5. Drafting the First Tax Expenditure Report – Structure and Important Clarifications

This chapter presents recommendations for the outline of the TE report that Colombia is planning to publish on a regular basis, following best practices in other OECD countries.1 The TE report should be a standalone report and not an annex of the MFMP. The chapters of this report could be used as a template for the outline of the TE report and for the specific topics it covers. In particular, Colombia’s TE report could include the following chapters:

  • The first chapter would describe the tax benchmark for all taxes that are covered in the TE report, and in particular the CIT, PIT and VAT, and include a discussion of the standard benchmark principles. The chapter would define key concepts such as TEs, tax benchmark and revenue forgone and describe the main assumptions underlying the revenue forgone method and provide guidance on how to interpret the revenue foregone estimates. It would also briefly discuss the type of data that has been used when estimating tax revenue foregone.

  • The second chapter summarizes changes regarding tax policy, the benchmark and the methodology applied to measure revenue forgone compared to the previous TE report. The chapter would first list relevant tax policy reforms introduced in the last year that lead to the creation of new TEs or to the abolishment of existing TEs. It would then zoom in on the changes made to the benchmark compared to the previous editions of the report. This is relevant as it will have an impact on the tax provisions that are identified as TEs. In addition, the chapter could discuss major changes to the methodology applied to measure revenue foregone. Any changes to the data sources that have been used compared to previous editions of the report could also be presented.

  • The third chapter would present a summary of the main findings; it would include the main revenue forgone estimates. This chapter could include a summary table that lists the revenue forgone by type of tax and main types of TEs (Table 5.1). This chapter could also include a table that lists the largest TEs and their revenue forgone.

  • The following chapters would zoom in on the TEs within the main taxes covered in the report. The report would include a separate chapter for CIT TEs, PIT TEs, VAT TEs and, possibly, TEs within other taxes (including excise duties, property taxes, etc.).

  • The report would then include annexes with the details that would be useful background information.

Some clarifications that would be useful to add in the first chapter of the report include:

  • The fact that a tax provision is identified as a TE does not mean the provision needs to be removed or reformed. The TE report aims at providing a comprehensive overview of deviations from the benchmark tax system, and the associated revenue forgone. From a tax policy perspective, certain TEs may be well justified.

  • The fact that a tax provision is a TE does not mean the associated revenue forgone can be quantified. For some provisions, measuring revenue forgone is difficult or impossible because the data that would be required to measure the TE is not available. However, concerns regarding costing are not supposed to impact whether a provision is considered a TE or not. Whether or not a tax provision is a TE is solely guided by comparing the current tax system to the benchmark tax system. If the revenue forgone of a provision cannot be estimated, the provision is nonetheless included in the comprehensive list of TEs in the annex to increase transparency.

  • Low revenue forgone does not necessarily mean that the distortions caused by the TE are small. TEs can result in significant distortions that do not show up as revenue forgone.

To reduce the risk of misinterpretation, the report should remind the reader of the limitations of the revenue foregone method that is typically applied to quantify TEs.2 As a static concept, revenue forgone does not take into account the potential behavioural responses by businesses and consumers to the removal of a TE.3 In particular, taxpayers may change their behaviour when a specific TE is abolished to, possibly, benefit from another TE. Revenue foregone is therefore not the same as revenue gain.

This subsection list key recommendations regarding the reporting of revenue forgone:

  • The main focus of TE reporting should lie on revenue foregone of the TEs. Less attention should be given (if at all) to the total amount of exempt income and deductions, as the revenue foregone of specific TEs will depend on the tax rate that applies to that specific item.

  • Revenue forgone could be summarised as in Table 5.1 and then discussed by type of tax, in line with the reporting suggestions included in the VAT, CIT and PIT chapters. The tables should include both item-by-item estimates and totals by type of TE.

  • Values would be expressed both in thousand million pesos and as a percentage of GDP. Previous year estimates could be included if estimates for more than one year are available. If estimates for the latest year are projected this should be clearly stated in the notes to the tables.

  • For special regimes, the table should include the total number of firms that are subject to each regime as is currently done in the MFMP (Minhacienda, 2022[1]).

  • The tables could possibly include revenue foregone for several years. This would allow for an analysis and discussion of the changes over time. However, the reader of the TE report should be made aware of the fact that changes in the benchmark would induce changes in the TEs over time. In addition, if revenue forgone estimates for two consecutive years are based (with adjustments) on the same underlying data source, the change in the estimates will only capture the change in the design of the TE and the tax value that corresponds to this tax provision. It will not capture the extent to which taxpayers have taken up the tax provision.

  • An overall total of all measured TEs could be provided as part of Table 5.1 (8 out of 16 OECD countries reviewed in this report provide such a statistic) even though the resulting figure should be interpreted with caution.4 If the total is provided, the report would explicitly mention that the revenue forgone concept does not take into account the interaction between individual TEs, and that some TEs could not be included in the total because the necessary data was not available.

  • A breakdown of CIT and PIT revenue forgone by agents and/ or economic sectors would ideally only be reported once all types of TEs are measured and aggregate revenue forgone from each type of TE can be measured accurately. Otherwise, the tables would present biased results across types of agents and economic sectors.

Following the format recommended in Table 3.3 and Table 4.4, the individual chapters for each specific tax would include separate tables for each TE that has been measured, including the legal reference, the reason why the provision deviates from the benchmark, the revenue forgone estimate as well as the other information such as the methodology and data used to estimate it.

Leveraging the potential of the tax return microdata, the individual chapters for each specific tax could include a distributional analysis of specific TEs. More guidance has been included in the CIT and PIT chapters.

The annex to the report could include a comprehensive list of all TEs, regardless of whether they have been quantified or not. In addition to their legal reference, these tables could point out why these TEs are not part of the benchmark. Annex C, Annex D and Annex E provide the list of TEs that have been identified in Colombia in the income and VAT. Reduced rates from withholding taxes should be reported in a separate table rather than being included under the income tax.

The annex of the report could include estimates of the tax provisions that are part of the tax benchmark but that are nevertheless interesting to quantify in terms of their foregone tax revenue. For example, even though the zero-rate bracket or the deduction for dependents under the PIT do not constitute a TE, their foregone revenue would be useful input in the tax policy debate. The foregone revenue of these tax provisions could therefore be presented in the annex of the TE report. The same could apply to exempt income from Andean Community treaties.


[1] Minhacienda (2022), Marco Fiscal de Mediano Plazo 2022.


← 1. Annex A summarizes practices in other OECD countries.

← 2. It is standard practise to quantify TEs by means of the static revenue forgone method. The alternative method (tax revenue gain) strives to take into account behavioural responses but its results depend on reliable estimates or assumptions about how precisely taxpayers would react to removing a TE. Regarding VAT, this would for example imply having access to estimates on price elasticities (ideally on the product and industry level), which are not available.

← 3. Two other limitations of the revenue forgone method are that it does not account for interdependencies between TEs and the potential non-take up of a tax relief.

← 4. Table A A.5 reports practices in sixteen OECD countries on this matter.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD 2022

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.