On 28 May, 2019, Paris-based international quango, the Organisation for Economic Cooperation and Development, or the OECD, has published the latest agreement between the Inclusive Framework countries. The agreement provides a road map for future reforms across the Framework’s 129 member countries in the area of harmonisation of international taxation rules applying to the multinational corporations (MNCs). The document places a special emphasis on companies trading in the digital economy, augmenting the broader Framework reforms that aim to reshape international tax rules for all MNCs. As such, it defines the key options for sweeping changes in tax regimes going forward.
The OECD efforts to drive harmonisation of taxation rules are not new. The organisation has steadfastly pushed for a drastic revamping of the international tax laws, invariably aiming to drive forward the creation of a new global bureaucracy charged with eliminating “harmful” tax competition and stripping away the core basis of the modern taxation system – the national level autonomy in tax policy. As such, the latest proposals represent not only an assault on the sovereignty of the national fiscal policies, but a dangerous push toward greater monopolisation and monopsonisation of the global economy.
Paving the road to a global minimum tax rate
The latest OECD report outlined three options for profit attribution rules, covering “digitally-trading” MNCs. They involve a modified split, a fractional apportionment and a distribution-based option.
The first of the three, the modified residual profit split, will require calculating corporate non-routine profits (or profits in excess of “norms”) and allocating these to different jurisdictions for taxation purposes. The basis for allocation can either be transfer pricing accounts, or a yet-to-be-defined rules-based approach.
Calculation of this share of profits earned by an MNC would involve assessing the rate of profit earned by a comparable third party and applying it to the affiliates of the MNC. The residual allocation may be based on external market benchmarks or estimation based on costs. Under this mechanism, profits attributable to the subsidiaries and affiliates of an MNC will be treated as if they are earned by an entity that does not share any risks or uncertainties with the MNC, nor benefits from the global nature (economies of scope and scale) of the MNC.
However, multinationals are integrated entities. Basic theory of the firm, formulated first by Ronald Coase, says as much. The activities of MNCs are also integrated at the level of shared risks and uncertainties. The more modern variants of the theory of the firm say as much.
MNCs’ sub-entities are also integrated at the level of sharing benefits, costs, uncertainties and complexities of accessing global financial and inputs markets. The financial theory of the firm says as much. Digitalisation of the economy, and a broader technological evolution, add more complex couplings between risks and uncertainties to modern firms’ strategies and operations. Modern risk management says as much.
Thus, the idea of treating subsidiaries and divisions of MNCs, or even their fully independent partners, as being wholly separable from the main firm runs contrary to the basic tents of modern economics and business practices.
The OECD view under the modified residual profit split proposal means that a more productive MNC, enjoying higher profit margins, can potentially be taxed based on the profit expectations (or average norms) of the less productive competitors operating in the same market. Economics-wise, this is an equivalent of more profitable firms regressing toward the mean, irrespective of whether their profits are generated through tax arbitrage or tax optimisation, or through better, more efficient use of technologies, innovation and/or better business strategies.
In higher tax states, this regression to the mean will harm the better run companies and indirectly subsidise less efficient producers. In lower tax states, more productive MNCs will be transferring funds to higher tax jurisdictions, penalising more competitive economic environments and rewarding less competitive ones.
Let us take a simple example based on “extreme differences” between two hypothetical MNCs operating in the same markets. Suppose, Company A trading across borders has lower quality corporate governance, and is contrasted by Company B, identical in all respects with Company A, but having stronger quality of management and decision-making. Academic literature strongly suggests that, holding everything else constant, Company A will be earning lower risk-adjusted profits in the long run. Yet, the averaging of the two companies profit shares distribution across various jurisdictions, as proposed by the OECD, will impose higher taxation on Company B and lower taxation on Company A, precisely because Company B is “abnormally” more profitable relative to the average of Companies A and B.
The only world where such a possibility makes sense as a basis for policy formation is the world inhabited by the OECD – a quango hell-bent on dressing up higher tax extraction into the garbs of “social reforms”.
The above example does not consider the positive correlation between risk-taking by companies, innovation and profitability. This correlation is hard to ignore in the real world. Smaller, faster-growing companies trading across borders require higher profit margins, due to higher cost of capital, greater performance expectations by the markets, and other factors. These firms are likely to face a tax penalty from the OECD proposals than their larger, less profitable enterprises. Which is why the OECD “reforms” are favouring larger, legacy firms over smaller, more innovation-driven ones: the OECD simply cannot comprehend the simple reality that some abnormal profits are necessary to sustain competition and value creation.
The second option is a fractional apportionment under which the tax authorities will determine the overall profit of an MNC and then allocate this profit across different jurisdictions in which this MNC is trading based on a set of “allocation keys”. This approach would require defining the tax base division process, determining the allocations keys (rules and metrics, based on which the tax base can be distributed across different jurisdictions in which the MNC operates), and the weighting of these allocation keys. The factors that can determine jurisdiction share of the total taxes paid globally can take into account sales, assets and employment figures for the MNC in each specific jurisdiction in which it operates.
The above option, in spirit, closely relates to the third one: the distribution-based approach.
This aims to apply a “simple formula” to determine what baseline profit is attributable to a specific jurisdiction in relation to the intangible activities or assets, such as intellectual property, customers lists, or marketing, distribution and consumer support services.
Defining a simple formula for either the allocation keys or the distribution-based factors, however, is a major problem. All three OECD options involve complex, ambiguous and uncertain definitions of key concepts, rules and metrics, as well as the mechanisms for their imposition, supervision and enforcement. More ominously, all three options are associated with a risk of eroding national sovereignty over tax policies, as well as potentially distorting the markets through asymmetric tax treatment of the MNCs, and smaller internationally trading firms.
The latter point is non-trivial. Costs of adhering to the new rules and associated recurring costs of reporting and compliance of taxable activities are likely to be significant, as increased complexity of tax regimes, rules and bases under the Inclusive Framework will have to be borne out by every company trading across borders. Firms operating in higher margin, more monopolistically-competitive sectors, such as provision of differentiated consumer services (e.g. search engines, social networks and online advertising platforms) should be able to pass at least a part of the new costs onto their end users. Smaller companies operating in more competitive segments of the markets are already trading on thin margins and have no market power to share these costs with the customers. These firms will have to consolidate and/or lobby for market protection in order to survive.
Under the OECD proposals, therefore, rent-seeking will increase, competition will decrease, and the rate of entrepreneurship in the digital services sectors will decline. In the age of already anaemic growth in productivity, this is a dangerously daft idea.
Dividing the pie is not the same as baking it
These arguments go to the heart of the OECD’s “nexus rules” relating to what will replace the current system that imposes tax on the basis of companies’ physical permanent establishment.
So far, we have no significant understanding from the OECD’s Inclusive Framework proposals as to what will replace the status quo.
One proposal is to redefine physical permanent presence concept to include locations where a company “exhibits a remote yet sustained and significant involvement in the economy”.
An alternative is to introduce “a new standalone provision giving market jurisdictions a taxing right over the measure of profits allocated to them under the new profit allocation rules, which would require:
- identifying and defining a new non-physical taxable presence separate from the [physical establishment] concept;
- identifying and defining a new concept of income taxable in the source jurisdiction (i.e. income derived from a particular source in a jurisdiction); and
- the interaction between the new taxable presence or source income and existing provisions…”
Both of these approaches are rich in ambiguity and complexity. Determining any given firm’s contribution to economic activity in any specific jurisdiction is an act of art more than of science. Within this art, doing so on the basis of specific one-type-fits-all factors is more of a theatre of absurd than a docudrama.
Take one example: Apple’s iPhones. Based on the IMF research, Ireland, Korea and Taiwan are now “the main beneficiaries of the new tech cycle in value added terms”. More specifically: “In Ireland, where the intellectual property of Apple Inc resides, staff estimate the contribution in value-added terms of iPhone exports to account for one-fourth of the country’s economic expansion in 2017. At the same time, it is important to note that the income generated from smartphone sales does not fully contribute to the Irish economy.”
This means that, in the case of iPhones-related taxation, Ireland is on the hook when it comes to the OECD-proposed frameworks. But no one, in Ireland or outside, can tell exactly how much of Apple-related economic activity is generated through network effects and other complex and tightly coupled sources of value added. Does Ireland’s tax regime, as opposed to its intellectual property laws, or quality of business environment, or Ireland’s position as the country being a gateway to the massive European market serve as the source of a specific dollar in value added and revenue? In other words, no one can tell just how much of iPhone-related revenues taxable in Ireland is organically attributable to the Irish economy, and how much is attributable to, say, Germany or Belgium who supply iPhone customers.
The nexus rules, just as the rest of the new agreement proposals, are supposed to first apply to digitally enabled businesses, apparently leaving traditional businesses covered by the existent set of rules. The problem, of course, is differentiating digital users from traditional users. Traditional businesses increasingly and seamlessly bundle digital services with other goods and services. With time, the OECD-targeted distinction between digital and traditional businesses is becoming more and more ambiguous in practice.
Take, for example, a hospital providing physical surgery and that relies on digital pre-op and post-op services. Is the hospital a digitally-enabled business or a traditional one? What if the treatment involves nanotechnologies that rely on software that can be updated remotely?
The same applies to car manufacturers who increasingly rely on provision of digitally enabled services bundled into their vehicles, up to the levels of cloud-based, software-enabled and customer-specific autonomous driving. Now, to complicate things even more, add car insurance products that in the near future will have to cover both human error-caused accidents and incidents involving software-operated systems that are sensitive to data flows between vehicles and systems. These services can, and will be, provided from remote locations, but they (and the value they create) will be partially locally based. How does one decouple multiple and highly complex first- and second-order effects of local data, global data, local software, global software, local systems, global systems on local value-creation?
How can one attribute the value created by the self-learning autonomous driving system installed on a German vehicle, administered from Ireland, when the software co-uses data from Belgian and French drivers to optimise vehicle performance for a passenger in India, while simultaneously localising data insights from the Indian market?
In recognising these complexities and their dynamic evolutionary and revolutionary nature, the OECD notes that the entire new mechanism for disputes resolution and arbitration must be put in place to address the evolving nature of new regimes and rules.
This is simply not good enough to be a basis for structural policy reforms. Firstly, dispute resolution systems are ex post interventions in the markets, and cannot serve as a basis for assessing the economic and business impacts of the taxation system. Secondly, the OECD cannot provide a coherent system for such resolution that would be accessible to smaller and medium-sized enterprises when these trade across borders. Dual taxation incidence is currently relatively costly to manage, when it comes to conflicts and disagreements arising at the national taxation level, cumbersome, inefficient and prohibitively priced for smaller enterprises. Once again, the OECD framework provides a de facto bureaucratic barrier to trade for smaller firms, pushing the markets for digitally enabled services into potentially greater degree of monopolisation.
Bad politics, worse economics
The OECD work plan of May 2019 states that member countries need to reach a political agreement on the Inclusive Framework reforms plans in order to allow for completion of the technical work by the end of 2020. Herein lies the major problem. Political agreements on taxation rules are, well, political. This means they have little to do with realities of the modern economies.
The OECD-proposed framework for dealing with digitalisation of the global economy is a clear step toward setting a global minimum tax on corporate activities, a tax that is going to erode both the states’ sovereignty over taxation and fiscal policies, and the process of technological innovation-fostered, competition-based development of the markets. Where the OECD approach will create a complex web of new accounting laws, rules and tools, and a gargantuan whirlpool of tax compliance and reporting regulations, the costs associated with the companies trading across borders using digital services and platforms for delivery of traditional goods and services will undoubtedly become prohibitive for the majority of smaller and medium-sized enterprises. Likewise, entrepreneurship in the areas of innovation and technological development will likely be funnelled into more concentrated, less competition-driven “unicorns”, or larger-scale entrepreneurial ventures. In this, the key risk stemming from the OECD reforms is that of a secular decline in the global economy’s growth potential.