Use of tax expenditures in budgeting

3. Findings from case study analyses

This section analyses the information from the four case studies in the context of the key elements of a tax expenditure system for fiscal and budget control, as discussed in Section 2.

3.1. Definition and measurement of tax expenditure figures in budget

All four case study countries use the revenue foregone approach to measure tax expenditures, but formal definitions of tax expenditure vary from country to country. While the Netherlands used to have a specific definition, they abolished this following a government study which concluded that as there is no international consensus on the definition of tax expenditure and it is preferable to use a broad guideline. Generally, the guideline states that something is considered a tax expenditure if it has an intended policy objective that ultimately results in an intended loss of government revenue.

The United Kingdom takes a similar approach. HM Revenue and Customs (HMRC) distinguishes between non-structural and structural tax reliefs, where the former are integral to the functioning of the tax system and the latter actively designed to advance economic or social objectives (i.e. tax expenditures). However, beyond this, it does not disclose how they decide under which category of tax reliefs fit, instead stating that their categorisations remain under constant review.

On the other hand, Canada and Ireland identify tax reliefs in relation to a benchmark. Canada defines its benchmark by including the unit of taxation, the taxation period, the tax base and the tax rate structure, as well as more specific references such as the treatment of inflation and treatment of government and governmental entities. Ireland’s basis for its benchmark was established by a 2009 Commission on Taxation, and includes measures that are inherent in the design of the tax system, measures that are tax neutral, and measures related to the state, among others.

3.1.1. Defining a benchmark

Benchmarking decisions can be controversial within countries. For example, in the Netherlands, a change in the benchmark definition was made in 2018 due to a decision that all provisions triggering revenue foregone should be considered not to be part of the benchmark, even if they did not meet the previous definition of tax expenditure. In Ireland, two government departments produced different lists of tax expenditures based on different benchmark criteria, and only consolidated the lists into a master list in 2023 (Aliu and Redonda, 2024[12]).

Canada defines the tax base for goods and services tax (GST) as comprising all goods and services consumed in Canada, meaning the tax applies on a destination basis. As such, any exemptions from GST or zero rating (where GST is not charged) are counted as tax expenditures (Department of Finance - Canada, 2024[21]). This same logic is used in most OECD Member countries, although some consider lower VAT rates for food as part of the benchmark system on the basis that they are redistributive measures, where redistribution is a core goal of the tax system. As such, they are not reported as tax expenditure measures.

This variation in definitions of tax expenditures and their tax benchmarks makes international comparability of tax expenditures challenging. This in turn can make it difficult for governments to understand how much they spend on tax expenditures in a budget relative to other countries.

3.2. Treatment of tax expenditures vs direct expenditures in the budget

3.2.1. Development and management of tax expenditures

Generally, in the case study countries, it is ministries of finance, or equivalent organisations, which have responsibility for the development and management of tax expenditures. This is the case in the United Kingdom, where HM Treasury is responsible for strategic oversight of the tax system. In Canada, the Department of Finance is responsible for developing tax policy, and is thus the most relevant actor for tax expenditures, while in Ireland and the Netherlands, the Department of Finance and Ministry of Finance respectively serve as the main tax expenditure focal points. This differs from direct expenditure, where development and management usually lies with the relevant line ministry.

Ministries of finance do not generally develop tax expenditure policies in isolation. This is most noticeable in the Netherlands, where government budget guidance requires that tax expenditure provisions be reported in the budget of the relevant ministry, and as such fall primarily under their responsibility. In the United Kingdom, HMRC and HM Treasury jointly have that primary responsibility, but often collaborate with relevant departments on tax expenditure policy design and implementation. This collaborative approach exists but is less common in Canada and Ireland, where the respective Departments of Finance are primarily responsible for the design, costing and reporting of tax expenditures, although other departments do assist in the design of measures, for example when expertise on a particular industry or activity is required.

In most cases, the design of tax policy and its administration is managed by separate organisations. In the United Kingdom, HMRC is responsible for tax administration. While both HM Treasury and HMRC conduct analysis on the impact of changes to tax expenditures, HMRC is responsible for implementation and maintenance of these expenditures. In Canada, tax administration is the responsibility of the Canada Revenue Agency, which includes advising the Department of Finance on issues during implementation, and reporting any issues with compliance. In Ireland, the Revenue Commissioners (more frequently referred to simply as ‘Revenue’) are responsible for taxation matters, which includes publishing statistics on expenditures and providing analysis of particular tax expenditures. As above, the Netherlands is an exception, with administration split between the Ministry of Finance (which includes the Directorate-General for Tax and Customs Administration) and the line ministries.

3.2.2. Incorporation of tax expenditures into the budget process and budget totals

The treatment of tax expenditures within the budget varies significantly across the case study countries relative to the treatment of direct expenditures. In some cases, tax expenditures are documented separately from the budget, while in others they are integrated into budget documentation, but are not subject to the same approval procedures as direct expenditure. Generally, even where tax expenditures are considered in budgets, they are part of overall tax figures or as part of revenue projections. Dedicated line items showing how tax expenditures contribute to budget totals are rare.

Some countries do not specify tax expenditures in the budget, while in other cases, tax expenditures are considered in the budget totals, but are not explicitly labelled as tax expenditures. For example, in Ireland, the government publishes tax policy changes alongside the budget, and includes tax expenditures within them.

In the Netherlands and in the United Kingdom, tax expenditures are considered both alongside tax expenditure policy and as part of the budget. In the Netherlands, changes to tax expenditures are proposed through the Tax Plan, which is presented alongside the budget. The Dutch budget also includes two annexes highlighting which department is responsible for each tax expenditure and estimated costs. In the United Kingdom, tax expenditures in the budget include both costings and a description of purpose. HM treasury has also published policy costing methodologies for tax expenditure measures in each budget since 2011, which include information on the level of detail and behavioural assumptions behind estimated costs.

Most of the four case study countries do not approve tax expenditures through the same budget processes that they approve appropriated expenditure, even when they include information on them in budget documentation. In general, tax expenditures are approved through the standard processes used to introduce new legislation, as opposed to the processes within the budget cycle. In Canada, direct expenditure is authorised through appropriations and statutes, while tax expenditure is authorised through changes to tax laws. In the United Kingdom, the Finance Bill, which includes tax expenditures, is treated like any other Bill – i.e. a proposal for a new law. This means it receives multiple readings and can be amended, unlike the Supply and Appropriations Bill (used for authorising direct expenditure), for which no amendments are allowed. The Netherlands departs from this approach, as there is no significant difference in Parliamentary process between the budget and standard legislation. For both the budget and the Tax Plan, only the House of Representatives can propose and adopt amendments, while the Senate can only either accept or outright reject.

Some countries also set requirements on tax expenditures separate from the budget process, which are not required of direct expenditure. For example, in the Netherlands, a framework on the Regulation on National Budgeting Instructions requires that tax expenditures answer eight questions, which ensure that the purpose of the tax expenditure is clearly defined and the need for intervention with an expenditure (as opposed to a subsidy) is clearly explained. For measures with impact over EUR 20 million, the appraisal must be included in the bill presented to parliament. In the United Kingdom, HMRC has established mandatory guidance on tax reliefs, and a framework for recording information on those valued at over GBP 40 million per annum.

Table 1. Overview of the treatment of tax expenditure vs direct expenditure in the budget process

Country	Type of Expenditure	Costs published in budget	Tax expenditures explicitly labelled	Same approval process as direct expenditure	Committee oversight	Regular scrutiny after enactment
Canada	Direct	✓	n/a	n/a	✓	✓
	Tax	✓	✓	✗	✓	✓
Ireland	Direct	✓	n/a	n/a	✓	✓
	Tax	✗	✗	✗	✗	✗
Netherlands	Direct	✓	n/a	n/a	✓	✓
	Tax	✓	✓	✓	✓	✓
United Kingdom	Direct	✓	n/a	n/a	✓	✓
	Tax	✓	✓	✗	✓	✗

Source: OECD authors’ elaboration

3.2.3. Forecasting of tax expenditures vs direct expenditures

Tax expenditures are generally not treated as equivalent to direct expenditures in forecasting practices in the budget. Instead, they are incorporated into forecasts through lower tax revenue predictions, with formal data on tax expenditure levels provided in separate reports.

In Canada, forecasts are published as part of annual budgets (as well as fiscal updates). The starting point is an economic forecast, which forecasts GDP, inflation, employment, and other macroeconomic indicators. These economic assumptions, alongside any new budget proposals, are then used to forecast fiscal impacts, including predictions of changes to tax revenues. Again, tax expenditures are implicitly included in these forecasts, but as part of the tax revenue change predictions. However, a separate document on tax expenditures (the Report on Federal Tax Expenditures) is published alongside the Main estimates (the government’s detailed spending plan for the fiscal year, tabled in Parliament to provide detail on the spending authorities sought). This document provides forecasts for one year beyond the current budget year, as opposed to the budget year and the next four years, as seen in the budget.

In Ireland, the Department of Finance publishes a White Paper ahead of the budget, which gives estimates of receipts and expenditure for the coming year, based on macroeconomic forecasts. These projections then feed into the annual budget package. Tax expenditures are considered in these forecasts through lower tax revenues but are not explicitly identified as such. Instead, a tax expenditure report is submitted alongside the budget.

In the Netherlands, the national budget process is anchored by a multi-year expenditure framework, and a more flexible revenue framework, both of which are set at the start of a government’s term. The Netherlands Bureau for Economic Analysis (CPB) produces economic forecasts and macroeconomic outlooks. While tax expenditures are tracked, they are not treated as public spending, instead being highlighted through lower projected revenues. An annex to the budget provides detail on these tax expenditures, including forecasts, and shows the largest changes to tax expenditures, and distinguishes between policy changes and endogenous effects, highlighting that some of the cost from tax expenditures come from economic or demographic changes rather than new legislation.

In the United Kingdom, the OBR prepares five-year forecasts on the economy and public finances twice a year, supporting fiscal control through oversight. Tax expenditures are set out in the revenue forecast, not the expenditure forecast. The OBR also decides whether it believes a tax policy will impact GDP or other macroeconomic variables, and if so, how to incorporate it into its economic forecast. In both these cases, tax expenditures do not have to be explicitly identified. Instead, a separate tax reliefs publication includes forecast estimates for the following tax year. From 2024 onwards, the publication also compared estimated outturns to the forecasts published in the previous year’s publication.

The practice of explicitly reporting tax expenditures in dedicated reports but not in the budget highlights the importance of these tax expenditure reports (discussed in greater depth in “Reporting on tax expenditures in budgeting”). As tax expenditure reports are likely to receive less attention than the main budget and related forecasts, this practice also risks reducing the visibility of tax expenditures and their impact on the overall budget.

3.2.4. Treatment of refundable tax credits

In many cases, refundable tax credits are treated as expenditure. The Canada Revenue Agency is responsible for delivering tax credits, and they are reported in the Authority’s finances. However, in the budget process, they are authorised through changes to tax laws, not appropriations or statutes as in the case of direct expenditure. In the Netherlands, the largest refundable tax credits, known as the General Tax Credit and the employed persons tax credit, are both included in the Budget Memorandum with cost change estimates (Rijksoverheid, 2025[22]). In the United Kingdom, refundable tax credits are recorded as Annually Managed Expenditure – the portion of expenditure that is demand led, which includes welfare. While there is no cap for Annually Managed Expenditure, they receive greater budgetary scrutiny as a result of this definition. Similarly, in Canada, refundable tax credits are treated as expenditure, there are only a handful of them.

Ireland is an exception to this trend, as refundable tax credits are generally still treated as tax expenditures. This is evident from the Research and Development tax credit, which can either be refunded as cash or offset against a company’s tax liabilities, but is recorded as a tax expenditure.

Box 2. Refundable tax credits in Korea

The Earned Income Tax Credit was introduced in 2008 to provide low-income earners within income support, and to encourage their participation in the labour market. It has both income and wealth eligibility requirements. For a single person, the income threshold is KRW 22 million (USD 15 000). When the value of the earned income tax credit exceeds the income tax, it pays the difference in cash. As such, it can be considered a refundable tax credit.

The Child Care Tax Credit was introduced in 2014 to support the childcare expenses of low-income earners. The sum of annual income of the applicant and spouse must be less than KRW 40 million (USD 27 000). As with the Earned Income Tax Credit, when the value exceeds income tax owed, the difference is paid in cash.

The 2024 Korean tax expenditure plan emphasises continued provision of these credits to stabilise household finances and promote social mobility, ensuring that tax burdens do not disproportionately affect vulnerable populations. Both credits are subject to periodic evaluations, including preliminary feasibility assessments and in-depth reviews, to assess their policy effectiveness, fiscal impact, and alignment with economic and social objectives. Both tax credits are managed by the National Tax Service.

Sources: (OECD, 2024[23]; Ministry of Economy and Finance, 2024[24])

3.3. Fiscal rules and objectives related to tax expenditures

It is rare for countries to have fiscal rules or objectives exclusively for tax expenditure. However, in some cases, countries have rules which encompass tax expenditure alongside direct spending. In the Netherlands, it is standard practice for an incoming government to commit to a set of fiscal rules in a letter to Parliament (although such letters are not legally binding). This includes the development of both a revenue framework and an expenditure framework, with the goal of the former to maintain the tax burden at a steady level. The 2024 version of this commitment highlighted that tax expenditures are included in its revenue framework (Tweede Kamer der Staten-Generaal, 2024[25]). In the United Kingdom, the Welfare Cap is a limit on the amount that government can spend on certain social security benefits and tax credits, although it excludes pensions and payments that vary with the economic cycle (i.e. unemployment benefits). If the cap is breached, this must either be justified, or measures must be brought forward to reduce welfare spending.

In Canada, Budget 2025 established two fiscal anchors. The first aims to balance day-to-day operating spending with revenues by 2028–29, shifting spending toward investments that grow the economy. The second maintains a declining deficit-to-GDP ratio to ensure disciplined fiscal management for future generations. In Ireland, there is a 5% National Spending Rule, which aims to limit core net spending growth to 5% each year. This accounts both for spending and tax cuts, although it is not legally binding. In 2025, United Kingdom government had three fiscal rules. One is the welfare cap, while the other two state that the current budget should be on course to be in balance or surplus, and that net financial debt should fall as a share of the economy, both by 2030.

Such deficit rules can be argued to implicitly account for tax expenditures, as, all other things being equal, a reduction in tax revenue will of course worsen the deficit. However, such rules are not necessarily effective. If not properly regulated, governments are able to expand tax expenditures in periods of high growth without negatively impacting the budget surplus, while in periods of low or negative growth ministries can make a political argument that such tax expenditures should be kept to support economic recovery.

In the Netherlands, certain tax expenditures have a set annual budget, meaning their spending must be monitored during the year and halted when they reach their limit. For example, the Environmental Investment Deduction, a corporate tax incentive for businesses investing in environmentally friendly assets, has a maximum budget of EUR 25 million for 2025. Once this ceiling is reached, no further deductions can be granted. At the time of writing, this ceiling has never been reached. However, such a limit does not exist for most tax expenditures.

3.3.1. Sunset clauses

The use of sunset clauses is relatively rare. In Ireland, it is the responsibility of the Department of Finance to decide whether a tax expenditure should be further extended. In the Netherlands, the assessment framework for new tax schemes requires that new tax expenditures have sunset clauses, although in reality deviations from this framework can apply. There are also some examples in the United Kingdom, although they too are rare.

However, sunset clauses are not necessarily effective in and of themselves. One reason for this is that just because a tax expenditure has to be reviewed does not mean it will be removed. Indeed, without other measures in place, a sunset clause can simply lead to repeated renewal of a tax expenditure. The effectiveness of sunset clauses also presupposes that effective evaluation systems are in place. However, if evaluations are partial, rushed, or based on insufficient data, they may not be able to determine whether or not a tax expenditure should be extended.

3.4. Reporting on tax expenditures in budgeting

Tax expenditure reporting practices are strong across all four case study countries. In Ireland and the Netherlands, this is in part due to EU requirements to publish detailed information on the effect of tax expenditures on revenue, although these requirements do not specify exactly how countries should go about this (Turrini et al., 2024[26]). Canada and the United Kingdom also publish such information.

In Canada, both Budgets and fiscal updates (i.e. updates to the fiscal plan within the budget that often include new spending commitments and initiatives) include new changes to tax expenditures. These are not presented explicitly as tax expenditures, but instead are introduced with less technical terms such as tax credits, deductions or exemptions. An annex to the budget also provides supplementary information on the tax measures included in the budget. Detail. Detail is instead provided in the Annual Report on Federal Tax Expenditures, published and tabled in Parliament by the Department of Finance. This report commenced in 1995, with the aim of providing a source of information to parliamentarians and government officials who wished to analyse Canada’s tax system (Government of Canada, 1995[27]). This report also summarises the main policy changes to tax expenditures, but supplements this with descriptions of all tax expenditure measures.

In 2015, Ireland’s Department of Finance started publishing an annual Report on Tax Expenditure, following on from the government’s 2014-2020 medium-term economic strategy, which stated that tax expenditures should be limited to cases where there is a demonstrable market failure and where direct expenditure would be less efficient (Department of Finance - Ireland, 2014[28]). This report includes estimated costs of the main tax expenditures in Ireland, an overview of the ongoing reviews of tax expenditures, and best practices for the review of tax expenditures. While, as mentioned previously, new tax expenditures are included in a tax policy document published alongside the budget, this annual report offers a comprehensive overview of all such expenditures.

A similar setup can be seen in the United Kingdom, where changes to tax expenditure are included in the Autumn Budget and Spring Statement Changes, but the full costs of tax expenditures are only disclosed in two dedicated tax expenditure reports. These two reports look at structural reliefs (i.e. the benchmark tax system) and non-structural reliefs respectively (i.e. tax expenditures), and make use of statistics on tax reliefs published by HMRC. They commenced in 2017, the same year HM Treasury piloted monitoring of tax expenditures. Greater detail is provided for a subset of tax policies, including estimates on cost and claimants, some limited distributional analysis, and links to relevant research.

Unlike the other three countries, the Netherlands includes all its information on tax expenditures within its budget. This has been the case since 1999, following a period during the 1990s in which Parliament repeatedly requested greater information on tax expenditures that the government was initially not able to provide (Spanjers, 2021[29]).

The timetable on the publication of tax expenditure reports varies, and is not always co-ordinated with the budget process. Canada’s Report on Federal Tax Expenditures is released by 1 March alongside the Main Estimates, which contains the resources required by departments and agencies for the upcoming fiscal year. The fact that its publication coincides with the Main Estimates helps parliament stay informed about changes to tax expenditures and includes all announcements from the previous year. In Ireland, the report on tax expenditures has historically been published in alignment with other budget documentation in September or October. In the United Kingdom, the two reports on tax reliefs are published each year in December, with another report on non-structural reliefs published in May with previously unavailable costings. This is in accordance with the publishing timeline for official statistics, which follow the tax year, while the budget follows the fiscal year. This can mean that the most recent statistics are not always available for scrutiny by parliament.

Table 2. What is included in tax expenditure reports?

Documentation on tax expenditures	Canada	Ireland	Netherlands	United Kingdom
	Report on Federal Tax Expenditures	‘Tax Passport’ published from 2025 alongside Tax Expenditure Report	Annex 4 of Annual Budget	Official statistics on non-structural reliefs. This is for a subset of tax expenditures
Cost projections	✓	✗	✓ (for some)	✓
Description of policy objective	✓	✓	✓	✓
Description of TE’s history	✓	✓	✓	✓
Explanation of estimation method	✓	✗	✓	✗
Explanations of changes in cost/revenue foregone	✓	✗	✓	✓
Information on beneficiaries/distributional analysis	✓	✓	✗	✓
Information on legal reference	✓	✓	✓	✗
Number of claims/beneficiaries	✓	✓	✗	✓ (for some)
Reason to not include measure in the benchmark	✓	✓	✗	✗
Results of recent evaluations	✓	✓	✓	✓
Revenue forgone quantification	✓	✓	✓	✓
Source of data	✓	✓	✓	✗
Type of measure (exemption, deduction, credit)	✓	✗	✗	✗
Type of tax (Corporate, Personal)	✓	✓	✗	✗
Whether a sunset clause is included	✓	✗	✓	✗

Source: OECD authors’ elaboration.

Several countries also include tax expenditure in their distributional and welfare analyses. In Ireland, since 2022, a Beyond GDP report has been published alongside the budget, intended as a complement to the annual Economic and Fiscal Outlook. This includes distributional analysis of the policies introduced in the budget within both the tax and welfare systems, although they do not account for behavioural responses, and do not account for all measures in the Budget. The 2025 version included analysis of the rent tax credit, changes to the earned income tax credit, and several other tax credits. It also analyses the impact of general tax increases. However, tax expenditures are not explicitly highlighted. A chapter in the Tax Policy Changes document also provides distributional analysis, looking at the overall impact of all changes reported.

In the United Kingdom, distributional analysis of both tax and spending changes is published with the Autumn Budget and Spring Statement. Similarly to Ireland, its tax focus is often on credits, with analysis of tax expenditures beyond this limited. In both Ireland and the United Kingdom, distributional analysis focuses on tax measures generally, rather than looking at tax expenditures specifically. This is not the case in Canada, where the tax expenditure report dedicates a chapter to distributional analysis using the Gender-Based Analysis Plus framework, as required by the Canadian Gender Budgeting Act of 2018. This analysis focuses on gender, although also considers other factors including age, family types, income groups and place of residence. Canada’s budget also provides a brief overview of expected beneficiaries (by gender, income, and age), including for tax expenditures. The Netherlands does not incorporate distributional analysis into its tax expenditure reporting.

3.5. Evaluation of tax expenditures

All four case study countries have a system for the evaluation of tax expenditures. In Canada, evaluation is not a requirement for tax expenditures as it is for direct spending. However, the Canadian Department of Finance publishes several evaluations of tax expenditures each year. Since 2020, this has taken the form of extra studies published within the Report on Federal Tax Expenditures, involving additional distributional analysis. The focus topics are set on the basis of priority issues and issues of interest within the Department of Finance.

In Ireland, reviews of tax expenditures are generally conducted by the Department of Finance, with the co-operation of Revenue. In some cases, other government departments and agencies can be involved, and in some cases such a review can incorporate public consultation. These reviews are intended to ensure that tax expenditures are amended or ceased if they no longer serve a purpose, and extended if they do. Evaluations follow a proportionate approach, meaning the costliest tax expenditures are subject to the highest frequency and rigour of review.

The evaluation process in the Netherlands follows an evaluation calendar published annually, which details which tax expenditure is to be evaluated and when. These are conducted by various different people, including the Ministry of Finance, sometimes the Netherlands Bureau for Economic Policy Analysis, and sometimes outsourced, although timely completion of the evaluation is always the responsibility of the relevant ministry. Generally, tax expenditures are subject to evaluation every 4-7 years, although several exceptions and criteria for prioritisation means that this does not always occur.

Since 2021, HMRC, typically responsible for conducting tax expenditure evaluations, has also published a tax reliefs evaluation plan and a list of tax expenditures currently being evaluated. However, as in the Netherlands, there have been several constraints causing this plan to not always be on schedule, including funding constraints and unclear tax expenditure objectives hindering assessment. In the cases of both the Netherlands and the United Kingdom, this has resulted in some of the largest tax expenditures remaining unevaluated.

3.5.1. Evaluation frameworks

The tax expenditure evaluations of the four case study countries are guided by an evaluation framework. In Canada, direct programme spending is subject to the Treasury Board Policy on Evaluation, published in 2009. This policy is not required for tax expenditures, but the Department of Finance applies an evaluation framework that considers relevance, effectiveness, efficiency and equity of tax measures. However, it is rare for evaluations on tax expenditure to be published in Canada beyond the studies published with the tax evaluation report, with only two formal evaluations published between 2020 and 2025 (Government of Canada, 2026[30]; Auditor General of Canada, 2015[31]).

In the Netherlands, tax expenditures must be regularly evaluated following the criteria in the assessment framework. The ministry conducting a tax expenditure evaluation must consider whether a direct expenditure could have served as an alternative. This is in line with OECD guidance on tax incentives, which states that consideration of tax incentives should include an exploration of complementary and alternative policy options, including direct interventions and other alternatives such as training programmes and public investments (2026[19]). A negative evaluation, in theory, must mean either removal or modification of the tax expenditure. In reality, few negative evaluations have led to policy changes, as the rules also identify exceptions that can favour retaining the tax expenditure, e.g. when there is little new information since the last evaluation.

In Ireland, the first guidelines for evaluating tax expenditures were released by the Department of Finance in 2014. They include best practices for both ex ante and ex post evaluations. While these guidelines initially established that all tax expenditures would be examined over a five-year time period, this has proven challenging. In 2024, these guidelines were updated, with major changes including ex ante equality-proofing and beneficiary analysis more generally, as well as consideration of environmental implications, and review of whether existing data is sufficient for a review. These guidelines are not legally required, and highlight that not all evaluations will be able to meet the standards set out in these guidelines, although there is a requirement for tax expenditures with sunset clauses to be reviewed before the end of the sunset period.

In the United Kingdom, this takes the form of HMRC’s Evaluation Framework, which outlines criteria for deciding whether to evaluate, sets out three different types of evaluation, and describes how it connects with United Kingdom government evaluation guidance (such as the Green Book and Magenta Book). This framework, along with the list of tax expenditures mentioned above, was established after a National Audit Office review found in 2020 that there was no systematic evaluation of tax expenditures. Specific criteria for the evaluation of tax reliefs also exist, which include considering the relief’s cost and design, as well as both its direct and indirect impacts (National Archives, 2024[32]; HM Revenue and Customs, 2021[33]).

3.5.2. Spending Reviews

There is little evidence of spending reviews being conducted for tax expenditures. For example, both Canada and the United Kingdom conduct periodic spending reviews, but neither do so for tax expenditures. In Canada, the Auditor General explicitly recognised this in a 2015 speech to the Parliament (Office of the Auditor General of Canada, 2015[34]). Budget 2016 announced a review of tax expenditures and Budget 2017 announced the elimination of some inefficient tax measures or the amendment of others.

Ireland and the Netherlands conduct spending reviews as an annual exercise, and the Netherlands includes the conclusions from these reviews in its multi-annual budget planning. However, these reviews did not cover tax expenditures.

As tax expenditures can often be viewed as equivalents to direct expenditure, there is no reason they should not undergo spending reviews in the same manner. The current lack of such spending reviews increases the risk of inefficiencies and waste within existing tax expenditure programmes.

3.6. Oversight, scrutiny and audit of tax expenditures

3.6.1. Oversight and scrutiny

Several countries have oversight organisations whose responsibilities encompass tax expenditures. The key independent institution in Canada is the Office of the Parliamentary Budget Officer. While explicit tax expenditure review does not form part of the budget process in Canada, the Office does regularly publish fiscal outlooks which include tax expenditures. The comparable institute in Ireland, the Irish Fiscal Advisory Council, is responsible for assessing official budgetary projections and monitoring compliance with fiscal rules. However, its focus is on fiscal perspectives, rather than individual tax or spending measures (IFAC, 2025[35]). In the Netherlands, the Bureau for Economic Policy Analysis (CPB) checks the costings done by the Ministry of Finance on new tax expenditure policies, checking for any bias or uncertainty. In the United Kingdom, this is The Office for Budget Responsibility, who provides an independent assessment of tax forecasts and fiscal risks, certifies the costs of changes to tax policies (as prepared by HMRC and HM Treasury), and reviews the sustainability of the tax base.

Generally, the parliamentary committees responsible for financial oversight also take responsibility for scrutinising tax expenditures. In Canada, standing committees with roles relevant to tax expenditure include the Standing Committee on Finance, looking at the spending of the Department of Finance, and the Standing Committee on Government Operations, looking at the Treasury Board of Canada Secretariat and the budget estimates. This second one has a mandate to scrutinise tax expenditures, although the extent to which this occurs in practice is unclear. In the Netherlands, the Standing Committee for Finance, reviews tax expenditure provisions and also has responsibilities in relation to the national budget, the tax plan, and general matters for which the Minister of Finance has primary responsibility. In the United Kingdom, Parliament’s Treasury Select Committee is responsible for oversight of both HM Treasury and HMRC, including through the scrutiny of proposals in the annual budget. This includes reviewing tax expenditures.

There is some evidence of ad hoc parliamentary scrutiny leading to governance reforms. In Canada, the Standing Committee on Finance, responsible for oversight of the Department of Finance, has produced several reports which include recommendations related to tax expenditure. A 2016 pre-budget consultation included a recommendation that the federal government simplify the income tax system. In response, the government stated it would conduct a review of federal tax expenditures (Minister of National Revenue, 2016[36]). A 2022 pre-budget consultation report included a recommendation to undertake a public review identifying tax expenditures, loopholes and other tax avoidance mechanisms that particularly benefit high income individuals and large corporations (Standing Committee on Finance, 2022[37]).

In Ireland, the Committee on Budgetary Oversight has on several occasions released reviews of tax expenditure in Ireland, stating that they would incorporate examination of Department of Finance evaluations and reviews of tax expenditures more generally into their annual scrutiny programme. In a 2022 review, they concluded that the Department of Finance did not have enough staff to review tax expenditures, and so they should collaborate with the Department of Public Expenditure and Reform to increase capacity.

In the United Kingdom, the Public Accounts Committee held inquiries into tax expenditures in 2014 and 2020, examining the lack of transparency and value-for-money tests for tax expenditures, while in 2023 the Treasury Committee released a report calling for increased scrutiny and simplification of the current tax expenditure system.

Once enacted, tax expenditures are generally not subject to regular parliamentary scrutiny. Independent research in Canada has found that one reason for such a lack of scrutiny is a lack of information provided to parliament. While the Canadian Government’s Expenditure Management System means that policies are regularly reviewed, with performance information send to parliament, spending programmes authorised through tax acts are not subject to these requirements (Lester, 2024[38]). However, the Department of Finance monitors and evaluates tax expenditures on a regular basis. In Ireland, this was also highlighted as a point of improvement in a 2018 Parliamentary Budget Office report. In the Netherlands, regular scrutiny is more thorough. The tax expenditure report is included in the annual budget, which includes an overview of the fiscal costs of all tax expenditure provisions with revenue forgone above EUR 5 million, as well as summaries of recently performed evaluations (van Opstal, 2024[39]).

In the United Kingdom, government departments must publish annual reports and accounts, detailing their objectives, past performance, risk management, and financial statements. While HM Treasury’s Annual Report gives detailed accounts of government expenditure, treatment of tax expenditures is less systematic, although some reliefs appear in the narrative. However, HMRC’s report is more detailed, with estimates of certain tax reliefs, explanations of these forecasts, and estimates of error and fraud (HM Revenue and Customs, 2025[40]; HM Treasury, 2025[41]).

3.6.2. Audit

Regular reporting on tax expenditure audits is rare in the case study countries except the Netherlands, where tax expenditures are included in the Court of Audit’s annual Accountability Audit, in which the Court publishes its assessment of whether the government has spent taxpayer money wisely and according to the rules. However, in countries where refundable tax credits are recorded as expenditure, these credits receive a greater level of audit institution scrutiny. This is the case in the United Kingdom, where tax expenditures under this definition are included in departmental accounts and audited by the National Audit Office, and in Canada, where there is an annual audit of public accounts, which includes refundable tax credits.

In most cases, these countries have conducted ad hoc reviews of tax expenditures. For example, Canada’s Auditor General has conducted audits both on the systems used for managing tax expenditures and on the performance of specific tax expenditures. The Netherlands Court of Audit conducted audits on tax expenditure in 2020 and 2021, finding that the assessment framework was insufficiently applied. This led to the 2021 Cabinet tightening the rules of tax expenditures in the budget process. A 2020 review by the United Kingdom’s National Audit Office led to HMRC publishing a tax reliefs evaluation plan as well as a list of tax expenditures in the process of evaluation. The Office had also conducted two previous performance audits into the management of tax reliefs, as well as several audits on specific policies or tax expenditures.