TAX GAP : say what?

We've all heard, read and discussed about those numbers. They have been massively relayed through and by specialized bodies as well as by civil society and political bodies, and are repeatedly quoted in large audience media: the profits that are artificially shifted by Multi-National Enterprises (MNEs) and their impact on the the sustainability of government finance and -by extension- on public policies.

In these complex times, this important issue righteously gains traction again: not only in light of past budget limitations on the public healthcare sector of the economy but also in the perspective of the increased budgetary pressure economies might face in the aftermath of the current crisis.

I have to admit that despite my 20 years in Transfer Pricing I never devoted the necessary time to carefully review the numbers that stemmed from extremely serious and complex studies. That has probably something to do with this:

Those numbers that we either embrace or reject result from tremendously complex studies aiming to appraise the difference between the Tax MNEs should pay vs. the taxes actually paid, the TAX GAP. You have also observed that the amplitude of those numbers can be of 1 to 5 in many cases, reflecting differences in methodological aspects and on what is being measured.

Before you rush to your "back" button bear with me for a second: this is not going to be about equations. This is more about perspective because these numbers have fueled the debate on the necessity of the an international tax reform, and have shaped the political agenda of many stakeholders. Thus the necessity of rapidly reviewing some features on Tax Gaps to better understand their reach.

TAX GAP: falling down the rabbit hole

The corporate income tax gap (CIT Gap) is the gap between corporate tax revenues as they “should" - in theory be collected and as they “are” collected. The gap is an indication of potential CIT revenue losses. Three main sources of Tax Gap:

A conceptually simple metric, and as often with those, a tremendously complex concept to properly quantify, or approximate at best. One of the main limitations of all methods is the availability of relevant data. In absence of data with the required degree of granularity, economists and tax administrations have developed several approaches through proxies. A comprehensive review of the available methods developed over the last years for the assessment of Corporate Income Tax gaps was prepared by the Fiscalis 2020 Tax Gap project group and can be found here.

Models can be either top down models or rooted in firm specific data that is more difficult to obtain (explaining why mainly tax administrations use this approach).

Typically, Top Down methods tend to provide significantly accurate results on VAT and to some extent on Personal Income taxes. Applied to CIT, Top Down methods require a substantial interpretation effort on the linkage between macro economic factors and taxable basis of MNEs. That interpretation can sometimes lead to the definition of the "true tax" of the tax MNEs "should pay" that actually substantially depart from the existing international taxation system.

In other terms, and in a very superficial way, some TAX GAP numbers can actually include the difference that could exist between a radically different taxation system and the current system. Beyond a traditional policy gap, this is a "paradigm shift" gap that by nature is difficult to insulate from compliance gaps, or from behavioral elements such as tax avoidance or tax evasion.

My point here, that will be explored in a further section, is that most of the times, studies carried by scholars are fairly clear on what they measure and on their limitations, while the communication based on those studies is seldom nuanced, to say the least.

VAT, Personal Income Tax and Corporation Tax

A second step in building some perspective is not to consider Corporate Income Tax in isolation, but in relation to the economy and actually to the other taxes that compose a country budget. This is not to artificially devalue the questions on MNE's profit shifting but to better understand the entire problem of Tax Gaps.

The UK's tax administration (HMRC) is one of the few official bodies to publish every year Tax Gaps analysis. The latest edition was based on 2017 and 2018 data compiled and processed by HMRC, the results of which can be summarized as follows:

The tax gap in Corporation Tax of Large Businesses (+250 employees -100 if foreign owned - and more than 30 M ￡ of turnover) represents 3,7% of the total Tax Gap in the UK. As previously mentioned, HMRC calculates this gap with a bottom up approach, using a degree of extrapolation and modelling.

Such exercise for other countries is made difficult by the fact that -surprisingly- most EU countries do not engage in such assessments for CIT. In order to derive aggregates for France, it is necessary to pool different ressources. Two European studies - EU taxation papers WP76-2019 and a study prepared for DG TAXUD based on respectively 2016 and 2017 data- suggest tax revenue losses VAT of 12,5 Bn € and of 10 Bn € in Personal Income Tax. In contrast, tax revenue losses generated by MNE profit shifting would range from 2 bn€ to 8 Bn € depending on the methodologies and scope of studies (other studies would even suggest a gap of 20 Bn €).

The issue at stake in measuring Tax Gaps is to provide additional guidance on base extension initiatives, prioritize remediation actions and measure the adequation of regulations and their enforcement. Measurement is key, and when it comes to Profit Shifting and CIT gaps, as suggested in the review of French data, we (public opinion and practitioners) are left with more questions than answers.

Two factors play in my view a role: i) the lack of nuance in communication on numbers and ii) the large differences in estimates. While the latter can be justified by scope or method differences, the former tends to shake the credibility of the estimates and -more importantly- prevent a balanced dialogue. Below are comparisons on the CIT impact of profit shifting on 5 studies (source : Jalansky & Palanski 2019):

Worth mentioning as well, the OECD study from which stemmed the BEPS estimate of revenue loss in the range of 100 Bn / 240 Bn € (Johansson, Skeie, Sorbe & Menon 2017) based on 2010/2016 data for all OECD and G20 countries (a loss of CIT revenues ranging from 4% to 10% of overall CIT revenues).

"A wise man proportions his belief to the evidence" (David Hume).

And here comes the problem: while there is undebatable evidence of Profit Shifting and Base Erosion, is there enough evidence to believe that it cannot be fixed through i) the implementation of the BEPS derived framework, ii) the reinforcement of both anti-abuse rules and double taxation mechanisms and (often forgotten) iii) increased capabilities of tax administrations?

(I am on purpose not engaging in the debate whether it's easier to finger-point MNEs rather than individuals and small businesses in an already complex political environment).

TAX GAP "should be paid" vs. "actually paid" : when tax gaps measure "something else"

Reaching the moment when wheels touch ground. The easiest way to approach this is by taking some examples of studies and identify - in a very simplistic way, I can acknowledge that - the main drivers utilized to derive the profits that "should" be paid by MNEs.

Scholars and analysts have been hoping that the Country by Country Reporting could be an interesting set of data to analyze. A recent report published by Cobham, Garcia-Bernardo and Bou Mansour (April 2020) quote a study based on CbCR data on US multinationals and conclude that by the effect of US MNEs profit shifting EU countries have surrendered 27 Bn € of tax revenues.

While the details are much more complex, the basic principle seems to be the determination of "true profits" based on CbCR metrics such as Sales, Employees and Assets (obviously not including intangible assets as they are anyway not reported in CbCR and are probably considered as the vehicle of profit shifting). I am very conscious of the shortcut, but in my view, a very significant share of the gap cannot be qualified as tax avoidance because it relates to the comparison of the tax system as it is (with probably its share of avoidance) to a fundamentally different paradigm closer to Global Formulary Apportionment/CCCTB. The Report is not misleading at all as the agenda is explicit, but it is difficult to engage into a constructive dialogue on the basis on those numbers or the approach because it simply disavows all the components of value creation that are pivotal in the BEPS framework.

Other studies use as a proxy the differential of profit reported by MNEs affiliates vs. the profits reported by domestic operators with similar economic attributes, and the difference between the results - corrected to account for different biases- corresponds to profit shifting (recognized or unrecognized). Some studies accurately take into consideration productivity gaps to correct the derived outcomes.

The point should then be: yes there is profit shifting, including and not limited to profit shifting to favorable tax jurisdictions coming from high tax jurisdictions. All the studies evidence this, and I believe no practitioner in its right mind would deny it. The main question that these Tax Gap studies are actually raising is whether Tax Gaps that are based on profitability differentials, or based on allocations on employees or assets, are justified or not.

Indeed, one can claim that the surplus of profit once the production factors are equalized, are either attributable to additional factors such as intangibles or network effects - or are instead to be re-distributed with an arbitrary -though maybe politically acceptable- allocation key (employees, assets or sales). However, with the current framework, the allocation of residual profits ("non routine profits" in TP language) is to be performed in light of value generation considerations and the contribution of the different functions and assets behind the value generation (that generate the value and control the risks). As such, many Tax Gap - while capturing it - are not as much about tax evasion or tax avoidance, but about a change of regime.

In other terms, this is not about questioning the Tax Gap studies, but instead ensuring that we are asking ourselves the right questions. In my view, Tax Gaps could accurately indicate what the tax revenues would look like if nothing else could justify the allocation of the surplus to "other" entities or jurisdictions, as if any other claim on profits would be unsubstantiated. But let's be serious, put aside crass profit manipulation schemes that probably still exist or have hit the press, that is not the environment most Tax and Transfer Pricing practitioners - and the firms they are working for - evolve in.

OECD and Pillar 1/2?

This is actually why proponents of the Global Formulary Apportionment or CCCTB have been so vocal against the latest OECD move on international taxation and transfer pricing, because it does acknowledge - implicitly though- that a quite significant part of the Tax Gap can be resolved through the BEPS framework (control over risk, DEMPE functions, Intangibles returns and ownership). This is perfectly illustrated in the impact estimates of Pillar 1, that have a somehow modest impact on tax revenues redistribution: mainly sourced from "investment hubs", or Principal Companies and distributed to markets, in a modest proportion with significant potential impacts for mid and low income economies.

It is actually Pillar 2 (minimum tax rate - GLOBE) that yields the most important shift in tax revenues, and the implementation of minimum tax rates might indeed generate behavioral change in a less - but still existing - tax rate distorted landscape.

One of the most far reaching outputs of the OECD proposals mainly concerns the digital economy and the impact on sales platforms by creating a de facto nexus, itself leading to a reallocation of taxing rights.

Nevertheless, the OECD proposals should not dramatically affect US-type structures operating through the intricacies of 1.482-7 Regulations of Cost Sharing Agreements and check the box structures. The most significant issue at stake for those companies will remain between the recipient of the Cost Shared rights and the country that enacted the Regs. The issue might at the end only marginally depend on the OECD post BEPS guidelines that in many cases generate an adequate answer to whether the "tax gap" is properly allocated or not (talking here about something that is actually closer to policy gap than tax avoidance or evasion).

That is why the most important question on Tax Gaps is not necessarily the number, or the methodology, but the degree of entitlement of other entities to those profits. It is perfectly acceptable to reject the current framework under which MNEs and Countries operate and call for a change of framework (GFA), but in that case, Tax Gap studies have to improve and properly delineate what is a policy gap, a tax avoidance and tax evasion gap, and to insulate a Tax Framework gap. To provide a fair picture of the current and hypothetical tax landscape.

The issue is that this is not always how communication around tax gaps is structured, preventing a balanced discussion about the accuracy of value drivers and value chain in MNE taxation, concepts that for sure are not yet perfectly appraised and implemented, by taxpayers and tax administrations. But concepts that deserve to be given a chance, both on policy levels and on the public debate.