Tax expenditures are among the most significant yet least visible instruments of public policy. Through exemptions, deductions, reduced rates, and preferential regimes, governments pursue economic, social, and regional objectives via the tax system rather than direct spending. Because these measures often operate off budget, they typically receive far less scrutiny than expenditure programs, even when their fiscal costs are substantial. The latest figures from the Global Tax Expenditures Database (GTED) indicate that average reported revenue forgone amounts to 3.7 percent of GDP and 23 percent of tax revenue globally.
This blog builds on four Tax Expenditures Country Reports published within the series hosted by the Council on Economic Policies (CEP) and the German Institute of Development and Sustainability (IDOS), and shows how Brazil, France, India, and the United Kingdom govern tax expenditures across three critical dimensions: benchmark definition, reporting, and evaluation. Two key takeaways emerge from the experience of two advanced and two large emerging market economies: First, tax expenditure policy making is deeply country-specific, particularly—though not exclusively—in the way benchmarks are defined. Second, tax expenditure analysis is technically demanding and resource intensive, making it especially challenging for lower-income countries. Yet even though high-income countries are generally better equipped to conduct such analysis, owing to stronger data infrastructure, administrative capacity, and technical expertise, many observed shortcomings are strikingly similar across contexts, including limited transparency, inconsistent methodologies, and weak evaluation practices.
In an era of rising debt and constrained fiscal space, governments at all income levels can no longer afford to overlook large quasi-spending programs delivered through the tax system without robust evidence of their effectiveness and value for money.
Benchmarks matter
At the core of tax expenditure analysis lies a deceptively technical—and inherently political—question: how to define the benchmark tax system, that is, the reference point against which deviations are identified. Because tax expenditures are deviations from this benchmark, what governments treat as “normal” taxation ultimately determines which measures are classified, costed, and evaluated as tax expenditures.
In Brazil, the Federal Revenue Service defines a benchmark for each major tax in a supplement to the annual tax expenditure report, clarifying why measures such as consumption tax exemption for the basic food items, incentives for operating in the Manaus Free Trade Zone, and the preferential regimes for small firms (the so-called Simples Nacional) are classified as tax expenditures. An explicit benchmark makes normative choices more visible. For example, the exemption of dividends received by individuals is considered part of the benchmark on the grounds of integrating corporate and personal taxation, despite ongoing debate about its distributive effects.
Similarly,France provides a clear benchmark definition. The French benchmark is relatively broad and excludes several social tax measures from the scope of tax expenditures. Tax benefits related to social security contributions are not counted as tax expenditures, and reduced value-added tax (VAT) rates on essentials such as food—often classified as tax expenditures elsewhere—are treated as part of the benchmark and are therefore omitted from reporting. Over time, some measures, such as reduced tax liabilities for families with children, have been absorbed into the benchmark without a transparent explanation of the decision and its revenue implications.
In India, statutory personal income tax rates above basic exemptions and statutory corporate income tax rates constitute the benchmark. Preferential thresholds, deductions, and concessions that lower effective tax burdens are classified as tax expenditures. The concept also applies to customs duties. However, the absence of a goods and services tax (GST) benchmark means that exemptions and reduced rates under the GST are excluded from tax expenditure reporting.
The United Kingdom takes a different approach. Since the late 1970s, HM Treasury has rejected the benchmark concept and has instead distinguished between “structural” and “non-structural” tax reliefs. While this avoids explicit debates about what constitutes the benchmark, it weakens comparability with direct public spending and limits transparency. The controversy surrounding the VAT exemption for private education—classified as structural until its abolition in 2025—illustrates the political salience of these classification choices.
Making hidden subsidies visible through reporting
Once tax expenditures are identified, estimating and publishing their fiscal cost is critical for informed policymaking.
In Brazil, the federal government regularly publishes federal tax expenditure reports linked to the budget cycle, providing revenue forgone estimates by provision and tax type, along with stated policy objectives and time frames. This framework aligns closely with good international practice. However, as in many decentralized countries, reporting at the state level is uneven: differences in definitions, estimation methods, and classifications limit comparability across states and over time.
France also publishes a comprehensive annual list of tax expenditures alongside the budget. The French report stands out for its detailed information, including number of beneficiaries—a rare but valuable feature for assessing distributional effects. However, methodological changes have at times lacked transparency and drawn criticism from the French Audit Office. For instance, a controversial methodological change reduced the reported cost of VAT-related tax expenditures by around €10 billion in the 2024 Budget Proposal. Although the Audit Office recommended reverting to the old methodology, this recommendation was not followed in the 2025 budget.
India also includes tax expenditure statements in the annual budgets. A notable strength of recent tax expenditure statements is reclassifications of certain export-related subsidy schemes to direct spending, which improves conceptual clarity. Nevertheless, frequent policy changes, multiple tax regimes, and very limited state-level reporting constrain consistency and transparency. The absence of GST-related data remains a significant gap.
The United Kingdom does not publish a comprehensive tax expenditure report. Instead, it releases statistics for a subset of provisions, reflecting its distinction between structural and non-structural provisions. Only about one third of tax reliefs are classified as non-structural and therefore subject to cost reporting; the majority, treated as structural, receive limited scrutiny. The National Audit Office (NAO) has also noted that cost estimates for many reported reliefs are missing or outdated, constraining effective oversight.
Evaluation remains the weakest link in the policy cycle
Across all four countries, and globally, evaluation is the least developed stage of tax expenditure governance. Although an increasing number of countries publish revenue forgone estimates, systematic assessment of whether tax expenditures achieve their stated objectives remains rare.
In Brazil and France, individual tax expenditures are occasionally reviewed, but these are not embedded in regular, criteria-based evaluation frameworks. Sunset clauses exist in some cases but are inconsistently applied and rarely linked to performance or ex-post evaluations. As a result, many provisions remain in place for years without rigorous scrutiny of their effectiveness or distributional impact.
In India, independent research on tax expenditures is limited, although the Comptroller and Auditor General conduct some assessment through statutory audits. Even so, systematic evaluations are uneven, and findings do not consistently inform policy decisions.
The United Kingdom evaluates selected tax reliefs, but not within a comprehensive evaluation framework. A 2024 report by the NAO found that only a small fraction of identified reliefs have been evaluated since 2015, with limited progress following the 2021 evaluation plan.
Reforming tax expenditures and political economy constraints
Reforming tax expenditures is inherently political. Benefits often accrue to small, well-organized groups, while costs are dispersed across the broader tax base—creating strong incentives for ineffective measures to persist. Ultimately, tax expenditure governance is less a purely technical exercise than a matter of institutional design and political choice.
Supreme audit institutions can play an important role in exposing weaknesses, as experiences in France and the United Kingdom illustrate. Clear benchmarks determine which provisions are visible and which remain outside scrutiny. Comprehensive reporting makes tax provisions legible within the budget process. Systematic evaluation—still largely missing—is essential to assess value for money.
At a time of mounting fiscal pressures, rising debt, and expanding demands on public finances, outlays delivered through tax expenditures are increasingly difficult to justify without robust evidence of effectiveness. Subsidies provided through the tax system should be subject to the same scrutiny, evaluation, and fiscal discipline as direct spending to safeguard fiscal sustainability and maintain public trust.
