A.4. "Scoreboard of indicators" to identify third...

A.4. "Scoreboard of indicators" to identify third countries subject to further evaluation

A.4.1. General approach

Tax avoidance can occur through three main channels (debt shifting, transfer pricing and location of intangible assets) and often involves specific financial constructions, often located offshore. Third countries may be used in some type of aggressive tax planning structures but not in others 97 . Furthermore, tax avoidance is a complex and multi-faceted phenomenon, which needs to be looked at from various angles. It is therefore important to rely on a range of indicators that are able to reflect various types of structures.

It is proposed to consider three dimensions:

- Financial importance: Third-country jurisdictions that are used in ATP structures should record (abnormally) high financial flows (such as interest income or royalty income that allow artificially shifting profit, or through the establishment of numerous foreign-owned subsidiaries for tax purposes). They are also likely to be specialised in trade in financial services to non-residents. 98 This category of indicators will therefore focus on the financial "symptoms" associated to ATP, both in absolute terms (the biggest financial centres), but also relative to the size of real economic activity.

- Importance of economic ties with the EU: This category of indicators aims to focus attention on those jurisdictions with which the EU has strong economic links.

- Institutional and legal factors: This category of indicators will reflect factors that facilitate or are necessary for engaging in ATP. The economic literature identifies as a main and obvious characteristic of a tax haven its very low rate of business taxes. Other findings relate to good governance and structure of the tax systems. 99 The ATP Study identifies additional indicators such as the absence of withholding taxes on interest, royalties or dividends. It also identifies which tax rules and/or practices those offshore jurisdictions would need to have in order to facilitate the setup of an aggressive tax planning structure.

It is important to stress that each potential single indicator cannot in itself suffice to draw conclusions on whether a country should be prioritised for screening. For example, most indicators relative to the "financial importance" of a jurisdiction do not allow to disentangle BEPS-driven activities from real economic activities. Similarly, it should be stressed that some indicators, for example in the category of "institutional and legal factors", may very well serve a positive function in the organisation of the country but are "abused" by aggressive tax planners. This is also true for the other categories of indicators. Therefore one needs to look at a range of indicators before reaching any conclusion.

Finally, the preselection of a third country does not rest on the requirement to have all indicators "in the red". On the contrary it is only necessary for a minimum of warning signals to be on. The aim of the information gathering exercise in the scoreboard is to enable a pre-selection. Therefore the adverse consequences of forgetting a potential jurisdiction that should be screened ("false negative") are much more important than the ones linked to the preselection of jurisdictions that only stand out for perfectly legitimate reasons ("false positive"), as there will be further checks and assessments carried out in the rest of the process.

A.4.2. Data selection and limitations

An indicative list of indicators which might be among those considered for the first step screening can be found in the table below.

The preliminary choice of the indicators has been guided by their relevance in identifying jurisdictions that merit further screening, and was focussed on elements for which information is available in a comprehensive and comparable manner. The chosen economic indicators are all based on macro-data for which the coverage was found to be more comprehensive and comparable than that for micro data (i.e. at company level or from tax receipts) 103 .

Coverage may differ per country and per indicator, with some risk of sampling bias, if the analysis would be limited to countries for which the dataset is complete. The use of a scoreboard with possible missing values already alleviates that risk in comparison with the use of a single number or formula. Furthermore, as missing data can occur for spurious reasons, but may also be an indication of secrecy (or of insufficient administrative capacity), jurisdictions for which some or most of the indicators are missing cannot simply be ignored from the analysis. On the contrary, those jurisdictions will be specifically indicated for further consideration.

With respect to the choice of which data vintage to use, there is a trade-off between using the most up-to-date indicators, and being misled by one particular data point corresponding to specific circumstances, in particular for very volatile data. Using averages of recent years is an alternative – this would also reflect more structural characteristics of countries. 104

A.4.3 Use of the scoreboard for setting priorities for further assessment

At the level of each indicator taken separately, ranking countries per indicator is obviously the best way to allow for later prioritisation. ( 105 ). This will be done for each indicator and for all jurisdictions, but jurisdictions with missing values will be marked separately.

In a second step, when aggregating or selecting among closely related indicators, recourse is made to the general guiding principle explained above: "false negatives" - third countries not being preselected while they would deserve to be further scrutinised - should be avoided. To avoid this risk of "false negatives", it is therefore suggested to take only the "worst" indicator (and not the average indicator for instance). In other words, the indicator reflecting the highest risk to be used for ATP purposes would be retained. To give an example related to the selection of the relevant time period: One indicator, such as the relative importance of FDI flows, will be ranked independently over several different time periods, corresponding to several different variations of the data: one for each of the last years and also the average of recent (10) years. Out of those rankings, the scoreboard will only reflect, for each jurisdiction, the one ranking based on the most suspicious outcome (in this case the highest of the rankings, which will be the one based on the time period, or on the average, for which the relative importance of FDI flows was the highest compared to other jurisdictions).

Finally, and for reasons linked to the phenomenon we try to apprehend, but also given the possible data limitations of which the most important is incomplete coverage, no rigid combination of the potential strands of indicators into a single composite indicator is currently foreseen. In addition to the necessity of perfect data coverage, this kind of methodology obviously entails numerous choices and selections, some of which are unavoidably of an arbitrary nature, and therefore predictably open to criticism. Rather, the exercise is one of gathering multi-faceted information along the three dimensions outlined above. This comes in contrast to the approach taken elsewhere, for instance by the Tax justice network in the recent release of its "financial secrecy index" which builds numerous indicators into a single number, allowing for a single listing of a total of around 100 jurisdictions but for a different objective, and a slightly narrower scope. 106 107

Finally, the scoreboard is seen as a tool for priority setting and therefore does not formally set a limit on the number of jurisdictions subject to further scrutiny.