amendments, while also guarding against the risk that changes will have a material impact on the implementation of the government’s policy agenda and legislative accountability.
4.5.2. Monitoring budget execution
Robust monitoring and enforcement mechanisms need to be in place to ensure that actual spending is in line with the expenditure ceilings. This is typically a responsibility shared between the CBA and the finance function in the line ministry. Other entities with an interest in budget execution include IFIs with a mandate to independently monitor the government’s adherence to its expenditure ceilings.
The main role of the CBA is to establish a consolidated view of budget execution to manage treasury functions and oversee budget performance for the government as a whole. It will also monitor spending under each portfolio in order to advise the government where decisions are needed in order to adjust policies or reallocate spending in order to keep expenditures within the agreed ceilings.
Germany: Budget officers in the Federal Ministry of Finance gather information from line ministries that is weighed together with various high-level and strategic priorities, before the Federal Ministry of Finance prepares its proposal for the aggregate budget allocations for each ministry. Mirror units have the ability to monitor expenditure line-by-line through a centralised ICT system. They do not generally intervene in budget execution unless the Ministry has concerns about budget expenditure or execution.
Netherlands: Budget officers must ensure that changes to the budget fall within the permitted rules. Forward estimates of the expenditure baseline are maintained in the government’s ICT system for budget preparation. Line ministries submit their changes through the system, including a categorisation of the nature of the change. The Budget Inspectorate has to approve all changes upon each update.
United Kingdom: The role of monitoring expenditure ceilings is divided between a “spending team” and the General Expenditure Policy Team which maintains a broad watch on the public finances. The monitoring of expenditure ceilings is largely based on updated baselines submitted by line ministries to the Treasury ICT system. The General Expenditure Policy Team undertakes most of the analysis to check if spending forecasts up to the end of the year are in line with the control total.
This illustrates how the combination of medium-term baselines and top-down expenditure ceilings changes the approach to budget management. It increases the importance of regular updates of the expenditure baselines and shifts the focus of budget monitoring to be more forward looking and policy oriented.
References
[1] Moretti, D., A. Keller and M. Majercak (2023), “Medium-term and top-down budgeting in OECD countries”, OECD Journal on Budgeting, https://doi.org/10.1787/39425570-en.
[3] Moretti, D. and D. Kraan (2018), “Budgeting in France”, OECD Journal on Budgeting, Vol. 2018/2, https://doi.org/10.1787/budget-v18-2-en.
[4] OECD (2023), “2023 OECD Senior Budget Officials Survey on Budget Frameworks”, https://www.oecd.org/en/data/datasets/public-finance-and-budgets-database.html (accessed on 4 August 2025).
[2] OECD (2015), “Budget Review: Germany”, OECD Journal on Budgeting, Vol. 2014/2, https://doi.org/10.1787/budget-14-5jrw4sxb32q4.