The centre for tax analysis in developing countries

Overseas Development Institue Institue for Fiscal Studies

Tax expenditures (TEs) are used widely around the world. Their role in the fiscal systems of low- and middle-income countries has recently attracted increased scrutiny. At the heart of this scrutiny lies a desire to understand the extent of revenue foregone due to TEs, often in countries that have not undertaken such an exercise before. While a policy-maker preparing a TE report for the first time will likely find ample guidance on the broad principles (either with dedicated technical assistance or through ‘how-to’ literature), there are myriad practical, often country-specific hurdles to overcome when preparing estimates.

This paper discusses the key steps and challenges in preparing a TE report. It draws where relevant from the authors’ experiences of supporting TE reporting in Rwanda and Uganda, two countries that have only recently begun the practice of regular TE reporting and have had fairly contrasting experiences.

Specifically, we focus on the following four key areas, and provide a number of lessons learnt:

1) Defining tax expenditures and the benchmark tax system

Arguably the most challenging part of producing a TE report happens before any estimation has begun; defining a benchmark tax system (BTS) against which to measure deviations is a practice that is carried out quite inconsistently across countries. We first discuss the conceptual differences between benchmarks defined according to the (i) legal and (ii) normative approaches. Next, we explore in more depth a number of criteria that might serve as a useful guide for policy-makers when thinking about which elements to include or exclude from the benchmark and, in cases where discretion is employed, justifying why a provision is or is not costed. Ultimately, we stress that there is significant transparency value in reporting on a wider range of TEs.

Lesson #1: To underpin a robust definition of the BTS, it is important to adopt (carefully) – and make public – a national definition of what constitutes a TE.

Lesson #2: Valid arguments may exist for including a ‘deviation from the norm’ in the BTS, while arguments for dealing with ‘grey areas’ may be weak. But identifying and applying criteria consistently and documenting them help inform decisionmaking and are important for transparency.

Lesson #3: Separation of ‘structural reliefs’ and ‘tax expenditures’ can help to overcome confusion about grey areas and broaden transparency by reporting revenue foregone from all provisions that represent deviations from the norm. However the benchmark is defined, local ownership is key.

2) Estimating revenue foregone

The estimation of revenue foregone is far from straightforward and the types of data and modelling techniques employed will differ from country to country. We explore how analysts should best identify the data required for costing revenue foregone and how even relatively poor or aggregated data can be used to produce findings. For each of the major tax types – namely income taxes, value-added tax and customs and excise duties – we discuss some specific challenges and outline appropriate modelling techniques for some common types of TE. Finally, a number of practical challenges in the estimation process are discussed, with some guidance on how these might be best overcome.

Lesson #4: Estimating revenue foregone is heavily dependent on data availability and quality. Administrative data is often a core requirement, but other sources, such as supply-and-use tables and survey data, can provide important complementary information. When aiming to build evidence for policy-making, TEs estimated at the provision level can be more informative.

Lesson #5: There is no one-size-fits-all approach to modelling revenue foregone under different tax types. Often the ‘best’ methods are those that utilise the available data and resources to their full potential. 

Lesson #6: Starting simple and making improvements to modelling over time, at a pace that increases as institutional knowledge is built up through repetitions, can enhance the overall sustainability of the TE reporting process.

3. Report writing and communication of revenue foregone estimates

The content and level of detail of TE reports can vary across countries. While this often reflects the underlying motivation for producing the report, a number of norms are emerging regarding the contents of a TE report that provide the most transparency. Reporting on dimensions beyond the estimate of revenue foregone by including details on, for example, the legal basis for provisions, who the intended and actual beneficiaries are and so on, can provide a more holistic view of TEs. We also cover issues related to how TE reports are communicated and discuss some common misconceptions.

Lesson #7: Communication becomes the key bridge between the technical exercise and the TE report’s objectives around greater transparency and informing policy. In this process, preventing certain misconceptions is paramount.

4. After the TE report: towards better governance of tax expenditures

The publication of a TE report not only represents an important gain in transparency of public expenditure, but also can serve as a springboard for further work that seeks to understand the costs and benefits of TEs and enhance their governance more broadly. In contexts where governments are pressed for revenue to fund development needs, a set of processes that attempts to rationalise the existence of current reliefs and subjects to scrutiny any new proposed reliefs can bring further gains in the transparency and governance surrounding their use.

Lesson #8: Estimating revenue foregone is merely the first step to understanding the ultimate economic impacts of TEs. A TE report tells little of the net benefits brought about by tax reliefs and how these compare to alternative policy choices.

Skip to main content