Tax expenditures are tax provisions (i.e. exemptions, non-standard deductions, credits, reduced rates and deferrals) within the tax system that is in place that deviate from a benchmark tax system. The term tax “expenditure” (TE) arises from the fact that they are equivalent to public expenditure implemented through the tax system. When properly designed, TEs can stimulate economic growth and improve well-being. However, TEs often come at a significant tax revenue cost and may also raise distributional concerns and create distortions.

In 2021, the Tax Incentives Commission presented an in-depth review of tax incentives and tax expenditures within the Colombian tax system. The Commission concluded that there is a systemic over-use of TEs to correct for structural deficiencies in the tax system. It advised Colombia to focus on fundamental tax reform and break the undesirable spiral of continuously introducing poorly analysed tax expenditures. The Report also encouraged Colombia to define a “benchmark” tax system relative to which TEs are identified. According to the Commission, the absence of a well-defined benchmark tax system implied that some TEs remained unidentified while other tax concessions that were reported as a TE were in fact not a TE; this resulted in a biased estimate of the total revenue foregone of TEs that was reported by Colombia. The Commission also recommended that Colombia publish a standalone annual TE Report that would list all TEs and quantify the associated foregone tax revenues. In addition, the Commission noted that there was scope to improve the estimation methodology and the data that was used to measure tax revenue foregone. By allowing for fiscal transparency, the stand-alone TE report would enhance government accountability and support evidence-based decision making on the allocation of public resources.

Since the publication of the report of the Tax Incentives Commission, Colombia has made significant progress with respect to the measurement of the revenue foregone of TEs. Estimates of revenue forgone from TEs are presented in the Annex of the Marco Fiscal de Mediano Plazo (MFMP), an annual report published by the Ministry of Finance, which covers TEs from the Personal Income Tax (PIT), the Corporate Income Tax (CIT), the Value Added Tax (VAT), as well as fuel excise taxes and the carbon tax. In the last two years, the range of TEs that has been estimated has considerably increased.

Nevertheless, there remains significant scope to further improve the identification, measurement and reporting of TEs in Colombia. For instance, Colombia has not yet defined a benchmark tax system, and therefore has not been able to publish a coherent and complete list of all TEs within the tax system. The most recent edition of the MFMP provides aggregate estimates of revenue forgone, but it does not present revenue foregone of TEs on an item-by-item basis. Moreover, the revenue forgone arising from non-taxable income and from special CIT deductions (other than the deduction for investment in fixed assets) is not reported in the MFMP.

This Review provides further guidance on how to identify, measure and report TEs in Colombia. As such, this report provides useful input into future tax reform discussions and evaluations of the tax system in Colombia. It deepens the recommendations of the Colombian Tax Incentives Commission in 2021, and fills in the remaining gaps so that Colombia can start publishing a standalone TE report as soon as possible.

This Review provides estimates of the tax revenue forgone of CIT and PIT TEs on an item-by item basis. The OECD has estimated the revenue foregone of a wide range of TEs based upon two samples of depersonalized tax returns and exogenous information data that were provided by Colombia’s Tax Administration (DIAN) for the purposes of this review. The OECD made a proposal for the formulas that can be applied to estimate the TEs for which data is available. These formulas have been applied by DIAN to all taxpayers that file a tax return, which have been merged with the corresponding exogenous information forms, if applicable. The revenue foregone estimates as well as the distributional analysis of a number of TEs calculated by DIAN have been included in this Review.

The Review defines an income tax benchmark for Colombia. The proposed “hybrid” income tax benchmark is based upon a theoretical concept of income that limits arbitrary decisions about what constitutes a TE, but is nonetheless adapted to include structural aspects of the Colombian tax system. Given the proposed TE benchmark, the review then seeks to identify all TEs within the CIT and PIT. In particular, the report lists 123 CIT TEs and 124 PIT TEs that have been identified in collaboration with Colombia’s Tax Administration (DIAN) and its Ministry of Finance (MHCP).

The Review also presents a benchmark for the VAT in Colombia. The benchmark VAT base consists of final consumption plus gross fixed capital formation by private households, the government and non-profit organizations; the VAT is levied on a destination-basis. Under the VAT benchmark, the VAT paid on gross fixed capital formation by businesses is recoverable. This gives rise to a negative TE within the VAT, as this is currently not the case in Colombia. This negative VAT TE is likely large, but it has not been reported previously. The current exclusion of VAT on goods and services that are taxed under the “National Consumption Tax” also results in a VAT TE.

The Review makes several technical recommendations on how to improve the estimation methodology of VAT TEs in Colombia. In particular, the Review presents a methodology that would allow Colombia to differentiate its estimate for VAT non-compliance by economic sector in order to better disentangle TE revenue forgone from revenue losses due to tax evasion. Colombia could also start measuring the negative TE linked to the non-recoverability of VAT paid on investment in fixed assets. Significant scope exists to improve the measurement of revenue foregone from the VAT exemption within Free Trade Zones.

The Review describes the approach that DIAN currently follows to measure and report CIT and PIT TEs, and identifies areas for improvement. The key priority regarding the measurement of CIT and PIT TEs is to improve the granularity of the available data. Currently, CIT and PIT returns are not sufficiently detailed to identify revenue foregone of TEs on an item-by-item basis. In addition, the tax returns require taxpayers to file aggregated information that combines tax provisions that are TEs as well as provisions that are part of the benchmark, and are therefore not a TE. The tax returns will have to be changed such that these different types of tax provisions are filed separately.

The number of taxpayers that are required to submit exogenous information forms when they file their CIT return will have to increase, as long as more granular information is not available within the tax returns. In 2020, corporations that were required to submit these additional forms (i.e. with turnover above COP 100 million) constituted only 44% of all corporations. Reducing the COP 100 million turnover threshold would allow DIAN to report more representative item-by-item revenue forgone estimates. Exogenous information forms should also be redesigned to allow for more granularity. Only about half of all identified CIT TEs can be quantified based on the exogenous information that is currently available. DIAN is also encouraged to explore information that might be available in other government agencies and that could possibly be merged with tax returns data as well as to collect and use information for TE measurement purposes that can be provided by third parties, such as the financial sector.

The Review makes additional suggestions regarding the calculation of revenue forgone from CIT and PIT TEs. The Review recommends to move from the application of standard CIT rates to company-specific CIT rates (which might be lower or higher, depending on whether the company is subject to a special regime) to calculate tax revenue forgone. Furthermore, the Review recommends that DIAN report separately the interaction between reduced rates and tax base narrowing provisions that constitute a TE. To estimate revenue forgone from base narrowing TEs, DIAN is also encouraged to switch from applying an average marginal PIT rate (to the value of the TE) to applying the actual PIT rate schedule (to the sum of taxable income and the TE and then take the difference with the current tax liability) for each taxpayer.

The PIT return will have to be redesigned in order to estimate the revenue forgone from PIT TEs on an item-by-item basis. Less than 2% of the taxpayers that submit a tax return are required to submit supplementary tax forms that provide more detailed information on the tax provisions they claim. Therefore, the usefulness of exogenous information for the measurement of PIT TEs is severely limited. As a consequence, revenue forgone estimates from PIT TEs need to rely extensively on data from the tax returns and only 10% of PIT TEs can be estimated on an item-by-item basis with reasonable accuracy. In addition to ensuring that provisions that are viewed as TEs are not reported together with provisions that are considered part of the benchmark, DIAN should collect more disaggregated information within the tax returns that taxpayers are required to file. DIAN should explore additional sources of information to estimate certain TEs, such as information provided by pension funds on voluntary pension savings.

This Review concludes by suggesting a structure for a future standalone TE report and includes a set of clarifications that could be added in the report to avoid any misinterpretation of the results. The first chapter would define the benchmark for all taxes covered in the TE report, and describe the main assumptions underlying the revenue forgone methodology. The second chapter would summarise the changes that have been made since the prior edition of the report with respect to the definition of the benchmark, the methodology that is applied to measure revenue forgone and the data that is available as well as changes in tax policy. The third chapter would present a summary of the main revenue forgone estimates by type of tax and main types of TEs. The report would include a separate chapter for the TEs within the CIT, PIT and VAT, respectively. A separate chapter for TEs within other taxes could possibly also included. The full list of TEs would be included in an Annex of the report.