A Practical Guide to Investment Tax Incentives

5. Evaluation

Abstract

Evaluation of the direct and indirect outcomes of tax incentives informs whether the policy is effective and efficient, and is key to building a case for reform or removal of incentives that are not delivering value for money. Practitioners can tailor evaluations to available resources and data, starting with descriptive analysis and building up capacity over time.

Evaluation involves assessing the costs and benefits of tax incentives. Evaluation is key to understanding if the tax incentive is best supporting policy goals, and to trigger reform or removal of inefficient or ineffective incentives. Evaluations ideally examine indirect as well as direct outcomes of the policy, such as spillovers to taxpayers that do not receive incentives. In many countries evaluations are not done or are limited in scope. Evaluation supports policy accountability, efficient use of public resources, and minimisation of distortions.1

Evaluation is challenging but essential to improve tax incentive policy, and can be tailored to account for capacity, data, and other constraints. It is not straightforward to identify the causal impact of a given tax incentive or potential outcomes in its absence. Robust evaluations require quality data as well as analytical tools and expertise. Depending on available resources, governments can adopt different approaches, from more descriptive to more in-depth impact assessments. Governments can also seek to build up evaluation capacity over time.

Table ‎5.1. Key steps for improving tax incentive evaluation

Key steps	Recommended Actions
1. Establish aims & governance
	1.1. Set goals	Determine aims, scope & frequency of evaluation2
			Identify evaluation criteria3, e.g., effectiveness, efficiency, distortions, equity, administrability, policy coherence
	1.2. Establish mandate	Require, by law, regular evaluations4
	1.3. Designate agency	Designate unit or agencies in charge of evaluation5
			Ensure sufficient staff, technical & financial resources
		Consider partnerships (across government bodies, with international or regional organisations, development agencies, universities, peer countries)
2. Determine most appropriate & feasible approach
	2.1. Assess & improve capacity	Assess capacity, data needs & availability for evaluation approach
		Set coordination mechanisms with data-collecting agencies
		Develop plan for improving capacity, data collection, widening evaluation goals
	2.2. Select or sequence approach	A. Build inventory of tax incentives6
		B. Construct descriptive statistics on tax incentive take-up, firm characteristics, observable outcomes7
		C. Estimate direct costs: revenue foregone, administrative & compliance costs8
		D. Estimate benefits & additional costs: impact, spillovers, distortions9
		E. Conduct cost-benefit analysis10
3. Report on findings		Determine audience, format & frequency of reports11
		Ensure findings clear and accessible for audience12
		Develop recommendations for reform
4. Consider and prepare for reform		Make a policy and political case for reform
		Consider including in a package of reforms to mitigate costs
		Consider best timing, potentially phased or sequenced changes
		Communicate plans to promote transparency and predictability

5.1. Establish aims and governance of evaluations

5.1.1. Establish a framework to conduct evaluations, including goals, scope, and frequency

Good ex-post evaluations need a clear framework that defines evaluation goals, their scope, and the frequency of assessments. The definition of impact varies based on the incentive goal, which agency requested the evaluation, how the government defines policy effectiveness, among other considerations. Setting goals of the evaluation can ensure that assessments are based on clear and impartial criteria to trigger potential reform, rather than used to justify policy continuation.

Several guides provide suggestions on how to set criteria for policy evaluations, based on considerations that relate to tax policy or other government programmes.13 These criteria are relevant for both ex-ante14 and ex-post evaluation, and include (Box 5.1, Example 1):

• Effectiveness: does the policy achieve its stated objectives?

• Efficiency: are outcomes achieved at low social costs, including revenue losses for government and other social costs?

• Distortions: does the incentive introduce economic inefficiencies or other distortions; does it bring additional or unexpected economic or social benefits or costs?

• Equity: what are the distributional effects of the policy, how does it affect horizontal and vertical equity and perceptions of fairness?

• Administrability: what are the compliance and administrative costs of the policy?

• Policy coherence: is the policy relevant and sound in the long-term? Does it align with wider development and policy goals of the country?

Evaluations should also support officials in understanding the trade-offs between outcomes, which governments will have to weigh when determining potential reform. A tax incentive might have achieved its objective, but is not efficient, for example, or may have introduced distortions by providing windfall benefits to other taxpayers beyond intended beneficiaries (GAO, 2012[1]). Or, an objective may have been achieved, but this goal is no longer coherent with current policy (Ministerie van Financiën, 2023[2]).

Capacity constraints can limit the scope of evaluations but should not prevent them. Rather than not conducting an evaluation, when capacity or data is constrained, governments can prioritise which incentives to evaluate and to what extent. While ideally governments should consider the widest range of possible impacts of the incentive, they might choose to evaluate incentives based on only a few criteria or in a less comprehensive manner (Box 5.1, Examples 2, 3). Other countries assess only the direct costs of incentives through tax expenditure reports, which seek to quantify the tax revenue forgone from tax incentives (as well as other tax expenditure), but do not address broader impacts.15

Once the goals of evaluations are set, assessments can occur regularly. For most countries, it is not feasible or practical to evaluate all tax incentives on an annual basis (Beer et al., 2022[3]). Countries might opt to conduct evaluations in multi-year cycles, focusing, for example, on temporary incentives about to expire, the oldest incentives, or the most costly in terms of revenue forgone, or group incentives by policy objective (Beer et al., 2022[3]; GAO, 2012[1]) (Box 5.1, Example 4).

5.1.2. Require evaluations in law, backed by high-level commitment

To ensure regular and systematic review, governments can enact legal requirements for tax incentive evaluation, depending on capacity. Legislation could define the full framework for evaluation (including evaluation criteria), or set the frequency and scope of incentives to be evaluated. Such requirements can be commensurate with capacity. A growing number of countries require by law assessing ex-post costs of tax incentives in terms of revenue foregone, through an annual tax expenditure report, which is a useful first step. However, only a minority of countries mandate, or conduct, evaluations that seek to assess effectiveness or impact beyond direct cost. According to the Global Tax Expenditure Database (GTED), only 16 of 106 countries that report on tax expenditures reference or include evaluations in these reports (Redonda, von Haldenwang and Berg, 2023[4]).

Beyond legal requirements, political commitment to tax incentive evaluation is key. Estimating costs of tax incentives can be a means to get political buy-in for further impact assessments, especially if revenue forgone is found to be substantial (Box 5.1, Example 5). Wider governance of state budgets is also essential for effective evaluation (ECA, 2024[5]).

5.1.3. Designate a team responsible for evaluation, and support coordination across relevant agencies

The legal mandate for evaluation should also designate which agency is responsible for assessments. In many high-income countries a designated Tax Policy Unit in the Ministry of Finance conducts evaluations. The Ministry of Finance is generally best placed to assess government revenue needs – comparing costs of tax incentives with potential spending elsewhere – and is likely to have access to the most relevant data for evaluation, either directly or through cooperation with the tax or revenue authority (IMF-OECD-UN-World Bank, 2015[6]). However, in many low and middle-income countries such units do not yet exist or are small, constraining evaluation.

Real and perceived independence of evaluators is important for the studies to support policy reform. For this reason or due to capacity constraints, some governments opt to outsource evaluations to non-governmental agencies, including international organisations or academics (see Section 5.2.1 below).

Box ‎5.1. Establishing aims and governance of evaluations: country examples

1. The United States Government Accountability Office provides additional guiding questions to help define different evaluation criteria (GAO, 2012[1]).

2. The Netherlands has a robust evaluation framework based on four criteria (effectiveness and efficiency, rationale for government intervention, complexity in implementation and enforcement, and feasibility), but acknowledges that evaluating for effectiveness is not always possible due to data constraints, or because incentive objectives are unclear. However, all incentives are assessed based on whether there is a clear and sound rationale for the policy (Ministerie van Financiën, 2023[2]).

3. The United Kingdom prioritises what to evaluate based on five considerations, including whether the policy is untested or new, and if data makes an evaluation methodologically feasible (HMRC, 2021[7]; HMRC, 2021[8]).

4. Germany and Ireland require evaluation of each tax incentives once over ten and five years, respectively, and Kenya aims to review tax incentives every five years. South Africa has no legal mandate to evaluate but the Treasury aims to not leave an incentive more than five years without an evaluation, prioritizing which incentives to evaluate based on number of beneficiaries or take-up. Canada has adopted a more staggered approach, evaluating one to two tax incentives each year (Redonda, von Haldenwang and Berg, 2023[4]).

5. In Uganda and Rwanda, requests for evaluation came after the publication of the first tax expenditure reports (Granger, McNabb and Parekh, 2022[9]).

5.2. Determine most appropriate and feasible approach

5.2.1. Assess what type of evaluation is feasible, developing plan to improve quality of estimates and widen evaluation goals and capacity

Evaluation approaches should depend on goals and capacity constraints. Governments can consider tax incentive evaluation as a gradual process, to be improved and expanded over time (Redonda, von Haldenwang and Berg, 2023[4]). As a first step, governments should consider qualitatively if the incentive is in line with broader policy objectives (Box 5.3 Example 1). Governments can also glean insights from descriptive statistics based on firm-level data on tax incentive take-up. Evaluators can then conduct cost estimates, before moving to more in-depth assessments that seek to estimate causal impact, to determine the effectiveness of the incentive. Box 5.2 outlines these different approaches, with examples of how countries have applied them. Evaluations can also draw on academic evidence or investor surveys to assess impact, though caution may be required in overinterpreting evidence (Box 5.3, Examples 2,3).

More sophisticated techniques can seek to examine causal impact, though this can be challenging. Identifying causal impact, and if an incentive triggered an effect that would not have occurred in its absence (such as additional investment, or additional jobs), is inherently challenging.16 Economic models and empirical analysis can support estimations, but many countries do not have teams trained to develop or run such assessments. Access to relevant data may also be lacking.17 Ministries of Finance and tax administrations in many low and middle-income countries often do not use advanced statistical software.

Governments can consider outsourcing evaluations through partnerships with external evaluators, which can also support impartiality. These include academics, international organisations or non-profits. Many regional and international organisations are actively working to support governments to conduct evaluations and build internal evaluation capacity (Box 5.3, Examples 4-8).

Regardless of scope of evaluations, governments should develop medium- to long-term plans for strengthening evaluation. This could include improving quality of tax expenditure reports and moving towards impact assessments (World Bank, 2024[10]). Governments can seek to build capacity by staffing junior researchers or explore peer learning exchanges with other countries.

Box ‎5.2. Evaluation of tax incentives: towards a maturity model

A series of approaches to evaluation with different levels of capacity and data needs exist. Governments could carry out the following steps, listed in order of least to most demanding in terms of resources. Governments can consider this as a phased approach to building up evaluation capacity.

Build an inventory of tax incentives

First, evaluators should understand if they have a full picture of all tax incentives offered; building an inventory of tax incentives can provide key qualitative information and reveal where information and data gaps exist. Inventories should at least list the policy goals of incentives, the type of benefit (i.e., exemptions, reduced rates, allowances, credits), the targeted taxpayers or activities, the legal basis, as well as which agencies are responsible for granting. This can provide some indication of the scope and generosity of incentives used, which can help to prioritise which incentives to further evaluate. An inventory should also facilitate regular reviews to assess whether measures remain relevant and fit for purpose, and to determine whether any should be discontinued or consolidated.

Construct descriptive statistics on take-up, firm characteristics, and observable outcomes; compare to policy goals

Second, evaluators can draw important initial conclusions by examining data on incentive take-up, including the number of companies that benefit and statistics on their characteristics and performance (e.g., sector, profitability, employment, firm size, multinational/domestic, location of investment), for example from CIT administrative data. Descriptive analysis of this data will not speak to causal impact, but it can give policymakers an idea of whether intended beneficiaries are using the tax incentive (Box 5.3, Example 9).

Estimate direct costs: revenue foregone, administrative & compliance costs

Estimating direct costs of incentives seeks to quantify the tax revenue forgone from incentives and are an essential first step into more robust evaluations. Most policy advice on tax incentive evaluation emphasises the importance of estimating revenue forgone. Numerous guides detail how to conduct and report on these estimates, using different methods (see Table ‎5.1). It is important to note, however, that tax expenditure reports are complex, and take time and resources to conduct (Box 5.3, Example 10). Tax expenditure reports that include statistics on expenditure by sector or other firm characteristics can also provide insights for additional evaluation.

Estimate benefits and additional costs: impact, effectiveness, spillovers, distortions

More robust assessments can contrast revenue forgone with the benefits from the tax incentive; that is, its (causal) impact on achieving policy objectives and positive spillovers. The latter could include benefits to productivity, employment or skills, which are even more challenging to assess. Evaluation can also seek to identify, and at best estimate, potential other costs related to negative spillovers or market distortions (e.g. investment displaced in activities not targeted by the incentive).

Short of identifying causal impact, governments could compare descriptive statistics on incentive beneficiaries to revenue forgone and the stated goals of incentives to infer initial conclusions on costs versus benefits (Box 5.3, Example 11). While such analysis should be read with caution as it does not allow conclusions on the causal impact of incentives on jobs and exports, it can present useful information; for example, highlighting costly tax exemptions to high-income sectors that have been in place for decades. Distributional analysis on beneficiaries can also provide useful insights for governments (Box 5.3, Example 12).

Conduct cost-benefit analysis: more in-depth impact studies

As a next step, governments can seek to estimate wider and indirect costs and benefits or compare these outcomes in more in-depth impact assessments and cost-benefit analysis. Multiple guides provide details on how to conduct such in-depth impact studies.18 Such evaluations are resource intensive.

Short of these assessments, and even using limited information, governments can pose critical questions about the impact of the incentive, identifying what is known and unknown. For example, if costs appear to be high, take-up low, and there are limited indicators that target outcomes have shifted, evaluators might suggest that the incentive has not demonstrably proven results. This is a key first step for more in-depth assessments.

5.2.2. Consider how existing data can be used to gain insights, and establish a strategy for collecting additional data that might be critical for evaluation

Governments should consider what additional data is required to conduct evaluations and establish processes to make these available. This can include estimates of foregone revenue but also assessments of wider impact (see Monitoring stage on data gathering and availability). Collecting additional data to examine impact has costs, but other agencies might already collect relevant information, for example national statistical agencies (e.g. household and business surveys), customs authorities, development agencies, or line ministries (e.g. energy emissions) (ECOWAS-UEMOA, 2022[11]).

Box ‎5.3. Determining the most appropriate and feasible evaluation approach: country examples

1. Governments can first assess qualitatively if in the incentive is in line with government priorities. In the Netherlands, some incentives were deemed ineffective because their aims were inconsistent with other government goals (e.g., a tax exemption for energy-intensive processes does not align with the GHG emission reduction goals of the country.) (Ministerie van Financiën, 2023[2]).

2. Evaluations can also draw on academic evidence. One state in the United States does not conduct its own evaluations but combines a literature review of economic studies on effectiveness of similar incentives with descriptive statistics of take-up to suggest impact in their specific case (OPEGA Maine State Legislature, 2020[12]).

3. In cases where outcomes are gleaned through investor surveys, as has been used in Indonesia, results should be interpreted with caution, particularly as they may not be representative and responses may be biased (Kronfol and Steenbergen, 2020[13]; BKF, 2023[14]).

4. Governments can also consider partnering with external evaluators. An econometric evaluation of a tax credit in Uruguay for investment promotion was carried out through a partnership between the Centre for Fiscal Studies, the Centre for Economic Research, and the Economics Institute at the University of the Republic in Uruguay, supported by regular interaction and data exchange with the government and relevant implementing agencies (Llambi et al., 2018[15]).

5. In the Republic of Korea, line ministries conduct a first self-assessment on operational performance of the tax incentive, and report this to the Ministry of Finance. More in-depth evaluations are then conducted by independent government research agencies (Redonda, von Haldenwang and Berg, 2023[4]).

6. The Netherlands often contract independent research agencies to conduct evaluations through an official procurement based on quality of the proposal, research staff and price (van Opstal, 2024[16]).

7. The United Kingdom has also commissioned research institutions for specific evaluations, when the study requires new research, data, and expertise (HMRC, 2021[8]; HMRC, 2022[17]).

8. ECOWAS has proposed a regional evaluation tool for tax expenditures to harmonise practice across the community (ECOWAS-UEMOA, 2022[11]).

9. In the Netherlands the evaluation report determined that some untargeted incentives (available and used by a wide range of taxpayers) likely have budgetary costs disproportionate to their purpose, while other incentives were deemed inefficient because they are not extensively used by firms (Ministerie van Financiën, 2023[2]).

10. In Rwanda, it took more than three years to build a process for preparing annual tax expenditure reports (Granger, McNabb and Parekh, 2022[9]).

11. Honduras looked at job creation and export performance of beneficiaries, and whether take-up is higher in high- or lower-income sectors, and duration of benefit (Secretaría de Finanzas, 2023[18]).

12. Australia assesses the distributional effects of large tax expenditures, including the share of the benefit granted to companies by business turnover and sector. It plans to add analysis by geographical region and by urban and regional area, which can support assessment of whether incentives meet aims of regional development (The Australian Government the Treasury, 2023[19]).

5.3. Report on findings of analysis

For evaluations to lead to policy action, findings must be reported on. Legal mandates for evaluation can set minimum requirements for frequency of reports and format (e.g., annexed to annual budgets). Prepared evaluation reports to coincide with the annual budget can encourage reflections on whether spending on tax incentives is merited considering other spending requirements or objectives (Heady and Mansour, 2019[20]).19

Clear, accessible, and public report can best increase policy transparency and accountability. For most transparency and accountability, evaluations can be reported to Parliament and made public. Governments may consider submitting more in-depth technical reports to relevant ministries, and summary reports to Parliament, taking care to make findings accessible to a non-technical audience (particularly if Parliament does not have a technical fiscal policy committee) (Berg et al., 2024[21]). To this end, reports can include statistics that contextualize tax incentive spending, for example, revenue foregone as a percentage of total tax revenue, average cost per capita (as done in Italy’s Tax Expenditure Report) and compared to other costs of direct spending programmes (e.g., education, health, public infrastructure). Including a breakdown of tax incentive cost by sector or even by top beneficiaries (firms) can support public debate on their merits.

Evaluation reports can include recommendations for reform, and all relevant information to inform reform decisions. In addition to how the information is presented, evaluation reports to Ministries in charge of tax incentive policy and Parliament or committees overseeing budget decisions can describe how the incentive performs based on the established evaluation criteria (e.g. effectiveness, efficiency, administrability, policy coherence). Clearly outlining the different outcomes of the incentive will support officials in understanding trade-offs and make informed decisions on whether tax incentives remain justified, should be reformed, or removed. Reports should acknowledge the limits of the evaluation, particularly in cases where no conclusive outcome is determined, along with the reasons for inconclusive findings.

5.4. Consider and prepare for reform

Reform strategies can range from removal to simplification and adjustment, varying with political and economic context.20 Reform takes time, and may be easier in some contexts, and at some times, than others (Box 5.4, Example 1). As discussed throughout this guide, even short of comprehensive reform or removal, governments can improve outcomes by addressing shortcomings across the incentive lifecycle; including by simplifying procedures to receive incentives, increasing transparency and enhancing monitoring and evaluation to provide evidence for whether incentives are meeting policy goals, and at what costs (Box 5.4, Example 2).

Political, legal, and administrative hurdles can impede reform. There may be significant political economy challenges in removing incentives that provide substantial benefits to a subset of taxpayers. In addition, as with all policy reform, there may be legal and procedural hurdles to removing incentives, including related to the legislative and budget process. Governments may also face legal challenges where incentives are granted through contracts or subject to stabilisation clauses in international investment agreements (IIAs) (see Conception).

Policymakers may seek to mitigate certain reform challenges before the incentive is implemented. Certain design choices could make reform easier. For example, sunset clauses with non-automatic renewal can support debates among lawmakers on the merits of continuing certain incentives (see Design). Incentives can also be granted long-term, but only for investment within a temporary window (Box 5.4, Example 3).

Successful reform requires a policy and political case. Evaluations that show an incentive is ineffective or inefficient can allow for a clear policy case for why reform is needed, and what type of reform, whether removal, redesign or improved implementation, is recommended. Impartial evaluation reports, based on empirical evidence, can help policymakers make informed decisions (Redonda, von Haldenwang and Berg, 2023[4]). Building political support, with buy-in from key stakeholders is often necessary (Jung, 2023[22]). This could include dialogue with taxpayers that will be affected by the reform, especially for investors that receive incentives through investor-state contracts. It also requires clear identification of affected taxpayers – underscoring the importance of monitoring incentive uptake (see Monitoring). Policymakers can also seek to effectively communicate the economic and fiscal case for reform to build public support for reforms.

Including tax incentive reform in a package of fiscal and administrative reforms can mitigate costs. Several examples of successful tax incentive reform have been passed as part of a wider set of tax policy reforms (Akitoby et al., 2019[23]; Jung, 2023[22]). Bundling reforms can build political momentum and public support, combine reduced benefits with improvements elsewhere in the tax system, and ease distributional impacts. This could occur as part of a wider strategy implemented over a few years. Depending on fiscal space, removing tax incentives can even fund lower rate, broader base systems, which may eventually encourage more overall investment with fewer distortions (Box 5.4, Example 4). Changes in international tax rules and agreements have also provided a catalyst for tax incentive reform.21

Careful timing of reform, with phased or sequenced changes can improve chances of success. Phasing out of incentives, or gradually reducing generosity can support existing beneficiaries of incentives (Box 5.4, Example 5). Often, reform or removal of an incentive applies only to new investors, with grandfather provisions that allow existing beneficiaries to continue to receive the incentive (IMF-OECD-UN-World Bank, 2024[24]).

Well communicated reform plans can promote transparency and predictability. Signaling reforms well in advance can give time for investors to adjust and plan (Box 5.4, Example 6). Ample time might be useful, but if reform timelines are too long, it can risk repeals or backtracks by new governments. Reforms that are announced and do not materialise, or reforms that are too frequent, can generate confusion and uncertainty for investors.

Box ‎5.4. Considering and preparing for reform: country examples

1. Burkina Faso, The Gambia, Maldives, Mauritania, Rwanda, Senegal, Uganda and Nigeria provide examples of successful revenue mobilisation reform, including in several cases removing tax incentives; see (Akitoby et al., 2019[23]; Jung, 2023[22]).

2. South Africa changed the process by which investors can receive the R&D tax incentive, from a self-assessment to a pre-approval system, following industry concerns and an internal evaluation

3. Argentina introduced a 30-year incentive package, including an incentive for certain large projects that commence within two years of passage of the law.

4. Examples of tax incentive reform within a wider package of reform include Tunisia’s simultaneous adjustment of CIT rates (increase in the CIT rate in strategic sectors, while lowering the standard CIT rate), Rwanda and Uganda’s removal of VAT and other indirect tax exemptions alongside improvements in tax administration after 2010, and Mauritania’s use of targeted cash transfers to protect the most vulnerable following removal of certain tax incentives (Akitoby et al., 2019[23]).

5. Phased reform includes in Mauritius, where firms in Freeports that received tax exemptions before a 2018 reform could continue to benefit until 2021, but not after (EDB Mauritius, 2018[25]). Jordan gradually increased reduced CIT rates for manufacturing industries over six years (OECD, 2025[26]).

6. In Namibia, substantial reforms to Export Processing Zones, including removal of tax holidays, were signaled several years before they took effect. The reforms were passed in 2020 with a five-year transition period (Namibia also has introduced a phased reduction in the statutory CIT rate).

References

[34] ADB (2023), Tax Expenditure Estimation Toolkit, Asian Development Bank, Manila, https://www.adb.org/sites/default/files/publication/932086/tax-expenditure-estimation-tool-kit.pdf (accessed on 4 July 2024).

[23] Akitoby, B. et al. (2019), “Case Studies in Tax Revenue Mobilization in Low-Income Countries”, IMF Working Papers, No. 104, IMF, Washington, DC, https://doi.org/10.5089/9781498314565.001.

[29] Aliu, F., A. Redonda and C. von Haldenwang (2022), The Global Tax Expenditures Database (GTED) Progress Report, German Development Institute / Deutsches Institut für Entwicklungspolitik (DIE), the Council on Economic Policies (CEP), https://gted.taxexpenditures.org/wp-content/uploads/2022/04/220513_GTED_ProgressReport.pdf.

[42] ATI (2025), ATI post-2020 Monitoring Framework, Addis Tax Initiative, Addis Ababa, https://www.addistaxinitiative.net/sites/default/files/resources/ATI%20Declaration%202025%20Monitoring%20Framework.pdf (accessed on 17 March 2025).

[43] ATI (ed.) (2024), Ex-Post Evaluation of TEs II – Concepts & Methods, United Nations University World Institute for Development Economics Research, Kathmandu, https://www.addistaxinitiative.net/sites/default/files/resources/Session%209_Ebrahim.pdf (accessed on 19 September 2024).

[3] Beer, S. et al. (2022), “How to Evaluate Tax Expenditures”, How-To Note, No. 2022/005, IMF, Washington, DC, https://www.imf.org/en/Publications/Fiscal-Affairs-Department-How-To-Notes/Issues/2022/11/How-to-Evaluate-Tax-Expenditures-525166.

[21] Berg, S. et al. (2024), Pocket Guide on Tax Expenditures for Parliamentarians, International Tax Compact (ITC), Bonn, https://www.addistaxinitiative.net/sites/default/files/resources/ATI%20Pocket%20Guide%20on%20TEs%20for%20MPs%20EN.pdf (accessed on 4 July 2024).

[14] BKF, R. (2023), Laporan Belanja Perpajakan 2022 (Tax Expenditure Report), Badan Kebijakan Fiskal Kementerian Keuangan (BKF) (Ministry of Finance Fiscal Policy Agency), Jakarata, https://fiskal.kemenkeu.go.id/publikasi/tax-expenditure-report (accessed on 13 September 2024).

[27] Bulman, A. et al. (2021), Guide on incentives for responsible investment in agriculture and food systems, Food and Agriculture Organization of the United Nations (FAO) and the Columbia Center on Sustainable Investment (CCSI), https://www.fao.org/family-farming/detail/en/c/1396923/.

[44] CIAT & UN (2018), Design and Assessment of Tax Incentives in Developing Countries, United Nations & Inter-American Center of Tax Administrations (CIAT),, https://www.ciat.org/Biblioteca/Estudios/2018_design_assessment_tax_incentives_UN_CIAT.pdf.

[31] Congressional Budget Office (2021), Using Survey Data for Tax Analysis, https://www.cbo.gov/publication/57540 (accessed on 17 March 2025).

[5] ECA (2024), A Framework for Assessing and Reporting Tax Expenditures in Africa, United Nations Economic Commission for Africa, Addis Ababa, https://repository.uneca.org/bitstream/handle/10855/50021/b12040423.pdf?sequence=1&isAllowed=y.

[11] ECOWAS-UEMOA (2022), Methodological guide for evaluating tax expenses in West Africa, ECOWAS, UEMOA, Programme d’Appui à la Transition Fiscale (PATF), https://www.patf-ao.org/wp-content/uploads/2024/03/GM-Depenses-Fiscales-Ang_V3bdef.pdf.

[25] EDB Mauritius (2018), Freeport Communique, https://edbmauritius.org/wp-content/uploads/2022/12/Freeport-Communique-2018.pdf.

[1] GAO (2012), “Tax Expenditures: Background and Evaluation Criteria and Questions”, No. GAO-13-167SP, United States Government Accountability Office (GAO), Washington, DC, https://www.gao.gov/assets/gao-13-167sp.pdf (accessed on 11 July 2024).

[9] Granger, H., K. McNabb and H. Parekh (2022), “Tax expenditure reporting in Rwanda and Uganda: challenges, practical guidance and lessons learnt”, ODIWorkingPaper, ODI, London, https://cdn.odi.org/media/documents/ODI_Working_paper_Tax_expenditure_reporting_in_Rwanda_and_Uganda.pdf (accessed on 18 September 2024).

[28] Granger, H., D. Phillips and R. Warwick (2021), An introduction to tax policy appraisal: A guide to assessing the effectiveness and potential impacts of alternative tax policy options, The Institute for Fiscal Studies, London, https://www.taxdev.org/sites/default/files/2021-11/TaxDev_Policy_Appraisal_Manual.pdf (accessed on 11 July 2024).

[30] Grote, M. (2017), “How to Establish a Tax Policy Unit”, How to Note, No. 7, IMF, Washington, DC, https://www.imf.org/en/Publications/Fiscal-Affairs-Department-How-To-Notes/Issues/2017/10/19/How-to-Establish-a-Tax-Policy-Unit-45322 (accessed on 3 July 2024).

[20] Heady, C. and M. Mansour (2019), Tax Expenditure Reporting and Its Use in Fiscal Management: A Guide for Developing Economies, IMF, Washington, DC.

[17] HMRC (2022), “Creative Industry Tax Reliefs Evaluation”, HMRC research report, No. 684, HM Revenue and Customs (HMRC), Ipsos, Olsberg SPI, London Economics, https://www.gov.uk/government/publications/creative-industry-tax-reliefs-evaluation (accessed on 13 September 2024).

[7] HMRC (2021), HMRC Evaluation Framework, HM Revenue and Customs (HMRC), London, https://webarchive.nationalarchives.gov.uk/ukgwa/20240502190415/https://www.gov.uk/government/publications/hmrc-evaluation-framework (accessed on 13 September 2024).

[8] HMRC (2021), Official Statistics: Tax relief statistics (December 2021), HM Revenue and Customs (HMRC), London, https://webarchive.nationalarchives.gov.uk/ukgwa/20240502212234/https://www.gov.uk/government/statistics/main-tax-expenditures-and-structural-reliefs/estimated-cost-of-tax-reliefs-statistics#annex-a (accessed on 13 September 2024).

[24] IMF-OECD-UN-World Bank (2024), The Platform for Collaboration on Tax: Tax Incentives Principles. Public Consultation Draft, IMF-OECD-UN-World Bank.

[33] IMF-OECD-UN-World Bank (2015), Background paper to the report prepared for the G-20 Development Working Group: Options for Low Income Countries Effective and Efficient Use of Tax Incentives for Investment, IMF-OECD-UN-World Bank, https://www.oecd.org/ctp/tax-global/background-document-options-for-low-income-countries-effective-and-efficient-use-of-tax-incentives-for-investment.pdf (accessed on 28 July 2021).

[6] IMF-OECD-UN-World Bank (2015), Options for Low Income Countries’ Effective and Efficient Use of Tax Incentives for Investment, A report prepared for the G-20 Development Working Group by the IMF, OECD, UN and World Bank., https://www.oecd.org/tax/options-for-low-income-countries-effective-and-efficient-use-of-tax-incentives-for-investment.htm.

[22] Jung, I. (2023), “Nigeria’s Tax Revenue Mobilization: Lessons from Successful Revenue Reform Episodes”, Selected Issues Papers, No. 19, IMF, Washington, DC, https://doi.org/10.5089/9798400234989.018.

[41] Kassim, L. and M. Mansour (2018), “Tax expenditures reporting in developing countries: An evaluation”, Revue d’économie du développement, Vol. 26/2, pp. 113-167, https://www.cairn-int.info/article-E_EDD_322_0113--tax-expenditures-reporting-in.htm (accessed on 4 July 2024).

[13] Kronfol, H. and V. Steenbergen (2020), Evaluating the Costs and Benefits of Corporate Tax Incentives IN FOCUS Methodological Approaches and Policy Considerations, World Bank, Washington, DC, https://documents1.worldbank.org/curated/en/180341583476704729/pdf/Evaluating-the-Costs-and-Benefits-of-Corporate-Tax-Incentives-Methodological-Approaches-and-Policy-Considerations.pdf (accessed on 8 March 2023).

[15] Llambi, C. et al. (2018), “Are Tax Credits Effective in Developing Countries? The Recent Uruguayan Experience”, Economía LACEA Journal, Vol. 18/2, pp. 25-28, https://economia.lse.ac.uk/articles/10.31389/eco.58.

[2] Ministerie van Financiën (2023), Official report on the approach to tax expenditures: Opportunities for lower tax rates, better tax expenditures, and a more understandable and simpler system, Government of the Netherlands, Hague, https://www.government.nl/documents/reports/2023/09/11/official-report-on-the-approach-to-tax-expenditures (accessed on 11 September 2024).

[26] OECD (2025), “OECD Investment Tax Incentives Database 2024 update: Corporate income tax incentives in emerging and developing economies”, OECD Business and Finance Policy Papers, No. 79, OECD, Paris Cedex 16, https://doi.org/10.1787/b0de19dc-en.

[47] OECD (2023), “Improving transparency of incentives for investment facilitation”, OECD Business and Finance Policy Papers, No. 35, OECD, Paris, https://doi.org/10.1787/ab40a2f1-en.

[32] OECD (2022), OECD Tax Policy Reviews: Colombia 2022, OECD Tax Policy Reviews, OECD Publishing, Paris, https://doi.org/10.1787/054722db-en.

[45] OECD (2010), Making Reform Happen: Lessons from OECD Countries, OECD Publishing, Paris, https://doi.org/10.1787/9789264086296-en.

[46] OECD et al. (2024), Reform Management Fundamentals: Setting Up a Reform Program, OECD Publishing, Paris, https://doi.org/10.1787/4d33b619-en.

[40] OFMA (2020), 2020 Indiana Tax Incentive Review, Office of Fiscal and Management Analysis (OFMA), Legislative Services Agency, Indiana General Assembly, https://documents.ncsl.org/wwwncsl/Fiscal/evaluationDB/2020-tax-incentive-evaluation-final.pdf.

[12] OPEGA Maine State Legislature (2020), “Maine Capital Investment Credit (MCIC): A Complicated Response to Federal Bonus Depreciation that Is Unlikely to Significantly Encourage Capital Investment in Maine”, Tax Expenditure Review: report to the Government Oversight Committee and Taxation Committee, No. Report No. TE-MCIC-17, The Office of Program Evaluation and Government Accountability of the Maine State Legislature, Augusta, https://documents.ncsl.org/wwwncsl/Fiscal/evaluationDB/MaineCapitalInvestmentTaxCredit.pdf (accessed on 13 September 2024).

[37] Redonda, A. and T. Neubig (2018), Assessing Tax Expenditure Reporting in G20 and OECD Economies, Council on Economic Policies, Zurich, https://www.cepweb.org/wp-content/uploads/2018/11/Redonda-and-Neubig-2018.-Assessing-Tax-Expenditure-Reporting.pdf.

[38] Redonda, A., C. Von Haldenwang and F. Aliu (2021), “Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa”, in Mosquera Valderrama, I., D. Lesage and W. Lips (eds.), Taxation, International Cooperation and the 2030 Sustainable Development Agenda, Springer, https://doi.org/10.1007/978-3-030-64857-2_9.

[39] Redonda, A., C. von Haldenwang and F. Aliu (2022), Global Tax Expenditures Database [data set], Version 1.1.4.

[4] Redonda, A., C. von Haldenwang and S. Berg (2023), “Evaluation of tax expenditures: Conceptual frameworks and international experiences”, Project Documents, No. LC/TS.2023/129, Economic Commission for Latin America and the Caribbean (ECLAC), Santiago, https://repositorio.cepal.org/server/api/core/bitstreams/7faad2b6-803b-4bed-859a-d45a493e872c/content (accessed on 4 July 2024).

[18] Secretaría de Finanzas (2023), Gasto Tributario 2023 (Tax Expenditure), Secretaría de Finanzas, Republica de Honduras, https://www.sefin.gob.hn/gasto-tributario/ (accessed on 13 September 2024).

[19] The Australian Government the Treasury (2023), Tax Expenditures and Insights Statement, The Australian Government the Treasury, Parkes, https://treasury.gov.au/sites/default/files/2023-02/p2023-370286-teis.pdf.

[16] van Opstal, R. (2024), Tax Expenditures Country Report: the Netherlands, German Institute of Development and Sustainability (IDOS) and Council on Economic Policies (CEP, Bonn, https://www.idos-research.de/publikationen/mitarbeiter-sonstige/article/tax-expenditures-country-report-the-netherlands/ (accessed on 13 September 2024).

[36] Vázquez, R. (2022), “Measurement of tax expenditures in Latin America”, Project Documents, No. LC/TS.2022/148, Economic Commission for Latin America and the Caribbean, Santiago, https://repositorio.cepal.org/server/api/core/bitstreams/be42490d-c86e-4f5f-a9a9-b664a50ba70c/content (accessed on 11 July 2024).

[35] von Haldenwang, C., A. Redonda and F. Aliu (2023), Tax Expenditures in an Era of Transformative Change: the GTED Flagship Report 2023, IDOS, Bonn, https://www.taxexpenditures.org/2023/10/06/tax-expenditures-in-an-era-of-transformative-change/ (accessed on 5 September 2024).

[10] World Bank (2024), Tax Expenditure Manual, World Bank Group, Washington, DC, http://documents.worldbank.org/curated/en/099062724151636908 (accessed on 14 February 2025).

Notes

← 1.  Evaluation is one of the key Principles to Improve Tax Incentive Policymaking developed by the Platform of Collaboration on Tax (IMF-OECD-UN-World Bank, 2024[24]).

← 2.  Guidance on frequency and scope of incentive evaluation and country examples in: Beer et al. (2022[3]).

← 3.  Templates on evaluation criteria and relevant questions for assessment in World Bank (2024[10]) and GAO (2012[1]) (most comprehensive qualitative questions), Redonda, von Haldenwang and Berg (2023[4]) (includes detailed country examples), Beer et al. (2022[3]) and ECA (2024[5]), general guidelines as part of the lifecycle in Bulman et al. (2021[27]); for ex-ante assessment see Granger, Phillips and Warwick (2021[28]). The Global Tax Expenditures Lab (GTEL) organises trainings on TE evaluation and plans to publish a repository of evaluations published by countries (Aliu, Redonda and von Haldenwang, 2022[29]).

← 4.  Broad guidelines on legal framework in (Berg et al., 2024[21]).

← 5.  On establishing a tax policy unit see (Grote, 2017[30]).

← 6.  Description on what to include in inventories see: (Heady and Mansour, 2019[20]; OECD, 2023[47]).

← 7.  For an example from the United States on matching tax returns with survey data for analysis, see (Congressional Budget Office, 2021[31]).

← 8.  Detailed guides on tax expenditure estimates and reporting include: (Heady and Mansour, 2019[20]; OECD, 2022[32]; ECA, 2024[5]; IMF-OECD-UN-World Bank, 2015[33]) for details on micro-simulation model, creation of excel-based model for revenue forecasting; (ADB, 2023[34]; von Haldenwang, Redonda and Aliu, 2023[35]). For application in country contexts see (Vázquez, 2022[36]) (Latin America), (Redonda and Neubig, 2018[37]) (G20 & OECD), (Redonda, Von Haldenwang and Aliu, 2021[38]; Granger, McNabb and Parekh, 2022[9]) (Africa)

← 9.  Guides on different methodologies for estimating effectiveness: (World Bank, 2024[10]; Beer et al., 2022[3]; Kronfol and Steenbergen, 2020[13]).

← 10.  Cost-benefit analysis framework in (World Bank, 2024[10]; IMF-OECD-UN-World Bank, 2015[33]).

← 11.  Examples of reporting on tax expenditures in (Redonda, von Haldenwang and Aliu, 2022[39]) and (von Haldenwang, Redonda and Aliu, 2023[35]); on examples of reports to legislators, see in the United States (OFMA, 2020[40]).

← 12.  Examples of elements to include in tax expenditure reports (Kassim and Mansour, 2018[41]; ATI, 2025[42]); and overview with country examples (Aliu, Redonda and von Haldenwang, 2022[29]).

← 13.  See among others (GAO, 2012[1]; World Bank, 2024[10]; Beer et al., 2022[3]; Kronfol and Steenbergen, 2020[13]; ECA, 2024[5]).

← 14.  While ex-ante evaluations seek to establish the rationale for policy intervention and foresee distortions, in order to inform better policy design, ex-post evaluations aim to understand impact through data analysis. For more on the ex-ante evaluation frameworks see Conception and Redonda, von Haldenwang and Berg (2023[4])

← 15.  While foregone revenue is crucial to understand direct costs, it addresses only one facet of the effects of tax incentives. It is important to consider that tax incentives can have wide impact across the economy and society, which can affect how governments determine whether the policy is effective and efficient.

← 16.  Identifying causal outcomes requires comparing beneficiaries to comparable taxpayers that did not receive the incentive (counterfactual). Identifying this comparison group and isolating the impact of the tax incentive (as opposed to other factors) requires data and statistical and econometric methods. For more on estimation tools to identify comparator groups and causal impact for tax incentives see for example, (Ebrahim, 2024[43]).

← 17.  For descriptions of different econometric tools for these assessments see (World Bank, 2024[10]; Kronfol and Steenbergen, 2020[13]; Beer et al., 2022[3]).

← 18.  Cost-benefit analysis framework in (World Bank, 2024[10]; IMF-OECD-UN-World Bank, 2015[33]; Kronfol and Steenbergen, 2020[13]; CIAT & UN, 2018[44]).

← 19.  For guidance on what to include in tax expenditure reports, see Table 5.1.

← 20.  Brief discussion on challenges of reforming tax incentives in (Redonda, von Haldenwang and Berg, 2023[4]). For examples of obstacles and strategies to implement wider tax reform, drawing on examples in OECD countries see (OECD, 2010[45]). The OECD, IMF, Inter-American Centre of Tax Administrations and Intra-European Organisation of Tax Administration (2024[46]) provide a guide for tax administration reform.

← 21.  For example, many governments have introduced substance requirements for CIT incentives deemed to have high BEPS risks, in accordance with Action 5 of the OECD/G20 BEPS project.