AustLII Home | Databases | WorldLII | Search | Feedback

University of New South Wales Law Journal

Faculty of Law, UNSW
You are here:  AustLII >> Databases >> University of New South Wales Law Journal >> 2000 >> [2000] UNSWLawJl 42

Database Search | Name Search | Recent Articles | Noteup | LawCite | Author Info | Download | Help

Li, Jinyan --- "E-Commerce Tax Policy in Australia, Canada and the United States" [2000] UNSWLawJl 42; (2000) 23(2) UNSW Law Journal 313

E-COMMERCE TAX POLICY IN AUSTRALIA,
CANADA AND THE UNITED STATES

JINYAN LI[*]

I. INTRODUCTION

Simply defined, electronic commerce (“e-commerce”) is the conducting of commercial activities through electronic means, such as the Internet and private networks, within multinational enterprises (“MNEs”). The multimedia capability of the Internet enables people in different locations to interact through voice, text, and/or image. Thus, e-commerce ignores distance, national boundaries and time differences. E-commerce makes it easier for enterprises to move their business operations across national boundaries and to shift income to tax havens. It also allows business functions of MNEs to be more integrated, while becoming more distributed throughout the world, taking advantage of local conditions, including tax benefits. For example, a MNE’s headquarters, research and development, production, customer services, administration, and financing functions may be located in different places and linked electronically to produce dynamic synergies. Global trading of financial instruments can be carried out on a 24-hour basis. Computer software, microchips and aeroplanes can be virtually designed in cyberspace, as engineers in different time zones work together through private networks or the Internet.

E-commerce directly challenges existing tax principles that were by and large conceived in an era that could not have foreseen the technological advances of the present. Virtually every aspect of the international tax system is challenged by e-commerce. In response to these challenges, tax authorities in Australia, Canada, the United States and many other countries have commissioned studies and stated their positions on the taxation of e-commerce. The stage was set for the current debate on this issue by a discussion paper released by the United States Treasury Department in 1996 – the Selected Tax Policy Implications of Global Electronic Commerce (“US Treasury Paper”).[1] The Australian Tax Office (“ATO”) published its first report on Tax and the Internet in August 1997 (“ATO First Report”) and its second report in December 1999 (“ATO Second Report”).[2] Canada released the Electronic Commerce and Canada’s Tax Administration – A Report to the Minister of National Revenue from the Minister’s Advisory Committee on Electronic Commerce in April 1998 (“Revenue Canada Advisory Committee Report”).[3] In addition, the Organisation for Economic Development and Co-operation (“OECD”) has played an important role in fostering a constructive dialogue among its member countries and between businesses and government. The Technical Advisory Group (“TAG”) of the OECD has already released draft comments on treaty characterisation of digital transactions[4] and on the definition of “permanent establishment”.[5]

This article provides a brief overview of the major international tax problems created by e-commerce and the policy position of Australia, Canada, the United States, and the OECD. The discussion is limited to income tax issues.[6] Although e-commerce also raises challenges for consumption tax, this issue is beyond the scope of this article.

II. NATIONAL TAX SOVEREIGNTY IN THE BORDERLESS WORLD OF E-COMMERCE

Taxes on international income are imposed by national tax laws. Taxation is purely a sovereign right. Each nation-state makes and enforces its own laws. In other words, true international tax law does not exist. Other than tax treaties, most of which are bilateral agreements between sovereign states, international tax law simply refers to the international aspects of national tax laws. Even tax treaties do not replace national tax laws, they merely determine which nation-state has priority where more than one country has a claim. There is no international authority that co-ordinates national tax policy and tax administration.

The notion of national tax sovereignty is challenged by e-commerce. E-commerce is truly global and highly mobile. It enables businesses to exploit differences in national tax laws in order to minimise their global tax liability. It also makes it easier for businesses to be located in tax haven jurisdictions, as e-commerce can be operated and remotely controlled by people located in other countries. Thus, the problem of e-commerce taxation requires an international solution.

National governments have recognised the inherent dilemma of applying national tax laws to e-commerce. One of the common themes that has emerged from various national reports is the need for international consensus. Thus far, consensus has been reached in respect of general principles governing the taxation of e-commerce. These principles include neutrality, certainty, fairness, efficiency and the need for international coordination.[7] There is also a common understanding that existing international tax concepts can be applied to e-commerce. At the same time, however, it is acknowledged, especially by the ATO Second Report, that the application of existing principles to e-commerce transactions is often unclear. Existing principles may have to be adapted and reinterpreted or, in extreme cases, replaced by new concepts.

The OECD has been identified as the international organisation that can best coordinate national tax policies and formulate internationally accepted rules. However, there are a number of reasons why it seems unlikely that any true international tax rules will be formulated in the near future. First, the commonly agreed principle of tax neutrality (for example, taxing e-commerce and traditional commerce in the same way) mandates an evolutionary, rather than a revolutionary approach to e-commerce taxation. This means that the ‘sovereigntist’ approach to international taxation will continue. Secondly, the OECD’s work has been limited to amending commentaries to provisions in the OECD Model Tax Convention, which forms the basis of the majority of bilateral tax treaties. As discussed below, these treaty concepts are based on physical presence of the taxpayer and on notional national tax sovereignty, and are therefore difficult to adapt to e-commerce. Finally, the OECD lacks sufficient clout, especially over non-member countries. Therefore, even if new rules are formulated by the OECD, they may not be accepted by non-member countries.

III. TAX JURISDICTION – WHICH COUNTRY HAS JURISDICTION TO TAX INCOME FROM E-COMMERCE?

A. The Problem

Australia, Canada, the United States and most other countries of the world impose tax on international income on the basis of residence taxation and source taxation. Residence taxation refers to the taxation of non-resident taxpayers based on their residence in a particular jurisdiction. Source taxation refers to the taxation of taxpayers based on the source of income. In general, residents are taxed on their worldwide income and are entitled to some relief for taxes paid to foreign countries on foreign-source income. Under the principle of source taxation, the taxing jurisdiction in which income is derived has jurisdiction to tax the income. Source taxation is generally applied on a ‘pigeonhole’ approach, that is, different kinds of income are treated differently. Investment income is generally subject to withholding taxes on a gross basis, whereas business income and income from services are subject to tax on a net basis. The source rules are tied to the characterisation of income.

E-commerce challenges residence taxation by making it easy for corporations to decide where they want to be located for tax purposes. The residence of corporations is generally determined by the “place of incorporation” test or the “place of management and control” test. The “place of incorporation” test is subject to taxpayer manipulation because e-commerce can be conducted anywhere, as it is inexpensive to establish a company in a tax haven. Also, an e-commerce business can be incorporated virtually anywhere. Therefore, the test of place of incorporation basically allows taxpayers to decide whether they want to pay tax on a residence basis or source basis.

Under the test of “place of effective management”, a corporation is resident in the country in which the board of directors holds its meetings. This test becomes difficult to apply because face-to-face meetings are being replaced by virtual meetings organised by means of conference calls, video-conferencing, interactive e-mail exchanges and other communications technology. There may no longer be a physical place for meetings between directors, shareholders and managers. In these cases, where is the central management and control? Even if it is possible to pinpoint the physical place of the meetings, that place is subject to taxpayer manipulation.

Moreover, e-commerce makes it easy for corporations to relocate to tax havens. As geographical boundaries become irrelevant, moving the location of a business may mean no more than electronically transferring key files to a new computer. It will be easier than ever to flee jurisdictions with high tax rates and relocate in those with low ones. As one commentator put it, “businesses used to be like cows in a field, waiting to be milked. Now the cows have wings.”[8] Another has said that “companies selling information over the Internet can call any place home, and the savvy ones are choosing jurisdictions with low or no taxes, financial privacy, governmental stability and decent communication systems (warm water and sandy beaches are also a plus).”[9] Finally, as discussed below, enforcing residence taxation will be more difficult, as information on taxpayers’ activities in the world of e-commerce may be difficult to obtain or verify.

Source taxation is also challenged by e-commerce. As discussed below, e-commerce makes it difficult to determine the territorial source of income. E-commerce knows no specific geographic location and transcends national boundaries. It also blurs income categories and defies income characterisation. Therefore, e-commerce allows taxpayers to manipulate the source of income to avoid source taxation.

Difficulties in applying traditional tax jurisdictional principles will cause income from e-commerce to be subject to either multiple (over) taxation or no (or under) taxation. Multiple taxation occurs in the following circumstances where:

• an e-commerce business is considered to be resident in more than one country and taxed on a residence basis;

• income from e-commerce is taxed in the source country and again in the residence country without full relief for double taxation; or

• income from e-commerce is considered to have a source in more than one country and taxed on a source basis. Under-taxation will occur where income from e-commerce is not taxed anywhere because the residence country levies no income tax and the income does not have a source in any country with a regular tax regime.

B. Policy Responses

In spite of the uncertainty involved in determining residence and the difficulty of enforcing residence taxation in an e-commerce environment, there have been suggestions regarding the possible ascendancy of residence-based taxation as a solution to the problem of taxing e-commerce. For example, the US Treasury Paper argues forcefully that a move away from source-based taxation and toward residence-based taxation is both advisable and inevitable. Paragraph 7.1.5 states:

Source based taxation could lose its rationale and be rendered obsolete by electronic commerce… U.S. tax policy has already recognized that as traditional source principles lose their significance, residence-based taxation can step in and take their place. This trend will be accelerated by developments in electronic commerce where principles of residence-based taxation will also play a major role.

Although the above statement seeks to justify the move towards residence taxation by characterising it as a continuation of current trends, the real basis for the argument seems to be the assumption that residence taxation would be more administrable than source-based taxation. The US Treasury Paper neither addresses the challenges in determining residence in the world of e-commerce nor elaborates on how residence taxation may be more administrable. It is assumed that all taxpayers must be resident somewhere. Even in cyberspace, individuals who run or operate a company generally must reside in a physical location.

Because the United States is the largest exporter in e-commerce and all other countries are source countries, it is not difficult to appreciate why the United States position has not been echoed by the ATO, Revenue Canada or the OECD.

IV. CHARACTERISATION OF INCOME: FITTING DIGITAL INCOME INTO TRADITIONAL PIGEON-HOLES

A. The Problems

The character of income is important in tax law because it affects how income is taxed. For example, income from the sale of goods is taxable on a net basis in the source country if the income is earned through a permanent establishment in that country, whereas royalties and other types of investment income are subject to withholding tax in the source country. Digital transactions must be classified as a sale of goods, service, licensing of copyright or other intangible property right, or a sale of intangible property. However, digital transactions defy the pigeon-hole approach of characterisation because of the hybrid nature of digital products, the modes of delivery, and the fact that digital products may be simply, accurately and cheaply reproduced. For example, the purchase of multiple copies may give rise to sales profit to the supplier, whereas the acquisition of an electronic version with right of reproduction may give rise to royalty income.

The most difficult and common issue at present is the characterisation of software payments. Characterisation of software transactions is difficult because it is tied to what the software transaction entails and how the software is delivered. A computer software program may contain a bundle of rights, each with different legal consequences: the copyright, a tangible good embodying the copyright, programming know-how, or technical services. Methods of delivery of software include single copy packages of a standardised product, site licenses, local area networks, enterprise licenses, electronic distribution without tangible media, reproduction by a distribution intermediary, enhancement and reproduction by a value-added reseller, bundling with hardware, limited duration licenses, custom programming, and others. Furthermore, software product deliveries may be accompanied by other ancillary transactions, such as installation, customisation of standard products, maintenance, training, development licenses, and others. Software payments can be characterised as sales proceeds, royalties or licensing fees, or service fees. There is virtually no international consensus on this characterisation. Similar problems exist in respect of other types of digital products and online services.

B. Policy Responses

On the issue of income characterisation, the consensus seems to be that the principles governing the characterisation of software transactions will also be relevant to the characterisation of other digital transactions. Thus far, only the United States has introduced legislation on the characterisation of software transactions.[10] The majority view of the OECD TAG on Treaty Characterisation Issues Arising from Electronic Commerce is, in principle, consistent with the US approach.

The US software regulations characterise software payments on the basis of the classification of software as copyright, copyrighted article (or copy of program), secret formula or know-how, or services.[11] A software transaction falls into one of the following four categories:

(a) a transfer of a copyright right in the computer program;

(b) a transfer of a copy of the computer program (a copyrighted article);

(c) the provision of services for the development or modification of the computer program; or

(d) the provision of know-how relating to computer programming techniques.

A transfer of a copyrighted article is generally categorised as a sale of goods. A transfer of a copyright right is further distinguished between a sale and a license. A transfer of information that is not protected by copyright laws may be characterised as a transfer of know-how or services.

A fundamental distinction is made between transfers of copyrighted articles and transfers of copyright rights, depending on whether any of the copyright rights has been transferred. These rights include: [12]

• the right to make copies for distribution to the public,

• the right to prepare derivative programs,

• the right to publicly perform the program, and

• the right to publicly display the program.

If any of the copyright rights is transferred, the transaction is a transfer of a copyright right; if none of the copyright rights is transferred, the transaction is a transfer of a copyrighted article. A transfer of a copyrighted article may be considered a “sale” or a “lease”, whereas a transfer of copyrights may be considered a “sale” or a “license”. Under the US software regulations, only a license of copyright rights results in royalties. The distinction between a sale and a license depends on whether “all substantial rights in the copyright” have been transferred. If all substantial rights in the copyright pass to the transferee, then the transaction is a sale. If less than all substantial rights are transferred, the transaction is a license.[13]

The US software regulations are likely to classify a transaction as a provision of services where a software company contracts with a customer to develop or modify the customer’s software program in accordance with the customer’s specifications, where the development or modification requires substantial labour, or where the developed or modified program belongs to the customer. On the other hand, where a software company sends a development engineer to the customer’s location to impart information on programming techniques, which will allow the customer to more efficiently create computer programs, the contract may be classified as a transfer of “know-how”. The term “know-how” is defined as information relating to computer programming techniques, furnished under conditions preventing unauthorised disclosure, and subject to trade secret protection.[14] A transfer of know-how gives rise to royalty income subject to source withholding tax, while the provision of services does not.

A similar approach is taken in the proposed revision to the OECD Commentary on Article 12 concerning software payments. The ATO’s views on software payments are not much different from the OECDs revised commentary. In Taxation Ruling 93/12 the ATO states that payments for the transfer of rights in computer software are royalties. These rights include the right to do any of the acts comprised in the copyright, such as modification, adaptation or reproduction. On the other hand, payments for rights in the tangible article (or program copy), or for rights to use the program, are not royalties. The ruling recognises that amounts attributable to the right to load a program onto the user’s computer would strictly be a royalty, but accepts that the amount, if quantifiable, is likely to be minimal.

Revenue Canada’s Advisory Committee does not make any specific recommendations on the characterisation of digital transactions. In general, though, the Advisory Committee is sympathetic to the prevailing view that goods that were previously sold in physical form should be treated in the same manner for income tax purposes independent of the form of delivery.[15] Hence, electronic delivery of software or services should not make any difference in characterising the transaction.

Some parallels may be drawn between digital products and payments for software. In characterising digital transactions, a distinction should be made between rights to use the copyrighted information and rights to use the copyright itself. However, there are some differences between software transactions and transactions in other digital products. For example, a software license will often specify that the original diskette or compact disk and any copies made of the software program remain the property of the licensor. In these circumstances, the consumer’s rights to use the program arise solely from the license agreement. Where the owner of digitised information grants the right to download that information, ownership of the resulting digital product will rest with the consumer and not with the vendor. In such cases, the rights to use the copyrighted information could be said to rise by virtue of ownership of the digital product and not by virtue of the any license granted by the vendor.[16] The application of traditional tax principles to digital products may lead to impractical results, with minor differences in the nature or mode of delivery of a product leading to significantly different tax results.

The OECD TAG on Treaty Characterisation Issues Arising from Electronic Commerce considers that a transaction where a customer orders and downloads digital products for purposes of copyright exploitation is a license of copyright, giving rise to royalties. On the other hand, the TAG is currently divided on the characterisation of a transaction where digital products are ordered and downloaded electronically for the customer’s own use or consumption. The majority of TAG considers that the payments made by the customer would not constitute royalties but, rather, would fall within Article 7 as business profits. The members who share that position view this type of transaction as equivalent to the electronic order processing of tangible products where products are delivered physically and consider that the mere fact that a digital product is delivered electronically should not change the treaty classification of the transaction. However, the minority view is that the payment made by the customer is royalty. Members of TAG who share this view argue that the payment cannot be seen as made to acquire a copy of the software or other digital product since that copy does not exist until it is made by the customer by copying to the customer’s hard disk or other non-temporary media. Since the customer makes the copy, the payment is made in order to acquire the right to make that copy and the payment must therefore be considered to be for the use or the right to use a copyright. The fact that the OECD’s TAG was not able to reach a consensus indicates the controversial nature of the issue.

V. THE SOURCE RULES: WHERE IS E-COMMERCE INCOME EARNED ON EARTH?

A. The Problem

Under traditional source rules, the territorial source of income is determined by the place where a taxpayer or his/her agent physically performs a function, or by the physical location of an income-producing asset. For example, the source of income from services is the place where the services are performed, the source of business profits is the place where the place of business (or permanent establishment) is located, or the source of royalties is the place where the licensed property is used. These rules were developed for a physical world where it is difficult to do substantial business operations in a country without a fixed place of business in that country.

Existing source rules are difficult to apply in the context of e-commerce. As one tax expert has put it:

To the extent commerce becomes electronic rather than physical, and to the extent that what is being sold also becomes electronic - information, entertainment, technology - the search for a physical presence such as permanent establishment takes on a touch of the quixotic.[17]

People can conduct business, negotiate, meet, revise documents, sell, plan, and otherwise conduct business functions from almost anywhere with limitless mobility and without any of the fixtures that once characterised the “place of business”. Virtual offices, virtual stores and virtual workplaces can be virtually anywhere. Where is the physical presence in a predominantly virtual world? For example, where a vendor resident in Country A sells digital goods and services on the Internet to customers in Country B, does the vendor derive any income from Country B?

One of the most difficult issues concerns the concept of “permanent establishment”. A permanent establishment is generally a fixed, and generally physical, place of business through which the business of a non-resident taxpayer is carried on. In the absence of a permanent establishment, the country where goods are sold has no jurisdiction to tax the business profit. In the context of e-commerce, what remains “physical” is the computer server that hosts the vendor’s web-site. Because computers can be moved easily, questions arise as to whether a computer is “fixed”. The Internet allows instantaneous worldwide communication at little cost, and there are generally no practical economic or technical restrictions on where the server could be located. If a computer server is considered to constitute a permanent establishment, no one should be surprised if the majority of servers are located in tax havens, or if the servers are moved around from one jurisdiction to another to avoid having a permanent establishment in any jurisdiction. Even within the same tax jurisdiction, a computer can be so mobile (moving from building to building or moving with its carrier) that it will not be geographically fixed anywhere. Much easier still, a business does not have to move computers; it can simply move its website from a server in one jurisdiction to a server in another. Such an operation could even be automated.[18]

B. Policy Responses

Most of the debate on the source of income has been centred on the definition of “permanent establishment”. Because this concept is used in tax treaties and is based on Article 5 of the OECD Model Tax Convention, a treaty solution is required. The OECD is currently revising its commentary on Article 5. The revised draft commentary states that:

• a web-site does not constitute a permanent establishment;

• a computer server may constitute a permanent establishment;

• an enterprise will not be considered to have a permanent establishment by virtue of an arrangement under which its website is hosted by an Internet service provider (“ISP”) on the ISP’s equipment, even if that equipment is located within the jurisdiction;

• except in extraordinary circumstances, an ISP cannot create a permanent establishment by agency for the enterprise whose web sites it hosts;

• a computer server does not constitute a permanent establishment if the e-commerce operations carried on through the server are restricted to the preparatory or auxiliary activities (such as providing a communications link between suppliers and customers, advertising, relaying information through a mirror service for security and efficiency purposes, gathering market data, or supplying information).

The above commentary is generally consistent with the position taken in the ATO reports, Revenue Canada’s Advisory Committee Report, and the US Treasury Paper.

The OECD believes that the principles which underlie the OECD Model Tax Convention are capable of being applied to e-commerce. It does not propose any changes to the existing definition of permanent establishment. The revised commentary does not change the fact that the permanent establishment concept is difficult to apply in the context of e-commerce. These difficulties will be compounded by the fact that there is considerable scope for an e-commerce business to structure its activities to ensure they do, or do not, result in a permanent establishment being established in the source country.

The physical location of a server is becoming increasingly irrelevant as bandwidth and response time problems are being overcome. An enterprise can access a market within a jurisdiction just as easily from a website located outside that jurisdiction as from one within that territory.[19] Therefore, the question of whether a server constitutes a permanent establishment will have little practical impact on tax revenues, as taxpayers can structure their e-commerce operations so as to avoid the need to store their data or software on servers located in countries where they do not want to risk having a permanent establishment.[20] This view is shared by the ATO, which states in its report that “measures to catch websites as permanent establishment may only provide revenue benefit in the short term and could force them offshore in the long term”.[21]

No changes have been proposed for determining the source of other types of income. The existing physical presence test applies to e-commerce transactions. That means, of course, that e-commerce businesses can virtually decide where to source their income.

VI. TRANSFER PRICING: SLICING THE DIGITAL PIE WITH
A CONVENTIONAL KNIFE?

A. The Problem

Transfer pricing, or the allocation of cross-border income among jurisdictions, is one of the most fundamental issues in international taxation. Transfer pricing rules are necessary to allocate income among jurisdictions in which a MNE operates. Most often these rules apply in the context of separate entities within a single enterprise that transact business with one another or that benefit from common expenses (such as management costs). They may also apply in the context of intra-firm (or inter-branch) transactions where a permanent establishment of a firm transacts with the head office or with another permanent establishment.

Traditionally, income from cross-border transactions has been allocated on the basis of the ‘arm’s length’ principle.[22] The underlying assumption of this principle is that members of a MNE group are separate accounting units and deal with each other at arm’s length. Allocations of income among related entities are evaluated in accordance with the way that unrelated parties transacting business with one another would structure their transactions. The establishment of arm’s length prices depends on finding comparable transactions among unrelated parties that could be used to establish the correct market price. In many cases, however, such comparables cannot be found, primarily because of the existence of intangibles that are proprietary to MNEs. This has led to the decline of the traditional arm’s length regime and to the adoption of the profit split method. The profit split method does not depend on comparable transactions, but is based on a functional analysis of each part of the MNE, resulting in the assignment of the appropriate profit to each function under a market-based return. Comparability in functions and profit margins in uncontrolled situations must be found in using the profit split method.

Implementation of the arm’s length principle has been difficult and costly to both taxpayers and tax administrations. Practical experience also indicates that the arm’s length principle provides ample opportunity for abuse. Where the transfer pricing practices differ from one country to another, the result is either over-taxation or under-taxation of cross-border income.

E-commerce will aggravate these problems. As a result of the nearly instantaneous transmission of information and the effect of the removal of physical boundaries, it may become significantly more difficult for tax administrations to identify, trace, quantify and verify cross-border transactions. Moreover, with the assistance of e-commerce, many small and medium-sized firms will expand their business beyond national boundaries, and the businesses of MNEs will become more integrated. Therefore, e-commerce will increase the number of transactions that need to be valued, and reduce the availability of comparable market prices. Automation of services and seamless integration of business functions will make functional analysis virtually impossible. Finally, transfer pricing transactions will increase in complexity, particularly if the MNE is purposefully attempting to shift income among related parties in order to minimise its global tax liability.

B. Policy Responses

By the end of March 2000, the OECD had not published any recommendations on transfer pricing in the e-commerce context. In its recent discussion draft on the taxation of global trading (a significant form of e-commerce in dealing with financial products),[23] the OECD reached the conclusion that existing transfer pricing rules and methodologies should suffice to ensure that each of the parties in a global e-commerce transaction reports an appropriate amount of income.

The US Treasury Paper notes that the increasing global collaboration facilitated by modern telecommunications may raise issues regarding the allocation of income and expenses from global dealing and other activities, but does not make a specific recommendation. The new global dealing regulations introduced in 1999 move away from the traditional arm’s length principle and allow more reliance on the use of profits comparison method and profit split methods.

The ATO First Report also recognises the difficulties of applying transfer pricing rules to e-commerce transactions, particularly transactions involving digital products or the transfer of information. The ATO acknowledges that e-commerce may alter its approach to transfer pricing issues, and increase the review of know-how transfers and the related transfer prices. Consideration may also be given in the future to the introduction of a safe harbour regime, which could assist in establishing a reasonable basis for transfer prices for e-commerce transactions between related parties.

Revenue Canada’s Advisory Committee recognises that the current transfer pricing rules may not be sufficient, but has clearly rejected the global formula apportionment as an alternative. The committee does not recommend any new substantive changes, and urges the tax administration to pursue stronger exchange of information and audit agreements and mutual collection agreements with Canada’s trading partners.

VII. TAX ADMINISTRATION: HOW CAN TAX AUTHORITIES KNOW “WHO”, “WHAT”, “WHERE”, AND “HOW” ABOUT
E-COMMERCE TRANSACTIONS?

A. The Problem

It has been said that tax administration is tax policy. A theoretically sound tax system means nothing if it cannot be effectively administered. Under the existing framework of international taxation, adherence to the principle of national tax sovereignty means that the powers of taxing authorities of any country exist only within their own jurisdiction. If taxing authorities in one jurisdiction seek international information, they must rely on the co-operation of other jurisdictions under applicable tax treaties. Where there are no treaties, or where the other jurisdiction does not agree to co-operate or does not have the requested information, difficulties will arise. The level of difficulties increases with the increase in the number of jurisdictions involved, especially tax heaven jurisdictions that do not participate in any international co-operation in tax matters.

E-commerce is multi-jurisdictional and can be easily located in tax havens. It poses great challenges to tax authorities. Effective administration relies on the tax authorities’ power and means to obtain information in order to assess a taxpayer’s tax liability by identifying taxpayers, identifying and verifying transactions, and establishing a link between the taxpayer and the transactions. E-commerce has the potential to make it difficult or impossible for tax authorities to obtain information or to enforce tax collection. Taxpayers may disappear in cyberspace, reliable records and books may be difficult to obtain, and taxing points and audit trails may become obscure.

The taxpayer’s identity may be difficult to establish. Because tax haven jurisdictions generally have bank secrecy laws and have no tax treaties, information on taxpayers established in these jurisdictions and their business transactions is shielded by these secrecy laws. A taxpayer’s identity on the Internet may have little to do with his/her real-life identity. The US Treasury Paper quoted a New Yorker cartoon featuring two dogs sitting in front of a computer with a caption that read “[O]n the Internet, nobody knows you’re a dog”.[24] If the identities behind e-commerce transactions cannot be established, they are useless as evidence, even if transaction records and contracts are available to the tax authorities.

Knowing what and when transactions are conducted by taxpayers may become very difficult. E-commerce has the potential to make audit trails disappear or less reliable. First of all, e-commerce may have an effect on tax authorities’ ability to locate and access records when they are kept electronically. The location of the records can be moved without much difficulty. The contents of the records may also be easily altered. Secondly, even when records are available to the tax authorities, they may be encrypted. These records are of little value unless the tax auditor is provided with the decryption key. Without access to the correct key, encrypted data cannot be converted to plain text without the expenditure of extraordinary computing power over long periods of time. Thirdly, electronic records may lack the attributes of authenticity and integrity. In a non-electronic environment, a taxpayer’s records chronicle all of its transactions with customers, suppliers, lenders, investors, governments, and other parties. The identity of the taxpayer and its transaction partners can be verified. Moreover, the reliability of electronic records is also questionable. It is much easier to alter e-records without leaving evidence of the alteration than paper-based records. Finally, audit trails may be vanishing. In many countries, financial institutions and other intermediaries provide tax authorities with an audit trail of most commercial transactions. Payments by cheque, debit card or credit card all create a trail of transactions. If electronic cash gains popularity as a means of payment for e-commerce transactions, that audit trail will probably vanish. E-cash is equivalent to cash and leaves no audit trail, as it can be freely circulated among individuals and businesses without involving a financial institution after its initial purchase by the individual or merchant. Unless e-cash becomes regulated, it can be totally anonymous and unaccounted for.

The way in which taxes are collected is tied to the system of information reporting and withholding of taxes by third parties, such as financial institutions and other intermediaries. For example, taxes on royalties, interest or dividends paid to non-residents are generally withheld by withholding agents. Withholding agents are generally persons who are residents of a particular country, understand their obligations, and can be identified and monitored by the tax administration. Local intermediaries who are convenient “taxing points” are, to a large extent, eliminated in e-commerce. E-commerce allows cross-border transactions to be conducted directly between the vendor and purchaser without the involvement of traditional intermediaries, such as brokers, agents, and other middlemen. The process of disintermediation leads to a rapid increase in the number of cross-border traders (individuals and small businesses) that are less sophisticated in tax compliance than traditional intermediaries. It, therefore, removes traditional taxing points and forces tax administrations to directly collect smaller amounts of revenue from a larger number of unsophisticated taxpayers. For example, traditionally, a local distributor of computer software or books may obtain the copyright from a non-resident copyright holder to reproduce the work for local distribution. The local distributor will deduct tax from royalty payments made to the non-resident. With e-commerce, local customers do not need to buy software or books from the local distributor, but obtain the products directly from the non-resident vendor’s website. If the payment made for the right to download and reproduce the software or book is treated as a royalty, the payment is subject to withholding tax. However, the customer may not be aware of the withholding obligation or get away from non-compliance, as it is difficult for the tax administration to audit small transactions.

B. Policy Responses

These administration and compliance concerns raised by e-commerce have been noted by tax authorities. For example, the US Treasury Paper flatly concedes that e-commerce developments will force tax authorities to develop new tax administration techniques. It acknowledges that the US Treasury and the Internal Revenue Service will not be able to develop or implement the necessary hardware and software on a unilateral basis and calls for private sector and international co-operation.[25] The ATO Report and the Revenue Canada Advisory Committee Report echo these views and make more specific recommendations.

More specifically, the ATO First Report recommends the adaptation of the existing tax compliance requirements to the e-commerce environment, mirroring the migration of some businesses from physical commerce to e-commerce. [26] Some examples include seeking to have Australian Company Numbers displayed on commercial Internet sites, seeking proper registration procedures for Internet businesses, and examining the e-commerce distribution chains for new intermediaries. The ATO Report also recommends that the ATO negotiate with major international credit card companies and electronic payment system providers to allow access to offshore transaction details. The ATO report broke new ground with respect to the specificity of its proposals to facilitate increased ATO scrutiny of e-commerce transactions. The main recommendations include:

• foreign credit card companies should be required to report e-commerce transaction details to the ATO;

• webshops should be licensed, reflecting their the current standard practice of registration of businesses;

• middlepersons (such as ISPs) and Online Service Providers may be used as tax reporting and collection agents;

• it may be feasible to develop a code, embedded in popular web browsers such as Netscape Navigator or Microsoft Internet Explorer, to enable reporting of financial transactions.

The way in which these administrative challenges will be resolved will significantly affect changes to substantive tax law and policy. If new taxes on e-commerce are adopted or existing principles modified, they must be capable of implementation in the digital environment. Solutions to these issues must be tailored to, and take advantage of, the technology supporting electronic commerce.

VIII. CONCLUSION

Australia, Canada and the United States have reached a consensus that the existing tax principles should apply to e-commerce and that e-commerce should be taxed on the same grounds as traditional commerce. The policy responses by these three countries and the OECD so far have been limited to reinterpreting existing tax principles in the context of e-commerce. In the long run, such an approach is likely to fail because e-commerce is a brand new medium for international business and will require fundamental changes to the international tax system. The status quo will result in more inequitable allocation of tax revenue between e-commerce exporting countries and importing countries, and increasing deflection of income to tax haven jurisdictions. On the other hand, new international tax rules are difficult to formulate until the time when the nature and full impact of e-commerce is known and when international consensus is reached on the basis on which e-commerce income is allocated among countries.


[*] Associate Professor, Osgoode Hall Law School of York University, Toronto; a senior research fellow of the Taxation Law and Policy Research Institute, Deakin University, Melbourne. I thank International Fiscal Association (Canadian Branch) for its financial assistance and Claudia Persichilli for her research assistance.

[1] Available at <www.ustreas.gov/treasury/tax/internet.html>.

[2] Both reports available at <www.ato.gov.au/drafts/ecomm/ecp>.

[3] Canada, Electronic Commerce and Canada’s Tax Administration, A Response by the Minister of National Revenue to His Advisory Committee’s Report on Electronic Commerce, September 1998 at <www.ccra.gc.ca/economy/>. This report does not represent the official position of Revenue Canada, although its recommendations have been mostly accepted by the Minister of National Revenue.

[4] OECD, TAG, Tax Notes International (3 April 2000) 1565-78 (first draft), available at <www.oecd.org/daf>.

[5] OECD, TAG, Tax Notes International (13 March 2000) 1198-1205 (second draft), available at <www.oecd.org/daf>.

[6] For further discussion, see articles cited in J Li, “Rethinking Canada’s Source Rules in the Age of Electronic Commerce, Part 2” (1999) 47(6) Canadian Tax Journal 1411 at 1418-22. See also R Avi-Yonah, “International Taxation of Electronic Commerce” (1997) 52 Tax Law Review 507; R Doernberg and L Hinnekens, E-Commerce and International Taxation, Kluwer Law International (1999); C McLure, Jr, “Taxation of Electronic Commerce: Economic Objectives, Technological Constraints, and Tax Laws” (1997) 52 Tax Law Review 273.

[7] See, for example, OECD, Committee on Fiscal Affairs, “Electronic Commerce: A Discussion Paper on Taxation Issues,” presented at the Ottawa Ministerial Conference, 10 October 1998.

[8] W Rees-Mogg, “A World of City States” National Post, 5 March 1999, p A18.

[9] M Murphy, “Cooling the Net Hype” Wired, September 1996, p 86.

[10] Treasury Regulations, s 1.861-18. These regulations were largely based on the Proposed Software Regulations issued in 1996. For discussions of these regulations, see R Chaze, R Jennings, “Guiding Taxpayers Through the US IRS Computer Program Classification Rules” Tax Notes International (4 December 1998) 1953; F M Dell, “Guide to Proposed Regs Classifying Transfers of Computer Programs,” Tax Notes (6 January 1997) 83; MJA Karlin, “Computer Program Proposed Regulations Are A Good But Cautious Start” (1997) 8 J Int’l Tax’n 64; A Levenson, A Shapiro, R Mattson and N Maguire, “Taxation of Cross-border Payments for Computer Software,” Tax Notes International (30 November 1998) 1723; LD Levin, “Tax Consequences of Electronic Commerce (Dealing with Magic)” (1997) 38 Tax Mgmt Mem 107; GD Sprague, “Proposed Regulations on Computer Software Revenue Characterization,” Tax Management Int’l (14 March 1997) 128.

[11] See Treasury Regulations, note 10 supra, s 1.861-18(b)(1).

[12] Ibid, s 1.861-18(c)(2).

[13] Whether all substantial rights have been transferred to the transferee is determined by the terms of the transfer. In general, under US law, the term “all substantial rights” in a computer program means all rights (or an undivided interest therein) that are of value at the time the rights to the computer program (or an undivided interest therein) are transferred.

[14] Treasury Regulations, note 10 supra, s 1.861-18(e).

[15] Revenue Canada Advisory Committee Report, note 3 supra at [4.2.3.1].

[16] See the ATO Second Report, note 2 supra, at [5.4.54].

[17] CI Kingson, “The David Tillinghast Lecture: Taxing the Future” (1996) 51 Tax Law Review 641 at 656.

[18] J Owens, “The Tax Man Cometh to Cyberspace,” Tax Notes International (2 June 1997) 1833 at 1847. See also JC Fleming, Jr, “U.S. Taxation of Profits from Internet Software Sales – An Electronic Commerce Case Study” Tax Notes International (16 August 1999) 675 at 677.

[19] See the ATO Second Report, note 2 supra, at [5.3.31].

[20] OECD, note 7 supra.

[21] ATO First Report, note 2 supra, at [7.2.1.7].

[22] See Article 9 of the OECD Model Tax Convention.

[23] See OECD, OECD Discussion Draft: The Taxation of Global Trading of Financial Instruments (1997).

[24] US Treasury Paper, note 1 supra at [8.4]. The same is quoted by The Economist, May 31st 1997, p 22.

[25] Ibid at [8.1].

[26] ATO First Report, note 2 supra at [11.1.2].


AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback
URL: http://www.austlii.edu.au/au/journals/UNSWLawJl/2000/42.html