UBER: A Test Case Analysis of Nexus in Digital Economy
Thesis
University of Amsterdam – IBFD
Submitted by
Lakshmi Narayanan
Date of Submission: 27 July 2018
Supervised by
Dr.Joanna C. Wheeler
Abstract
Digitalization of economy creates possibilities for borderless business operations without any physical presence in a jurisdiction. This has created policy challenges in lot of areas including that of international taxation. Especially the principle that profits of MNE should be taxed at the place of value creation is under attack. The direct tax policy challenges associated with digital economy are outlined by OECD under BEPS Action Plan 1.
In the above context, the nexus required for source taxation of cross-border profits of MNE has been found to be weakened. This vulnerability is owing to the nature of source rule codified in tax treaty rules. The tax treaty rules such as PE under article 5 and consequent allocation of taxing rights based on proportionate corporate profits to be taxed at source jurisdiction under article 7 of Model conventions are found nonoperational.
This thesis explores the scope of current tax treaty rules to accommodate digital business model and in case of inadequacy, the alternate options by way of modification of existing tax treaty rules to address the source taxation of such business are examined. To conduct a scientific inquiry, it is found necessary to work within a specific business model. UBER is taken as a test party to undertake such scientific inquiry.
To test the treaty rules for source taxation, the paper discussed the nature of UBER business model by drawing a brief intra firm legal and economic-operational relationship of UBER MNE to create context for discussion. Followed by that, the paper demonstrates that user data and user participation is one of the key value driver for UBER cross-border profits. Subsequently, the inability of tax treaty rules, especially PE under article 5 of the Model convention to address the source taxation of corporate profits derived by UBER is demonstrated. Further, an attempt has been made to restore the international tax system by remodifying source rule under article 5 based on deemed digital PE adopting user as a criterion for tax connection with source jurisdiction.
The key findings are that current treaty rules based on PE rules such as fixed place PE; dependent agent PE or services PE fail to create nexus for digital business in source jurisdiction and in spite of the fact that such profits are sourced from such jurisdiction. Further attempts at recodification of source rule based on deemed digital PE is not adaptable to an allocation of taxing rights through the use of transfer pricing or other principles as laid down under article 7.
It has been concluded that there is a compelling need to look for solutions for source taxation of digital business model outside the scope of PE as laid under article 5 of model convention and reject the idea of transfer pricing laid under article 7.
Further the paper also highlights possible options like withholding taxes under Article 12A of the UN model convention, 2017 to address the source taxation of cross border profits of UBER and the connected challenges. Simultaneously, it also lays emphasis on the need for additional research in this direction by highlighting areas of challenges in such a solution.
List of abbreviations
BEPS Base Erosion and Profit Shifting.
CV Commonditaire Vennootschap (Dutch Limited Partnership) MNE Multi National Enterprise
OECD Organization for Economic Co-operation and Development PE Permanent Establishment
UN United Nations
Table of Contents
1. Introduction ... 7
1.1 UBER as a Test Case for Digital Economy ... 8
1.2 Structure of Thesis ... 9
2. UBER Business Model-Introductory Remarks. ... 11
2.1 UBER- Business Model Analysis ... 11
2.1.1 UBER Business Model-Operational Overview ... 12
2.1.2 Operational Role of UBER ... 12
2.1.3 Salient Operational Features of UBER. ... 13
2.1.4 Interim Conclusions ... 15
2.2.1 UBER Business Model- Legal Overview ... 15
2.2.2 UBER Business Model- Intra Firm Legal Analysis ... 17
2.2.3 Salient Intra-Firm Features of UBER ... 19
2.2.4 Interim Conclusions ... 21
3. The key Value Driver of UBER and its role in UBER ... 22
3.1 A Debate on Value Driver- User Data and Participation Vs IP ... 22
3.2 USER Data vs. IP in UBER ... 22
3.3 Interim Conclusions ... 24
3.4 Role of User Data in UBER ... 25
3.5 Spectrum of User Activity in UBER ... 26
3.6 Concluding Remarks. ... 28
4. Tensions in International Tax Law caused by UBER ... 28
4.1 Source Concept and UBER ... 29
4.2 Source Rule and UBER ... 30
4.3 Fixed PE Rule and UBER ... 31
4.4 Dependent Agent PE and UBER ... 33
4.5 Service PE Rule and UBER ... 34
4.6 Interim Conclusions ... 36
4.7 Connected PE concepts for the purpose of Attribution of Profits ... 36
4.7.1 Separate Independent Entity Concept and UBER ... 36
4.7.2 Separate Independent Entity Rule and UBER ... 37
4.8 Concluding Remarks ... 39
5. Modified Source Rule and its Scope for Allocation of Taxing Rights ... 40
5.1 Reinventing Source Rule ... 41
5.2 Is Mere Creation of Market an Indicator of Income Production Function? ... 41
5.3 Customer in a Traditional Business ... 42
5.4 User in an UBER ... 42
5.5 Source Threshold for UBER ... 45
5.5.1 Website or Server as new PE Threshold for UBER ... 45
5.5.2 Users as new PE Threshold for UBER ... 47
5.6 Interim Conclusions ... 49
5.7 Application of Deemed User based PE for Allocation of Taxing Rights ... 49
5.7.1 Application of Separate Entity Principle to Deemed User based PE ... 50
5.7.2 Application of the Arm’s Length Principle to Deemed User based PE ... 53
5.8 Conclusions ... 54
6. Conclusions ... 55
6.1 Need for Locating Solutions outside the PE Concept ... 56
6.2 Characterization of Payments and Consequent Challenges ... 56
6.3 Conclusion ... 57
1. Introduction
Digitalization has become an omnipresent phenomenon touching all spheres of collective and individual life including the way economy functions. Digitalization of economy creates possibilities for borderless business having wider reach without physical presence [1]. These new businesses utilize digital technologies to render individualized services and products creating new streams of revenue through the innovative business model. Further digital businesses do not rely on tangible assets such as real estate; cars; inventory which are considered to be instrumental in the generation of business profits.1 The economy as a whole, in turn, adopted digital means of doing business which has brought in creative disruption in the entire operational process of global firms. This growing digitalization of economy has created challenges to the international tax system which is highlighted in the OECD/G20 BEPS project, Action 1 report [2].
The Action 1 report classified the direct tax challenges created by digital economy into three broad areas (i) nexus for taxation (ii) use and attribution of value to data and (iii) characterization of payments for digital products or services [3]. The aforementioned issues have drawn the attention of various stakeholders to address the issue of reviewing the international tax rules. Especially there is a strong political will to empower the source jurisdiction to tax such business.
This thesis is an attempt to broaden the debate on nexus for taxation in the digital economy. The business in digital economy are able to segregate their activities from customer markets [4]. The nexus for taxation is found to be an immediate challenge which needs to be resolved before addressing other connected challenges such as the use and attribution of value to data and characterization of payments. All the above cited tax challenges are interconnected and they coexist. However, for the purpose of this thesis the discussion is confined to the challenge of taxable nexus. The issue of nexus has two elements which are linked. The examination of a
right of a jurisdiction to tax and corresponding measurement of profits. These are inseparable elements of nexus which cannot be divorced from each other. Taxable nexus for the purpose of this thesis is thus not confined to the idea of tax connection with a jurisdiction but also examines connected issues of allocation of taxing rights.
Further, this attempt is guided by certain overarching principles propounded in the field of international taxation by OECD and accepted by other nation states under the BEPS inclusive framework. The principle of "value creation" which could be stated as “profits of business should be taxed at the location where the value is generated.” The thesis also explores the viability of principles of Ottawa convention [5] that digital economy should not be ring-fenced. Taxation of the digital economy should be akin to traditional business as principles of equity demand that choice of medium of business should not be a means for creating a distinction between traditional and digital business. But how far such principles could be sustained in case of UBER like business models needs to be tested.
To engage in a research about the nexus in the digital economy it is vital to work within a framework of a particular digital business model to have an informed discussion. This is carried out with the belief that a workable solution could be found for nexus issues in certain business model existing within digital economy. In this endeavor, this thesis focuses on UBER as a test case analysis to explore the issue of nexus in the digital economy.
1.1
UBER as a Test Case for Digital Economy
The primary motive for the choice of UBER as a test case is based on the fact that all major tax challenges related to digital economy highlighted by Action1 report are manifest in UBER business model. UBER business model provides context to extend the dialogue on the nexus issues found in the digital economy. However, it is acknowledged that real business keeps evolving in the process of innovation. But the endeavor here is to use UBER to launch an investigation on the broader issue of the nexus of the digital business model and their connected tax challenges.
The spontaneous choice of UBER is also driven by the fact that such business model does not emanate outside digital economy. The advent of communication technology made UBER like business model feasible. At the outset, it may be recognized that UBER is not a true representative for the entire digital economy but only a sample to address the issue of specific business model i.e sharing economy in the digital economy and not an attempt to address other business models in the digital economy such as e-commerce and cloud computing etc neither it covers digital tools such as robotics; virtual currencies and 3D printing etc. Further, the scope of the thesis is restricted to the discussion on nexus element of sharing economy through UBER as the test case. There might be connected tax issues2 for participants in sharing economy and future tax issues which are beyond the scope of the discussion of this thesis.
Availability of resource material on UBER was another factor of influence in such a choice. UBER case has been examined in the various jurisdiction such as the United Kingdom; the United States and Europe Union on different tax issues. This has provided a good starting point for understanding the business model of UBER.
1.2 Structure of Thesis
To achieve the above purpose of the thesis, the rest of the thesis is structured as follows. Section 2 helps in dissecting the business model of UBER. The focus is to highlight both the intra-firm economic operations and intra-firm legal contracts of UBER MNE (hereafter referred as UBER). This helps us to understand how UBER is legally structured and how the key assets, risks are economically owned and assumed for further tax analysis. In this attempt, brief mention of the consequent tax structure of UBER is also carried out. Tax case laws on UBER; UBER websites; Tech articles on UBER functioning are utilized as literature for building UBER tax framework. This illustration enables to identify the selected feature of the functioning of UBER which are relevant for the analysis of international tax law.
The focus of section 3 is to identify the key value drivers which drive the operations of UBER in the generation of profits. This analysis is carried out to understand the source of UBER business profits. Within key value drivers, an attempt has been made to address the priority between important value drivers through tax case law analysis on UBER. This informed position enables to critically interpret the norms of the international tax system and enhance it in case of gaps.
The ability of current international tax rules to address the issue of nexus of UBER in source jurisdiction is explored in section 4. The areas of stress points in international law; their implications and the need to reform is highlighted here.
Section 5 attempts to modify the international tax rules within accepted international tax law principles to address the issue of the nexus of UBER in source jurisdiction. This is examined as a response to limitations highlighted in section 4. The feasibility of modified international tax rules to act as an effective solution for UBER nexus issue is explored.
Finally, section 6 concludes by indicating the potential areas to look for the answers of nexus in UBER and identify topics for further research.
2. UBER Business Model-Introductory Remarks.
One of the essential requirement of the tax policy is to frame laws closely corresponding to the economic reality of business. The purpose of such tax laws is to identify a tax subject and object; define the extent of engagement of tax subject with a jurisdiction; the nature of tax object and the quantum of tax object. It is of paramount importance to understand the economic reality of a business in order to legislate international tax laws which will avoid double taxation and double non-taxation. Thus, the purpose of this section is to understand UBER business model. It is an attempt to delineate the salient features of the UBER business model around which informed tax policy discussion could be carried out.
2.1 UBER- Business Model Analysis
The UBER business model is embedded in two layers. One layer is the legal domain pertaining to intra firm contracts between entities of the UBER group. The other layer is the operational domain relating to core mechanical operations; prices; revenue and costs etc. The operational and legal domain represents two sides of the same coin i.e the business model. Examination of both layers of the UBER gives us a complete picture of the business model.
Here we will proceed initially by an operational analysis followed by a legal analysis to map the UBER business model.
2.1.1 UBER Business Model-Operational Overview
There is a number of sharing economy business model in the digital medium in the field of short-term housing; professional services; luxury clothing; finance etc [6]. UBER is representative of one such sharing economy business model in the digital economy. UBER business model operates through an application on the mobile device which connects service consuming users3 and service rendering users4 for the purpose of rendering specified services.5 These two set of users mostly are private individuals. The role of UBER is to regulate the interactions such as booking; payments; rating and reviewing between the two set of users through an automated platform like UBER app based on real-time information. In the process of facilitating such interactions, UBER earns a commission on payments made by the service consuming users to service rendering users at a margin of 20 to 30% [7] gross payments. This forms the primary corporate gross revenue of UBER and there shall be additional revenue streams through placement of advertisements in service rendering user’s navigation panel targeting both set of users. The taxation of corporate profit earned after deducting expenses for the cost of platform development spent on applications; algorithms; logos etc; sales and marketing; salaries of technical and managerial personnel from the gross revenue of the UBER Group is the core subject of debate especially the source taxation of such income streams.
2.1.2 Operational Role of UBER
The core operation of UBER business model is about monetizing interactions between two sets of users on the platform. This is carried out through a series of process as detailed out:
3 The term “service consuming users” is hereafter used to refer to the group of users who seek services like cab rides; delivery of food and groceries and delivery of couriers in UBER platform. It may be reminded that usage of term “service consuming users” is not restricted to rendering cab services. 4 The term “service rendering user” is hereafter used to refer to the group of users who render services like cab rides; delivery of food and groceries and delivery of couriers through UBER platform. It may be reminded that usage of term “service rendering user” is not restricted to rendering cab services alone. 5 UBER X; UBER Eats; UBER Fresh; UBER Rush are services in the field of transportation of people; food; groceries and courier services (hereafter referred as “specified services”) respectively. UBER online business operations are not merely in the domain of ride-hailing as often thought of but it is the future of the online logistics industry.
Firstly, UBER platform sets standards for service rendering users to engage in the platform by defining their requirements in terms of vehicle standards; driving skill set and past records etc. Later monitor their active hours and location details real time in the platform to render specified services to service consuming users. Secondly, the service consuming users through UBER applications on their mobile device key in their personal data followed up by request for specified services on the platform. Thirdly, UBER app matches service consuming users request for specified services with service rendering users present on the platform to create value for each other [8]. Fourthly, on completion of the service delivery, the fee is directly charged by the platform itself to the service consuming user’s bank card or credit card entered in UBER applications. The fee is calculated dynamically by the platform based on certain algorithms. Fifthly, the applications also provide for rating and review system wherein the users are allowed to provide feedback about each other and about the interaction of users with the platform [9]. This feedback acts as a mechanism for allocating further specified service requests between two set of users.
Thus, UBER collects user data through the applications and create real-time synergies based on interactions between user data and algorithms which outperforms conventional business.
2.1.3 Salient Operational Features of UBER.
UBER unlocks new sources of value and does not shift the existing source of value in the market on to its platform. The service rendering users such as taxi drivers are not employees of the UBER. These are new short-term labors who have joined the platform owing to the freedom and privileges offered by the platform. Further, UBER is in control of specified services which helps UBER to define such value created in the platform on its own and is not defined by end users, thus it is not the mere intermediary.6
6 Advocate General Szpunar Opinion on UBER case in ECJ case law C-484/15. The degree of control exerted by UBER in service delivery highlights that UBER is not the mere intermediary. This point is
The traditional business operates largely on four factors of production. Uber business model operates on the fifth factor of production i.e the user data and activity in the platform. The user data provided by users on both sides of the platform and the user activity is the engine of the platform. It is important to recognize data as any other input factor such as capital; labor or technology for business [10]. The role of user data and user activity would be discussed in detail in section 3.4 and 3.5.
The other key feature of UBER is the heavy reliance on intellectual property7. UBER platform is primarily driven by its technology compared to the traditional yellow cabs model of the taxi business. Almost all business operations of Uber are driven by technology right from demand identification; standardization of supply; matching of demand and supply; design of navigation route; pricing and settlement of service fees and ranking of supply performance etc. Uber currently owns patents covering a large part of its operation such as business model; app icons; design; pricing algorithm and navigation route etc. This heavy reliance on IP as opposed to tangible property, provides an opportunity for digital business to operate in a jurisdiction without any physical presence. This has important tax implications which will be discussed in section 4.
UBER business is a multisided market as it offers a range of specified services to a different set of service consuming users as detailed out in section 2.1.1. Multisided market enables users to network across jurisdiction and allow them to interact with each other through web applications [11] as in UBER. Typically, the multi-sided market has two core characteristics namely indirect network externality and non-neutral pricing [12]. Both these elements are present in the UBER. Firstly, an increase in the number of users on demand side (service consuming users) create more utility for users on the supply side (service rendering users) in the UBER platform; this demonstrates the indirect network externality. Secondly, pricing on a platform is also non-neutral as opposed to neutral. This allows pricing the specified services offered beyond the marginal utility of each new user added on to the platform. It allows charging less on one side of the market to add more service consuming users on the platform and overcharge on another side the service rendering users until a critical user mass is reached.
Thus, the price is not neutral with respect to distance traveled but its dependent upon the relative number of end users present on both sides of the platform.
A connected feature of the indirect network effect in UBER business model is the ability to scale up operations rapidly and operate a global value chain. A conventional business, which operates globally would require huge scale of investments in assets, people and technology. However, UBER does not need to buy a huge fleet of cars or to hire a large number of labors on a permanent basis to expand its business. Similar effect is achieved by adding more number of users to the platform. Further, the cost incurred on sales and marketing comes down due to the network effect of users pulling each other. This creates economies of scale for UBER business model which allows structuring pricing differently.
2.1.4 Interim Conclusions
The examination of operational business model shows that UBER creates a platform where users create and consume value. The core unit of such value is defined by UBER and not by end users. The cost of such a business model is in terms of platform development. The immediate revenue stream is the commission on gross service fee paid for consumption of value created and additional revenue stream based on targeted advertisements oriented towards users on the platform could also be found.
Further, the salient feature of UBER business model is that it is based on collection of real time user data which is the root of all operations. It creates new sources of value by bringing new parties to market. It has a high reliance on IP assets. It operates a global value chain. UBER business model demonstrates multisided market dimension with indirect network effect and non-neutral pricing model.
Legal analysis of UBER business model proceeds in two steps. First in identifying the contracting parties and the terms between them in the business. Second, what are the functions; assets and risks assumed by identified parties for business operations.
UBER
Service rendering users Service consuming users
From the above illustration, it is evident that there are three contracting parties in the UBER business and the legal relationship between these parties’ manifest in three axes. Firstly, the relationship between service rendering users and UBER is already a subject of intense debate, which has important ramifications for tax classification of service rendering users as independent contractors or salaried employees of UBER. This aspect of legal contracts between service rendering users and UBER was examined in great detail by the London employment tribunal case [13] and is beyond the scope of the discussion of this thesis. Further, the legal relationship between service consuming users and UBER are built into legal contracts in the digital world called as smart digital contracts. These smart contracts are concluded by the service consuming users expressing their “legal consent” for usage of their locational and other personal data and accept “rider terms”; “booking terms”; “payment terms” designed by UBER for specified services offered through the platform. Lastly, operational interactions between two set of users are very minimal and interaction between these two users in terms of matching; pricing; payment settlement and rating are controlled to a greater degree by the platform itself. The validity of legal contracts between both the users of UBER platform are subject to debate8. It is to be mentioned here that the functional relationship of both sets of users in furnishing user
8 There is a legal contract between both set of users on the platform. However, such contracts treated as pure fiction of law as the economic control of operations of such contracts are fairly regulated by UBER platform. This idea is elaborated in section 3.2.
data and performing user activity towards UBER group as a whole is elaborated in detail in section 3.2 and 3.3.
2.2.2 UBER Business Model- Intra Firm Legal Analysis
For users, UBER may be a single entity i.e. the brand UBER. However, UBER is not a single entity. The focus of this section is to analyze the intra firm legal relationship between different entities of UBER group. This analysis will be able to identify different entities and their functional obligations towards the UBER group.
Before addressing the above aspect, a brief analysis of the business architecture of UBER shall be made. The UBER business architecture constitutes a platform i.e. UBER web app; data storage unit; algorithms to process stored data; research and development activity to develop platform; marketing intangible like app and logo design; marketing and sales division at the global and regional level to promote the UBER network. This has been neatly segregated into value network analysis breaking such components as network infrastructure management; network promotion and maintenance and service provisioning by OECD [14].
The different components of the business architecture are distributed between these global corporate entities of UBER group. A graphic narration of the same is provided in Figure 1.9 It is relevant to mention that this graphic narration provided is a reasonable approximation capturing the essence of legal tax structure rather than a complete picture of all group entities involved.
9 This is perused from two sources: page 69 of OECD Full interim report on digital economy Annex Figure 2.A.3 and How uber plays the tax shell game? http://fortune.com/2015/10/22/uber-tax-shell/. This figure is only a reasonable approximation as complete UBER tax picture is not available on the public domain.
Figure 1 Intra firm contractual relationship of Uber group
UBER technologies Inc. is the parent company located at United States. This parent entity initially developed the network infrastructure such as platform, web applications, algorithms and also developed the marketing intangibles such as app; logos and operational intangibles such as business model patents. It is also responsible for global marketing and human resource management of technical personnel. This, in turn, holds two key subsidiaries10 in Netherlands namely UBER International CV incorporated in the Netherlands and headquartered at Bermuda and UBER BV incorporated and headquartered at the Netherlands. This CV- BV structure operates as the holding company for all of Europe and a few other countries like UK etc. UBER international CV holds all the IP assets as UBER tech Inc. US has sold the initial IP developed
10 Apart from other large numbers of subsidiaries it has in the Netherlands, it is reported it holds around 10 subsidiaries in the Netherlands, nine of which are located in the same building, Amsterdam central canal district.
to UBER international CV. Further, it also has a CSA with UBER tech Inc. to share costs and benefits of future IP to be developed. UBER BV has got the license to use all IP assets owned by UBER international CV to use outside the US. This is also responsible for local platform configuration; partial software development functions. Further, all payments made in the completion of specified services made by service consuming users are routed through UBER BV which in turn forwards to Raiser operations BV which settles payments to service rendering users. Further UBER BV being the holding company it supports all the foreign operating company in terms of funding the foreign subsidiaries operations. There is a string of subsidiaries throughout Europe such as Uber systems Spain SL; UBER Italy which render regional support services like marketing and sales etc. in the respective jurisdiction by enrolling drivers on to the platform initially.
The vast amount of data gathered globally from users are stored, retrieved, and processed. In this regard, UBER owns all the data it has gathered and controls the data in UBER Tech Inc US for US region and UBER BV controls the data gathered outside the US. Further, the group as a whole may engage external third-party server providers [7] to provide web hosting and cloud computing services which help in hosting UBER app and storing user data for further processing by UBER algorithms.
Further, it is important to note that in the above tax structure both set of users are outside the UBER group structure. Both type of users is not employees or consultant of UBER group on the contrary, they are unrelated third party performing certain activities for the group.
The above analysis helps us in identifying certain key features of intra firm structure of UBER. Only a brief mention of the same is made here for further discussion in the subsequent sections11.
There is multiple IP owner jointly responsible for the development and utilization of IP assets thus economic risks are shared between the entities. Centralization of assets and risks in few entities of the group in a jurisdiction is evident from the intra firm legal analysis [15]. There is a high degree of operational interdependence and integration of management and entrepreneurial risks. This leads to an integrated UBER group operating a global value chain. There is no linear process of production of value instead there is a nonlinear value chain which generates value out of a synergetic process of interaction between user data and entities owning the IP algorithm processing such collected data. Further, it is evident that taxable presence is avoided in source jurisdiction not as a matter of tax planning but the business model creates possibility to operate without a taxable presence in source jurisdiction.
As regards, the tax efficiency [6] of UBER structure detailed in Figure 1, it can be stated that UBER CV entity acts as a reverse hybrid structure treated as a controlled foreign corporation in the United States earning non-subpart F income that was allowed to defer U.S. tax until repatriation.12 The CV’s headquarters is in Bermuda, where there is no corporate income tax. The Netherlands view the CV as a company controlled by U.S. owners having Bermuda headquarters and no business operations or taxable income in the Netherlands. Moreover, BV remuneration for the local support subsidiaries in other countries may be reduced by interest payments on intercompany start-up loans. This aggressive tax structure having elements of base erosion and profit shifting is created to avoid residence taxation which is out of the scope of the subject matter of this thesis.
11 These ideas are expanded in section 3 and section 4 which identifies the key value driver of UBER and gaps in international tax rules in addressing such value driver.
12 With the introduction of the Tax Cuts and Jobs Act (P.L. 115-97) – with its repatriation, global intangible low-taxed income (GILTI), and dividend income provisions, U.S. taxation of the CV's royalty income may be different.
2.2.4 Interim Conclusions
The primary focus of legal analysis is on UBER intra firm contractual terms. UBER contractual analysis indicates centralization of key assets and risks. Further it could be identified that key assets of UBER is owned by multiple entities. The contractual analysis indicates high degree of operational interdependence. However, operations are carried based on user data provided by users outside the group as unrelated third party. UBER user data, in turn, is owned and controlled by UBER Tech Inc. US and UBER BV.
3. The Key Value Driver of UBER and its role in UBER
Certain key elements of the business model were identified in previous sections. The purpose of this section is to identify the key value driver among the identified elements of UBER business model. The emphasis is to identify the key value driver that allows UBER to be different from a traditional business and highlight the role played by such a key value driver for UBER.
3.1 A Debate on Value Driver- User Data and Participation Vs IP
UBER was found to be a platform operating based on a key factor of user data at a global scale with high reliance on IP. User data is created by users operating as free labor outside the domain of group entities. User data and IP are two important elements of UBER business.
Currently, there is continued debate on the contribution of user data and participation in the digital business model like UBER. Broadly there could be three groups of countries which can be noted in the BEPS inclusive framework discussion on the digital economy based on their appraisal of the role of user data and participation in business. The first group identifies it to be value drivers of business; the second group rejects the idea that user data and participation could create value and third group state that user data and participation should not be segregated from other business inputs derived from customers [16]. Similarly, scholars tend to disagree on the role of user data and participation vs IP in digital business. One segment of scholars argue that raw data does not create value in itself and it is the intellectual element of human capital that creates value [17]. On the other side scholars also argue that raw data in itself could be treated as know how or proprietary information which can enhance a commercial activity and create value [18]. This lays the emphasis to analyze the role of user data and participation vs IP in UBER business.
The user data on both sides of the platform is the key input for delivery of specified service in UBER business model [19]. It has a reflection in contractual terms in the agreement between UBER and two sets of users on the platform wherein users provide data in exchange for usage of UBER app services.
IP and data are two important assets which segregate an UBER from rest of the traditional business. The traditional Yellow cab taxi does not own large patents or trademarks to conduct its business neither does it collect and own a large pool of service consuming user data [15]. A traditional yellow cab taxi does not develop applications and algorithms to run its business. Further, a traditional yellow cab taxi shall not have real-time location details of customers and drivers; customers bank card/credit card details; drivers bank account information. The least data it stores is that of a group of drivers details at a rudimentary level. Neither does a traditional yellow cab regulate the business as a whole. It does not set prices for the ride dynamically; do not set vehicle standards; do not create a feedback loop in terms of driver ratings and customer review of drivers. However, UBER perform all the above functions. This is feasible because of data and IP assets of UBER.
It is critical to understand that the role of IP and data are complementary. The IP algorithms enables communication between users based on the user data. The challenge, is in fixing the order of priority for IP and data in terms of their importance. To find whether data gathered is as important as the processed data through data analytics provided by the IP. we need further analysis to arrive at a conclusion.
Any emphasis on IP focuses on largely the supply side of the UBER business model. IP algorithms could be conceived as the mere instrumentality in achieving the goal of being a specified service provider. It is the communication tool through which business is conducted. It could be aptly summarized that "Uber does not simply sell software; it sells rides. Uber is no more a "technology company" than Yellow Cab is a "technology company” It was adjudicated that UBER is not a technology company but it is a transportation company [20]. The IP algorithms were only instrumental in bringing the two set of user closer as a citizen band radio used for communication between cab drivers and cab facilitator would have done
in the yellow cab. Thus, technology is not the only dividing line between traditional business like yellow cab and UBER, however, it is a significant factor.
Further, an analysis of UBER business model with similar platform business models like hotel and airline booking platform brings the idea that technology or algorithms are the significant factors but plays a marginal role compared to data. It is important to note certain things. Hotels and Airline's booking platform and UBER are similar in terms of mechanical operations that they are rendered through dematerialized platform through the use of IP tools. However, hotel and airline booking platform cannot control the conditions of airline or hotel industry standards or performance of service or a fix price for such services rendered whereas on the contrary UBER can alter and control the conditions of specified services it renders through its platform and can fix price dynamically for such specified services rendered. Moreover, the service consuming users tend to associate delivery of such services with the brand UBER and not with service rendering users. Thus, UBER platform creates; define and allow the exchange of value on its own compared to hotel and airline booking platform which merely acts as an intermediary [21].
This is feasible because of the real-time data such as location details of both users/ preferred destination of service consuming user and the fungible nature of such data being capable of change over time etc. any focus on IP algorithms will wrongly equate all platforms into one category i.e. platform that acts as intermediaries and platform that actually create and control market. For both, these rely on IP to deliver services however the distinguishing feature is in terms of the role user data plays in these two different platforms.
3.3 Interim Conclusions
IP tools act as a significant mechanical tool but not the defining feature of UBER. However, UBER role in controlling and altering the conditions of specified service delivery is owing to the real-time availability of data and fungible nature of such data. From the above analysis, it could be stated in the affirmative user data is a key value driver of UBER.
3.4 Role of User Data in UBER
User data is the key value driver in the business model. However, a granular analysis of data needs to be undertaken. This granular analysis needs to discuss the type of data gathered and the utility value of data in a platform to understand how UBER monetize such data. Alongside it is important to access the role of parties in the creation of data critical for the functioning of UBER.
External users on two sides of the platform feed real-time data onto the platform which is fundamental for UBER operations. The platform collects data such as real-time location details; address; mobile number; Bank/credit card details; the current preferred destination from the service consuming users and stores their previous travel details etc. Similarly, it collects real-time location details; address; mobile number; bank account details; past conduct records; active hours spent on platform etc. from the service rendering users. Further user's data include reviews and rating of users on another side which are important in creating a feedback loop in the Platform.
The user data in UBER has the high degree of specificity and real-time value. Every data entered on the platform is by the user of the platform seeking a potential opportunity to consume or render a specified service. Thus, every user data can be readily monetized unlike user data on other platforms such as Facebook, LinkedIn etc. Secondly, data is highly fungible in the sense that it constantly changes demand and supply on a real-time basis as opposed to airlines or hotel booking platform where supply data is a static feature in terms of location; time and performance. This fungible nature of data lends itself as a raw material which is capable of exploitation by the platform to derive value.
As regards the utility of user data, it is the key input in all the algorithms. The IP tools in the business layer of UBER use this data to match both users [22]. In traditional business, there is always an element of idle time for production of value. UBER idle time of service delivery is very negligible. Service consuming users looking for specified services should get the services within seconds any longer idle time will lead to failure of the platform. This reduction of idle
time in the production of value is possible only due to the synergy derived out of interaction between IP and real-time availability of user data from both sets of users. This is carried out through the set of algorithms which matches users in the platform in real time. Secondly, the user data also serves as the input for the surge pricing algorithm nicknamed "geo surge" [22]. UBER sets the price according to the supply of service rendering users and demand from service consuming users for specified services at any given time. Thirdly, the user data in terms of driver reviews and driver rating by service consuming users help in the platform deciding whom to eliminate from the platform through IP algorithms. Further the rating of service rendering users on the nature of the transaction; number of rides performed by service consuming users and payment settlement by them help in future business functioning. This helps in building trust in the brand UBER and enhances the brand value of UBER. Fourthly, more activity is created on the platform by reviews done by users and this in turn helps in marketing the brand by creating more incentives for users outside the platform to move on to the platform. This helps in achieving scale without further marketing investments in the brand. From the above discussion, it is evident that user data is the building block of UBER IP; UBER brand; a tool for achieving scale and bringing a fresh stream of revenue.
The crucial aspect of UBER business model is that data collection from users happens in a dematerialized manner through applications available in mobile, hosted on the internet through servers. Thus, the UBER business model does not need to create a physical branch or office or factory to collect the user data in the jurisdiction in which UBER renders specified services.
3.5 Spectrum of User Activity in UBER
The activities carried out by users in UBER varies with the degree of user participation. The intensity of user participation can be distinguished based on the series of activities performed by the user. User activity chart is detailed below in Figure 2.
Figure 2 User activity chart in Uber
The above figure illustrates the range of such activity. Firstly, user login account in UBER app creates the ability of the user to use web app and feed his data. This stage of activity is the primary level of activity which is about user data collection by UBER from the user. The user site profile also helps in verification of user records such as his previous records of conduct especially in case of service rendering users and their level of expertise. The user activity of building connections by mutual reference builds a network effect of pulling further users onto the platform. This is crucial for building the critical mass of demand and supply for UBER. The user activity of payments helps in establishing the credibility of service consuming users and their ability to settle payments on time without any disputes. Further, it also helps in UBER offering structured discounts for frequent travelers. The user preference and previous rides help in building customized services for the user in terms of vehicle choice; service consuming user preference and time requirements of the user. The user activity of feeding geolocational data is highly valuable for generating different pricing structure for UBER specified services in real time through UBER IP. Further user activity of rating and review the performance of service delivered by the user is an important feedback mechanism which strengthens the trust of UBER. The user activity of comments refers to comments on the brand UBER itself as it is an important tool in furthering the brand of UBER.
All the above activities performed by USER could be treated as functions performed towards the generation of business profits by any other group entities of UBER. This notion of user activity as significant people functions would be elaborated in section 5.4 in the process of finding tax solutions for UBER.
However, it might be stressed that UBER business model allows users, the third party to not only provide data for operationalizing UBER but also perform critical functions necessary for the conduct of the business. In the process, they perform the functions of the UBER as free labor in return for providing or consuming specified services. This creates far-reaching tax implications for UBER business model as UBER is able to derive business profits without having the need to carry people functions from the jurisdiction from which business profits are derived. The users’ data and activity are significant value drivers in generation of corporate profits for UBER. However, such value drivers are not valued economically in spite of the fact UBER derives economic value from the user data.
3.6 Concluding Remarks.
This section demonstrates that user data is a significant value driver of UBER. Further the user data and user activity in the UBER platform obviates the need for having a physical presence to conduct business from within a jurisdiction. This causes stress points in current international tax rules.
4. Tensions in International Tax Law caused by UBER
The previous section highlighted about the important stress points in international tax rules owing to the nature of UBER business model. This stress emanates out of remote operations of UBER based on user data and user activity in a jurisdiction. The purpose of this section is to analyze the ability of current international tax law rules to capture the value created in source jurisdiction by UBER. It is important to note that tax law principle exists both in domestic and
international tax laws. The role of this chapter is to focus on international tax law concepts and rule13 such as source and single entity etc and its adaptability to UBER business model.
In this direction, initially, an attempt shall be made to carry out discussion on international law concepts and consequent codification of rules in law. Simultaneously, an analysis will be carried out to find the applicability of such tax principles and rules to UBER.
4.1 Source Concept and UBER
The source at the broader legal conceptual level can be defined as the place of income generating activity [23]. However, there is no singular narrative for the concept of source. In this thesis, the focus is on UBER business profits income. At an international level, there is a moderate level of consensus on the principle of the source. This consensual principle of the source is articulated through the benefit theory and the doctrine of economic allegiance.
The benefit theory is based on the idea that a nonresident deriving benefit from the source country especially use of country infrastructure; the legal system as provided to residents of source country should be treated as similar to residents [24].
In case of UBER, it benefits from user jurisdiction through various ways such as enabling and maintenance of digital infrastructure in source country; protection of UBER IP rights such as UBER app in market country; enforcement of service consuming users defaults in payments through legal system; general provision of energy supply to run the digital environment; providing legal rules for operationalizing such UBER platform in source country [25]. Further UBER utilizes public goods offered by source country such as the infrastructure for regulating traffic; issuance of driver license to the certain service rendering users; a reasonable amount of road infrastructure and enforcement of legal rules for violation of contracts by users against
13 It is not to state that domestic tax law concepts are not necessary rather the international tax law
principles are built on domestic tax law concepts. However, the scope of this paper is in the domain of international tax law domain. Indeed, domestic tax law needs to be addressed before international tax law debate happens. It is not an attempt to brush aside those concerns. Further, it is natural that UBER
UBER. All these public goods availed by UBER in source country is built by the contribution of resident taxpayers in the source country. This demands that UBER, the nonresidents who also avail of such services owes to contribute to the source country in terms of the principle of benefit theory. Thus, based on the benefit theory the source jurisdiction is entitled to tax a portion of UBER income.
The doctrine of economic allegiance postulates that relative contributions made by different nations states in the production and enjoyment of income needs to be examined to create basis for taxation. One of the factors of economic allegiance laying foundation for source taxation is the origin of wealth.14 This origin of wealth is the place where income is produced owing to the existence of economic life which is able to create the possibility of generation of wealth [26].
Discussion in previous sections 3.2; 3.4 and 3.5 establishes the user data and activity are factors of income production for UBER. This user base of UBER especially the service rendering users of the UBER are integral to the source country. This highlights the fact that UBER factors of production are integrated with the economic life of source country. Thus, the origin of wealth i.e the cross-border business profits of UBER is also derived due to monetization of user data and activity in source jurisdiction.
On a conceptual level source jurisdiction, right to tax an entity is satisfied in the case of UBER based on benefit theory and economic allegiance as demonstrated above. However, as per international source rules, the source jurisdiction is not able to exercise the taxing rights in case of UBER. This brings us to the point of examining the international source rule.
4.2 Source Rule and UBER
14 Origin of wealth is considered to be the primary basis of source taxation. This is evident from the fact that all committees and convention under league of conventions based source taxation based on principle of origin of wealth.
Firstly, source, as a concept differs with the character of income15. At the level of Models in terms of business profits source concept is expressed through the rule of Permanent Establishments (PE). The PE rule in international tax order allows source jurisdiction to exercise its taxing rights. This PE rule performs multiple roles in a tax jurisdiction. Firstly, the primary role of PE is acting as the nexus rule which establishes a tax connection with source jurisdiction; Secondly, PE as a threshold rule enables to address the compliance and enforcement problem by ensuring that only substantial business activity and continuing presence in a source jurisdiction is taxed [27]. This ensures tax certainty for nonresidents in source jurisdiction. Thirdly it also acts as income characterization rule in case of business activity associated with PE for reclassifying certain streams of income such as interest; royalty and dividends to be business income through the application of Model treaty term "effectively connected".
In case of UBER, all three above aspects of PE rule are subject to debate. However, the focus of this thesis is confined to analysis of nexus. The nexus and threshold element of source rule are subject to review. One of the roles of PE is to function as nexus factor which locates situs of business profits generating activity. This nexus is conceived to act as the link with source jurisdiction based on which taxing rights are exercised. Absence of the PE nexus creates a situation where only resident state is allowed to exercise the taxing rights in a treaty scenario. Thus, PE nexus acts as a balancing act for the exercise of taxing rights between source and residence taxation.
4.3 Fixed PE Rule and UBER
In case of business income, the PE concept is articulated primarily as a fixed place of business through which the business profits are sourced in a jurisdiction.16 Such a fixed place is endowed to be possessed with attributes such as a physical place; at disposal of the entity; fixed in terms of geographic and time (to a certain extent); that place acts as commercial and geographical
15 Source operates differently for interest; royalty; dividends or business profits.
16 The discussion at this juncture only place emphasis on the OECD 2014 Model and commentary in para 42.11 which does not differentiate business income and service income and do not focus on service
coherent unity; through which business is carried on partly or wholly. A non-exhaustive positive list of places to be conceived as PE are provided in the model which includes office; factory; branch; workshop; place of management; mine; oil or gas well etc [28].
From the analysis of UBER in section 2 and 3, it is clear that the business operations of UBER are automated through technology and this enables UBER to operate in a jurisdiction without places such as office; factory; or workshop. Thus, UBER is able to conduct business or trade with a jurisdiction but not from within the jurisdiction. It is also rapidly able to scale up operations in a jurisdiction by addition of users of the platform without the material need for being present in a jurisdiction. The only remote conceivable tangible presence in operational terms could be about the use of servers.
Further sections 3.2; 3.4 and 3.5, clearly indicates that user data and user activity in UBER platform are key value drivers through which business profits are generated. The substantial business activity of UBER is carried by the activity of the user through the virtual presence of a user in UBER and not through the physical presence of UBER itself. Further user activity and user data is the basis of value creation in source jurisdiction through brand strengthening; input for IP and marketing support through buildup of sizeable user base as discussed in section 3.4 and 3.5. From the discussion in section 3.4 and 3.5, it is clear that there is a value creation activity in a source jurisdiction by UBER without physical presence. Thus, it was found that UBER is able to generate business profits remotely without the need for physical presence in such jurisdiction.
But the current PE rules do not meet the threshold requirements for a source to exercise taxing rights over business profits of UBER. As a consequence, the user activity created value is not captured by current rules and consequently the business profits derived by UBER flows to the jurisdiction of corporate residence of UBER and other group entities in the intermediate jurisdiction. In this scenario, the source tax rights are not exercised and only resident taxation of UBER is exercised. Therefore, the function of the PE rule to establish a tax connection for business profit made in source jurisdiction is nonfunctional.
This failure of nexus element of PE fiction is owing to the current physical and material nature of PE. In a digital business model like UBER physical PE rule is outdated and assumptions surrounding such PE notion needs to be reviewed. Consequently, the scope of the legal rule of PE is in serious question which needs substantial deliberation.
4.4 Dependent Agent PE and UBER
In this regard, the other alternative to locating a PE in source jurisdiction is through the concept of dependent agent PE. A dependent agent with an authority to conclude contracts in its own name17 or has the authority to play a principal role in contract conclusion without material modification by head office [29]. In order to conceive that we need to find the agent of UBER who concludes contracts in the name of the principal or substantially negotiates the predominant material aspects of contracts without the intervention of principal in the source jurisdiction with the users of the platform.
The first difficulty in this regard is about the debate between which parties the contract is concluded in UBER for generation of business profits. The conclusion of contracts as claimed by UBER [30] for the performance of specified services is between the service rendering user and service consuming user. In UBER perception conclusion of contracts with service consuming and service rendering users with UBER are only for storage and monitoring of data. Further for the performance of specified service delivery such as rides the contracts are supposed to be concluded between users of the platform, not with the UBER. However, such an argument was clearly demolished by the court, the idea that legal contract between two set of users was declared to be pure fiction and which bears no relation to real dealings and relationship between parties [31]. Here primarily legal contract needs to be characterized in material economic terms to ascertain the contract is concluded with the UBER and not between two set of users in the platform for delivery of specified services.
Subsequently in conclusion of material contracts with users through which the business profits of UBER are generated whether a dependent agent PE for UBER in source country needs to be examined. Whether service rendering users like driver could be treated as dependent agent for the enterprise, the UBER? It is clear from the business operations that the service rendering users do not have authority to conclude contracts either in the principal UBER name or their own name. The authority to conclude contracts is bestowed with the UBER web applications. Further contracts oriented towards income producing activity are concluded through UBER web applications digitally without actual intervention by any agent in source jurisdiction. The agent cannot modify the terms of the contract or negotiate contract with service consuming users and the contract is automatically generated by the UBER web applications without the agent playing the principal terms. All the material elements of contracts such as price; control of material information for specified service delivery is controlled by UBER applications and not by service rendering users like drivers. However, there may be an element of argument over here that service rendering users like drivers have the right to reject the contract offered by UBER web applications. Does this right to reject the contract makes the driver a dependent agent by playing principal role in conclusion of contracts. This is far from clear that the right to reject the UBER app generated contract are completely at the control of service rendering users like drivers. On the contrary, UBER enforces acceptance of contracts indirectly through penalizing such users who reject the contract by logging off drivers who breach those requirements [32]. Further the rejection of contract by drivers indicate that there is no business relationship between UBER and service rendering users. In the absence of business relationship, it is not feasible to conceive a dependent agent. Thus, there is no clear sign that service rendering users have any principal role in material conclusion of contracts on behalf of the enterprise in this case. Therefore, there cannot be dependent agent PE for UBER.
From the above, it is evident that UBER need not have a dependent agent also in source jurisdiction to conclude contracts on its behalf through which UBER derives business profits.
4.5 Service PE Rule and UBER
A service PE is conceived in UN Model [33] based on the idea of provision of services by an enterprise in a jurisdiction through employees or other personnel engaged by the enterprise for
a period of more than 183 days in any 12 months commencing or ending in a fiscal year concerned. In case of UBER, service rendering users like drivers are not employees as per legal contractual terms between UBER and drivers. There could be a possibility to treat the service rendering users like drivers within the undefined treaty term “other personnel engaged by the enterprise”. But the possibility is not without an element of debate 18. The UN commentary does not provide adequate clarity on the above term to provide any meaningful conclusion. Further, what is the service rendered by the UBER? Is it a taxi services19 that is rendered by UBER or mere intermediary services of providing quick rapid matching of both set of users? This needs to be answered before constitution of service PE. As treating UBER as mere intermediary will lead to non-constitution of service PE in source jurisdiction as matching intermediary services are not carried out in source jurisdiction but through dematerialized web applications. However, as acknowledged by the courts decision if UBER considered to render taxi services than there might be a possibility to create service PE still with reservations whether service rendering users like drivers constitute “other personnel engaged by the enterprise”.
OECD commentary [34] also acknowledges with certain sense of reservations a formulation of service PE based on similar legal construct as discussed above in UN model under limb (b)20. However, the OECD choose the term individual and not employees or personnel as cited under UN model. Thus, service rendering users like drivers squarely fall within the scope of the provision. However, the above debate on what services UBER rendered in source jurisdiction is still a relevant question and a subject matter of contention.
From the above, it is evident that whether a service PE could be constituted for UBER in source jurisdiction is a subject of debate and not an objective fact of conclusion.
18 This needs deliberation on treaty interpretation under Article 3(2) found in Models as any undefined term whether needs to be given domestic tax law or contextual or ordinary or special meaning. This is beyond the scope of this paper.
19 As propounded in European court of Justice Case C-434/15, association professional elite taxi and in the case of Douglas O’Connor –Vs- Uber Technologies Inc19 before the North California District Court, United States.
4.6 Interim Conclusions
The above discussion has highlighted the user jurisdiction at a source concept has still the right to exercise the taxing rights of source jurisdiction based on principle of benefit and economic allegiance theory. Further it is evident that source rule based on fixed or dependent agent PE directly fail to capture the value created in source jurisdiction. There is a scope to consider existence of service PE for UBER in source but it is not a matter of fact but it is based on opinion. Thus, it is reasonable to conclude the current PE rules are not completely adequate in establishing the tax connection with a jurisdiction. Hence there is a need to reinterpret the PE rules otherwise the value created in source jurisdiction would not be taxable at source.
4.7 Connected PE concepts for the purpose of Attribution of Profits
The PE definition as codified under Article 5 of model convention is relevant to determine whether business profits of enterprise may be taxed in other state. The source rule codified under article 5 only provides a definitional threshold for exercising source taxing rights. However, it cannot allocate taxing rights on its own. Article 7 of Model convention needs to be symbiotically interpreted with article 5 to determine what profits could be attributed to PE established under article 5 based on certain criteria. Thus, in examining the nexus issue of UBER it is important to discuss the principles and rules based on which article 7 is laid down. These principles are “separate entity” and “arm’s length”.[35]
4.7.1 Separate Independent Entity Concept and UBER
The legal fiction of functionally independent entity in the international tax domain requires to segregate group entities under common control and ownership as a stand-alone entity. This is carried in order to divide the income tax base between these operational group entities. This operates at two level one at branch level other at the subsidiary level. This concept at subsidiary level is to an extent easier to formulate and enforce owing to the existence of legal contracts between subsidiary and head office, a challenging task as far as branches are concerned.
However, they exist to address the goal of income tax base division in an international tax order.
In conceptual terms, it is easier to create actual or fictional independent entities when we have clearly delineated operations within group entities who manage certain activities and assume the risk for those activities deviating from expectations. To be precise de-consolidated entities within a multinational firm separately managing and performing key functions of the firm would be consistent with the Article 7 and Article 9 foundational assumptions.
It is difficult to conceptualize UBER as neatly de-consolidated national entities within the group managed separately for specific assigned functions. UBER functions are highly integrated and interdependent which is carried out through constant interaction in the digital medium between various group entities of UBER. It might be legally organized as de-consolidated national entities as evident from the intra firm analysis carried in section 2. But operationally there is a high degree of cooperation within the group in terms of managerial and operational functions with shared economic risks for such integrated activities. In essence, this operational consolidation leads to the de-facto cross-border merger of relevant business activities within the group [36]. Thus, the level of operational consolidation within UBER group entities pose significant challenges to both conceiving and working with a fiction of separate legal entity approach. These challenges are twofold. Firstly, this operational consolidation could be a bottleneck for conceiving neatly segregated legal entities and whether the high level of integration achieved with shared economic risks for operations could be captured adequately within the framework of the legal fiction of independent entity. Secondly, whether the de-facto cross-border merger of relevant business activities within the group help us in creating a separate fictional PE of one single entity within UBER group.
4.7.2 Separate Independent Entity Rule and UBER
This fictional rule of independent entity is codified in article 7 of the models in case of branch and article 9 of the models in case of subsidiaries. The first step in the application of an independent entity of group rule is subjecting that entity to a factual and functional analysis. The crucial element of such analysis is the identification of significant people functions
