Electronic Services: Suggesting a Man-Machine Distinction

by Howard J. Levine, Sanford H. Goldberg, Ellen Seiler Brody
Published: August 15, 1997
Source: Journal of Taxation


The movement in the United States from an industrial to a service economy has been markedly accelerated. Innovations in technology, based upon the digitalization of data, have increased productivity and competitiveness and introduced a flexibility in how activities are performed. Development of the Internet and other electronic networks has facilitated the ability to operate a global business from a single, multiple or even mobile sites. The increasing uses of electronic networks to communicate, produce, and consume raise novel practical as well as theoretical tax issues for administrators, policy makers, planners, and taxpayers.

The Treasury Department's "White Paper" on electronic commerce(1) suggests that governments can and will address these issues as they arise. The White Paper's advocacy of what might be referred to as "digital-analog neutrality," namely that transactions occurring electronically and the "traditional way" be treated similarly for tax purposes, together with its conclusion that there need be no new taxes on Internet transactions, make common sense. The proposed regulations characterizing income from computer software transactions(2) (the "Software Regs") underscore the importance of distinguishing between services rendered and transactions that involve the exploitation or disposition of intellectual property. The Software Regs assume that once the proper characterization of a transaction has been established, current tax law is adequate to deal with the various factual situations likely to arise and to reach the proper tax results. Nevertheless, due to the complexity of the rapidly changing medium and what seems like unlimited flexibility, the challenge of addressing the taxation of services in the context of electronic commerce ("electronic services") seems daunting.

This article is based upon a rather radical view, namely that, for tax purposes, the category of personal services is not really expanding, it is contracting. At the same time, a new category of economic activity, arising from the exploitation of the expanding classification of intellectual property, is emerging. From this perspective, the fundamental framework of the U.S. income tax system would still appear to provide workable rules for application in domestic and international operations involving electronic services, even though the existing tax rules were developed in an era when manufacturing was the predominant aspect of economic commerce.

With a view towards developing an approach to the appropriate tax principles that should apply to electronic services on a global basis, this article first considers the kinds of electronic services that are now available. By offering a new classification of what constitutes personal services, this article then suggests that where mechanical (including software) systems, rather than human beings, are providing an electronic service, the existing place-of-performance rules cannot and should not apply. Instead, rules applicable to the exploitation of intellectual or tangible or intangible personal property should apply, depending on the proper analog. These suggestions are buttressed by a review of the current U.S. and international tax rules applicable to services and intellectual property generally.

Are Electronic Services Personal Services?

At first glance, the creation of electronic services and the delivery of products over electronic networks such as the Internet seem to create novel types of income, unlike those seen before. But what really is occurring involves bifurcated developments -- one development involves growth of traditional services rendered in a nontraditional manner, and the other development involves emergence of a new class of mechanized/software products which only on first glance appear to fall within the services category. In a colloquial sense, the entire experience of interfacing with an electronic network can be viewed as involving new and different "services." A closer look quickly reveals, however, that in many cases, these electronic interactions merely involve more efficient communication.(3) Thus, for example, the Internet (or an intranet) may facilitate production and delivery of the professional services of an architect, attorney, or engineer, which traditionally have been delivered through face-to-face meetings, telephone conferences, telexes, faxes, and "hard copies." Such traditional personal services are referred to in this article as "Type I" services. A hallmark of a Type I service is that at least some significant aspect of the service is provided directly (although perhaps on a deferred basis) by a human being.

The second category of electronic "services," referred to here as "Type II" services, generally involves the use of electronically accessed data bases. These "services" are largely mechanized and do not require any significant direct human input, One example of Type II services is provision of information searches (e.g., legal, medical or financial research, confirming or providing credit histories, or providing prerecorded information like diagnostic assistance or technical support). Such Type II services may prove to be little more than a substitute for purchasing books or magazines, renting a video tape, or paying for access to an index (even though a proprietary "search engine" may be used). To illustrate this point, consider how similar the use of a printed telephone directory is to accessing on line a telephone-listing data base. Should it matter that the on-line directory is "interactive," particularly where no human being is rendering any direct services?

Other Type II services involve a limited amount of relatively nondiscretionary human activity, such as on-line security monitoring, on-line gambling, and on-line banking. Now that computers can best human chess champions, we anticipate that the human element in many of these Type II "services" might someday disappear altogether, as the category of Type II services grows and the category of Type I activities shrinks.

Do We Need a Whole New Set of Rules?

International services income is generally subjected to income tax in the U.S. and sourced by reference to the place of performance. Other international income (e.g., passive income from royalties) is currently subject to source-based taxation (generally without the benefit of deductions), or net-basis taxation on income effectively connected with a U.S. trade or business or, under treaties, attributable to a permanent establishment (a "PE") in the U.S. With some notable exceptions (e.g., Brazil, India), these rules are fairly common worldwide. Different nations may have different rules for determining when a separate charge is required for the service component of a broader transaction, however, and may characterize income from any given activity differently. As suggested further below, taxpayers may be inspired to exploit the differences found in such "hybrid" cases.

Given the difficulties in determining the performance location of a broadly defined category of electronic services, the White Paper suggests the place-of-performance rule should be changed or redefined. Eliminating the place-of-performance rule, however, could create more uncertainty than clarity in the tax system. We believe that a better approach is to restrict the place-of-performance rule to "true" personal services -- namely the income attributable to (generally nonministerial) human activity (Type I services) and to subject the growing category of Type II services to the existing (and future-refined) international tax rules applicable either to the exploitation of intangible property, including its creation (e.g., cost sharing), licensing, lease, or sale, or to the disposition of tangible or intangible personal property, depending upon the particular Type II service involved. Some of these choices are explored further below in the context of examples.

As suggested above, it would be very difficult to determine the place of performance in the case of electronic research, Internet gambling, or on-line banking, where all important activity of the provider -- except perhaps a limited amount of supervision -- has been mechanized. In contrast, it should be possible to determine the location of an immediate (or initial) consumer in most Type II services. Although shifting the source of income to the residence of the immediate consumer, or the place of delivery, could be subject to some manipulation, once Type II services are treated as "other than personal services," it becomes clear that the residence of the consumer can be highly relevant (consider, for example, the state-of-consumption source rule applicable to royalties(4)).

If the White Paper's common-sense plea for digital-analog neutrality is to be adopted broadly, then it may soon become necessary to ignore cases where only ministerial activities are performed by human beings in connection with Type II services. One reason for this conclusion is that, sadly, it may soon not be possible for either the recipient or the renderer to determine whether there is any (and, if some, how much) direct human input there is in the Type II services being rendered. These observations suggest that it will be necessary to apply an appropriate model of taxation to Type II services, such as the rules applicable to royalties, where appropriate.

If Type II services involve the delivery of protected trade secrets instead of personal services, should the royalty rules apply to determine the source and character of the income? In a case where no human services are performed, is it appropriate for Type II services to attract little or no worldwide taxation? Another question -- based on unresolved questions of current law -- is what if the Type II service provider subcontracts functions that are, in fact, performed by human beings? Should the Type II service provider's tax result be different than where the subcontractor performs all of its functions digitally? We appreciate that by drawing this distinction between Type I and Type II services, and by treating those types differently for tax purposes, it will be necessary to resolve questions like these.

Consider a case where a corporation that is organized under the laws of the Netherlands ("Neth Co.") develops or otherwise acquires computer software to replace all of its human security monitors. Assume that sales into the U.S. are contracted for over the Internet and that the critical software is placed on mirror servers located in different tax havens. Under conventional rules, the Type II service provider would not be considered to have a U.S. or Netherlands's PE, yet it is far from clear that no tax should be paid on such income anywhere in the world. In this case, the consumer is presumably indifferent to (and does not know) where the monitoring takes place and whether or not human monitors are actually being used. If Neth Co. hired human beings to monitor the site, in order to avoid taxes under current law the human beings would have to be located in a tax haven. The beauty of the distinction between Type I and Type II services is that in the case of a Type I service, both the renderer and recipient should be aware that the renderer is using human beings to render the service, and at least the renderer should be able to determine where those human beings are located at all relevant times.

Distinguishing Services from Intangible Property Under Current Law

The case law has long recognized that entities as well as individuals can render "personal services" as that term is used in the Code. This presumably occurs through performance by their dependent agents, such as officers and employees. In Hawaiian Philippine Co. v. Commissioner, 100 F.2d 988 (9th Cir. 1939), for example, the court held that a commission earned by a corporation for processing services was compensation for "personal services."

In many cases, it is not obvious when income is attributable to services or when it is attributable to the exploitation or disposition of property. The tax law has long relied upon distinctions drawn by legally predictable rights to distinguish property from services for tax purposes. Nevertheless, efforts of individuals are normally required to generate intangible property. For example, the Second Circuit held in Ingram v. Bowers(5) that payments designated as royalties and paid to the performer Caruso actually represented income from the performance of personal recording services rendered in New York. The Court drew this conclusion largely from the fact that Caruso had no legal interest in the masters he recorded. He had contracted to perform the services that led to the recordings, rather than independently creating the recordings and then trying to find a purchaser or licensee. The Court found that the proper source for taxation purposes in that case depended upon where the creative processes occurred, rather than how the intellectual property was exploited (e.g., by lease or disposition of the recordings). Compare this with Rafael Sabatini, 32 B.T.A. 705 (1935), aff'd 98 F.2d 753 (1938), where an author "sold" the distribution and motion picture rights to his novels in return for amounts characterized by the courts as royalties.(6)

Even within the category of intangible property, there is often a tension under current law between whether an item is sold or whether it is licensed. If the intangible is licensed, payments for the use of the property are generally sourced where the intangibles are used or where the rights to such use are granted.(7) When the items are sold, however, the income is normally sourced at the place of sale under the longstanding title passage rule. The difference between a license and a sale for tax purposes typically depends on what, if any, rights the transferor retains in the property, a distinction also found in non-tax law.(8)

Under current U.S. law, a corporation could be deemed to be engaged in a U.S. trade or business through Type I services performed by an independent agent. Cf. DeAmodio, 34 T.C. 894 (1960), aff'd 299 F.2d 623 (3rd. Cir. 1962); Miller v. Commissioner, T.C. Memo 1997-134 (no withholding on amounts paid for services where subcontractors performed all the work and some contractors used U.S. labor). Under a typical treaty, however, no PE will be attributed to an independent agent unless the agent has the authority to conclude contracts in the name of its principal. Presumably, even activities excluded from a PE are subject to tax at the location of residence of the enterprise.

In Taisei Fire and Marine Insurance Co., Ltd. v. Commissioner, 104 T.C. 535 (1995), acq. 1995-2 C.B. 1, the use of an independent agent to underwrite insurance for the taxpayer in the United States did not rise to the level of a PE. Similarly, in Rev. Rul. 77-45, 1977-1 C.B. 413, the IRS ruled that a Canadian consulting engineering corporation, which was engaged to plan and design a U.S. manufacturing plant and performed the majority of its services in Canada, did not have a U.S. PE under the former treaty. The fact that the client provided it with work space at the construction site and that the company sent employees into the United States to inspect performance and quality of materials, to make minor adjustments to the plans, and prepare progress reports, was insufficient to rise to the level of a permanent establishment under the treaty. The duration of the construction was to last less than one year, and the activities conducted in the U.S. were solely of a planning and supervisory nature.

Separating Services from Sales of Products

Once it is determined that services have been rendered, when does the tax law currently require their separation from the delivery of a product? Much of the authority in this area arises in the context of section 351, which permits the tax-free transfer of "property" but not services, to a corporation. If services are also rendered in connection with the transfer of property in exchange for stock of the recipient, the transfer is generally tax free under section 351 as long as the services are merely ancillary or subsidiary to the property transfer.(9) Rev. Rul. 64-56, 1964-1 C.B. 133. See also OECD Model Treaty, Art. 5(3) (1995); U.S. Model Treaty, Art. 5(3) (1996). If the services are not ancillary or subsidiary, however, an allocation of the stock value must generally be made between the property and services transferred. Whether or not services can be considered ancillary or subsidiary is a question of fact. In contrast, the Software Regs would require a separate allocation to transfers of copyright rights, transfers of copyrighted articles, services, and know how, unless the item at issue is "de minimis." See Prop. Reg. Search7RH1.861-18(b)(2).

Rev. Rul. 64-56 provides examples of services that might be considered ancillary or subsidiary. Demonstrating and explaining the use of the property to promote a transaction can be considered ancillary.(10) On the other hand, teaching employees a new, necessary skill is essentially educational and is considered a separate teaching service, not part of a transfer of property. Similarly, continuing technical assistance provided after operations have begun is normally viewed as separate professional services, to which separate consideration should be allocated for tax purposes. Assistance from architects and engineers as to how to construct a plant or the proper layout of machinery and equipment is also considered the rendering of separate advisory services. See also Rev. Rul. 55-17, 1955-1 C.B. 133; Rev. Proc. 69-19, 1969-2 C.B. 201; Commentaries to Article 12, paragraph 11, 1992 OECD Model Convention.

Although these rules may prevent a taxpayer from avoiding separate services income in certain cases, it is clear that a taxpayer can normally structure a transaction, if desired, to separately allocate to the service component.

Where services are rendered by affiliates, multiple branches, or by dependent or independent agents, allocation (i.e., intercompany pricing) issues as well as agency questions often arise. (E.g., are the agent's activities attributed to the principal, and do they subject the principal to tax in the jurisdiction where the services were performed?)

Allocation of personal services income and expense.

Once it has been established that a transaction involves simply a Type I service, so that the income will be sourced at its place of performance, an appropriate allocation is still required in cases where such services are performed in more than one jurisdiction. Treas. Reg. Search7RH1.861-4(b) states that compensation for labor or personal services performed in the United States should be allocated on the basis that most accurately reflects the proper source of income under the particular facts and circumstances. See also section 863(b)(1) (allocation of income to sources where services are rendered partly within and partly without the U.S.). In the "real world," however, the allocation of both income and expense is often based simply upon time spent.(11)

Thus, if a foreign architect is hired to design and supervise construction of a building located in the United States, the architect must allocate his income to the various components of his services, including background development efforts, discussions with clients about design preferences and objectives, drafting plans and specifications, including incorporation of customer changes and local building requirements, as well as on-site time to view the site and to inspect the actual construction. The majority of the "brain power" and, therefore, value, of the architect's services may be located abroad. Yet, the allocation seems typically to be based upon the time spent on each activity, weighted perhaps by the higher salaries paid to more skilled workers; but as in many service businesses, "location savings" may exist if highly skilled draftsman charge considerably less in different jurisdictions, so a weighting by salaries does not necessarily appropriately reflect where profits truly belong.

Section 482 generally requires that compensation for personal services performed between related parties be based on an arm's length charge.(12) The arm's length charge equals the costs and deductions incurred in performing the services unless the services are considered "integral," in which case an arm's length amount is the amount that would have been charged for similar services performed for an unrelated party in a similar situation.(13) Unfortunately, little further guidance is provided for such cases. Even when services are not integral, the regulations allow taxpayers to establish alternative rates, other than costs and deductions incurred, if they can show that is appropriate.(14)

The section 482 regulations continue to outline four different tests under which services will be deemed integral to the business activity of either the party rendering the services or the party receiving the services. Under the first test, services are deemed integral if either party to the transaction is engaged in the trade or business of rendering similar services to unrelated parties.(15) Thus, a potentially independent agent will be considered to render integral services. Under the second test, services are deemed integral if rendering such services is a "principal activity" of the renderer of the services. Services are considered a principal activity if the costs associated with rendering the services exceed 25% of the total costs or deductions of the renderer for the taxable year.(16) Under the third test, services are integral when the renderer is "peculiarly capable" of rendering the services and such services are a principal element in the operations of the recipient.(17) The examples in the regulations do not provide an adequate definition of what is peculiarly capable. For instance, is at least a small amount of know how, i.e., protectible trade secrets, necessary? Under the fourth and final test, services are considered integral when the recipient has received the benefit of a "substantial amount" of services from one or more related parties during the taxable year. A recipient will be deemed to have benefited from a substantial amount of such services when its costs for the taxable year attributable to such services from related parties exceed 25% of its total costs or deductions for the taxable year.(18) Especially after the check-the-box regulations, the percentage tests are subject to taxpayer manipulation.

Contrast these rules to the alternative available to developers of intangibles who are willing to split ownership of the resulting property, if any. See generally Treas. Reg. Search7RH1.482-7 (cost sharing).

There is very little authority that sets forth the criteria for the application of an arm's length charge for services under section 482. In one important transfer pricing case, for example, the Tax Court allocated 75% of the income from a hospital management contract to a U.S. parent corporation, but the method for obtaining this result does not appear in the case.(19)

In the area of global trading, the Service has agreed in several advance pricing agreements ("APAs") to using a profit, split method reflecting the contribution of each trading location to the overall profit of the "book" of positions. In Notice 94-40(20), the IRS described several of the APAs that it entered into with taxpayers having fully integrated global trading operations covering commodities and derivative financial products. The Notice outlines the methods used in those APAs to allocate income among the various jurisdictions involved. Three factors were used in each APA: a risk factor, a value factor, and an activity factor. The relative weights assigned were different in each APA, to reflect the particular facts and circumstances of each situation, and different overall contributions to profitability. In this way, the IRS and the affected U.S. treaty partner relied on the determination of the parties to allocate income among the various taxing jurisdictions. This illustrates the importance of trader's compensation in real world allocations where services are a leading component: bonuses are normally tied to productivity and provide some indication that the profit allocation among personnel in different jurisdictions is legitimate. The allocation of income from the accompanying goods and intangibles also followed the allocation of income from services. In the vernacular of the authors, global trading normally involves a mix of Type I and Type II services (as well as actual income from sales). Like the Notice, it seems appropriate in such complex cases to use allocation methods that result in a proper reflection of the value of the functions performed by man, woman and machine.

Foreign Viewpoint

Most foreign jurisdictions appear to apply rules similar to those of the U.S. regarding the taxation of services, with a focus on the place of performance. There is, however, at least one minority view. Countries like Brazil apparently feel that the appropriate jurisdiction to impose a tax on services is the one where the recipient resides. While few nations have adopted this view under their income tax, a different conclusion normally applies for purposes of VAT and other foreign consumption taxes, just as one might expect. Differences in approach under an income tax regime provide the possibility of double taxation (or zero taxation), depending upon where the services are rendered. Thus, services rendered in Brazil to a U.S. customer would not attract any income tax under these divergent rules, while services rendered in the U.S. to a Brazilian customer would presumably result in the imposition of double tax, prior to the application of any available tax credit (but consider the source problem of the taxpayer). The absence of an income tax treaty between the United States and Brazil simply makes the situation worse.

Different tax rules applicable to the same sort of income is not the only problem or planning opportunity facing taxpayers and tax policy makers outside the world of electronic commerce. For example, taxpayers may also attempt to take advantage of different tax characterization rules between different taxing jurisdictions. For example, a construction site of limited duration normally does not give rise to a PE of the contractor under the internal laws of many nations and under the PE article of many income tax treaties. If a jurisdiction in which a major construction project (e.g., power plant) is built views the contractor as rendering a service, but the jurisdiction in which the contractor is a resident views the activity as a sale (of a completed plant), it may be possible for the contractor to avoid at least immediate taxation in both jurisdictions. Taxpayers naturally attempt to exploit such hybrid interpretations,(21) in characterizing their activities, as well as by creating hybrid entities, hybrid instruments, and by using hybrid accounting methods. Thus, while within each affected jurisdiction, the basic characterization of a transaction (e.g., as a sale, lease, license, or service) may become the key determinant of its tax treatment, consideration of the same economic activities in the other affected jurisdictions is vital to determining the overall direct tax impacts.

What Factors Should Govern Source?

The expanded reach of electronic commerce has also caused difficulty in establishing which jurisdiction's tax rules should be applicable. As noted in our previous article regarding the impact of the Internet on sales of tangible products, the commentaries in paragraph 10 to Article 5(4) of the 1992 OECD Model Convention state that a permanent establishment can exist when a business is carried on through automatic equipment when the activities of the personnel are solely to set up, operate, control and maintain the machine. Since maintenance and control can now be conducted via the Internet and other communication servers, one possibility is that nations could adopt the view that there is a permanent establishment of the service provider in each jurisdiction where it has a gaming or vending machine.

As electronic commerce grows and it becomes more difficult to determine the source of a transaction, the White Paper suggests that residence rules may become the easier standard to apply. The White Paper appears to stop short of advocating taxation at residence of the seller, however, but any other rule could result in a dramatic surrender of U.S. taxing jurisdiction.(22)

Treaties commonly require countries to sacrifice source-based taxation in favor of residence when no PE exists in the source country, so it seems plausible that the U.S. could get agreement from many of its treaty partners to adopt a similar approach. In adopting a residence standard, however, it is not at all clear that countries will or should relinquish their ability to tax electronic services rendered on sales made to customers in their country. The Type I - Type II distinction may help solve at least some of the apparent dilemmas here.

Nor is the White Paper clear on whether residence is meant to be determined under internal law (e.g., section 865, which does not include a branch or PE rule for U.S. corporations) or under the residency articles of treaties (which are often based upon internal law and rarely include a PE). If residence can be said to include a PE, however, much of the simplicity alleged to come from a residence-based source rule may not be achievable. In other words, the taxation of a PE could be viewed as being consistent with residence-based taxation (if residence includes a PE(23)) or, alternatively, as being inconsistent with residence based taxation (if residence is based upon place of formation or management and ignores a PE for source purposes).

Under the authors' view, the current taxation rules are consistent with a residence-based approach only if the location of performance is equivalent to residence. An individual who performs personal services generally performs them in the place in which she is resident, and her ability to do so increases with the use of electronic communications. If a foreign corporation provided sufficient services -- through its agents -- in another jurisdiction, it could be deemed to be a resident of that other jurisdiction and thus subjected to tax there, with the first jurisdiction providing a tax credit.

Under most income tax treaties, if a corporation of one State employs an independent contractor to perform personal services in another State, the income will be subject to taxation in that latter State only if accompanied by an additional substantial presence in that other State, such as the maintenance of a fixed base. On the other hand, if personal services are performed by an employee of the corporation in another State, the income may be taxable in that latter State regardless of whether or not it is attributable to a fixed base, unless the services are only of an auxiliary or preparatory nature.

What Will Happen to Type II Services?

Type II services are not traditional personal services, and so should be viewed (and taxed) in accordance with rules more appropriate to royalties or sales of tangible or intangible personal property. A royalty payment, for example, is sourced in the jurisdiction in which the property on which the royalty is being paid is exploited. For many Type II services, it will be relatively easy to determine where the ultimate "user" is located. Thus, a bank may perform a Type II service by establishing automatic teller machines in numerous locations. The user makes deposits, removes cash, and transfers funds at the machines' locations, even though the bank may be located in a different jurisdiction. In this situation, and assuming the customer is charged a user fee, it is easy to determine where the customer is benefitting from the service -- namely, at the location where it is "exploiting" the bank's resources. Similarly, a consumer who pays to play computer games on-line, is in the authors' view, really just licensing the software, not purchasing a "service". Therefore, the tax rules applicable to licenses would be expected to apply to govern such transactions too.

There are still difficult situations, however, such as cases involving blends of Type I or Type II services. Even the case of mobile electronic trading of securities is not simple. Arguably, under a royalty approach, for example, the proper jurisdiction to source the income is where the customer's account is located. What happens, however, when the customer takes his laptop computer on a plane and transacts various trades while flying over the ocean? Should a residence rule be applied in such cases?

Consider also the case of simple electronic research for which a fee is charged. This may be similar to purchasing rights to use an encyclopedia that is updated continuously through an online service. This is not really a personal service, but instead involves the license of a database. Under digital-analog neutrality, the source rule here should mirror the treatment of a publisher who delivers "updated" news to its customers.


It appears that despite continued movement from an "industrial" to a "service" economy, the category of pure personal services (Type I services) is probably shrinking, at least in relative terms, while the category of mechanized replacements for previous personal services (Type II services) is expanding. This apparent contradiction is explained by the explosive growth of what is probably intangible property. As legal protections afforded intellectual property grow, so does the category of intangible property under the Code and Regulations.(24) As but one example, the Software Regs are informed and guided largely by reference to U.S. copyright law. Indeed, Prop. Reg. Search7RH1.861-18(b)(1) separates services from the provision of know-how, which is defined partially by reference to matters that are subject to legal protection as a trade secret. Expansion of intellectual property protection under non-tax law is bound to continue as new knowledge gives rise to new products and their legal developers or owners seek new protection. As the categories of protected property expand, the relative amount of total economic activity devoted to human personal services will, logically, contract.

The White Paper suggests that the U.S. income tax treatment of electronic commerce may require reworking and shifting from a title-passage source rule for sales to a residence-based tax system. A logical question is whether such a shift may also be required for electronic services. The White Paper states that residence-based taxation should play a larger role as technological developments increase the capabilities of service providers. The authors believe that if residence-based taxation plays a larger role in electronic services, it should be through a narrowing of the place of performance rule to Type I services and treatment of Type II services consistently under the proper analog of intellectual and tangible personal property. The authors also believe that the current international tax rules for personal services (i.e., place of performance) should continue to apply only to Type I services, whether or not rendered or delivered in whole or in part electronically. Type II services should be treated as a license, royalty, or purchase of property, whichever is most appropriate.

As long as distinctions continue to be drawn between the taxation of services and the taxation of sales of intangible and tangible products, however, tax tensions (meaning planning opportunity and the risk of double taxation) will continue to exist in "cyberspace," as elsewhere. This article merely attempts to draw the line at some discernible point between man and machine, and to explain how such a distinction can and should be useful to tax planners, administrators, and policymakers. By creating global distinctions between Type I and Type II services, it is hoped that the current tax rules, rather than a whole new set of tax rules, can continue to be applied to electronic services (as broadly defined). Preserving the existing body of international tax rules to reach predictable solutions in this era of rapidly changing technology should be viewed favorably.


1. Department of Treasury Office of Tax Policy Selected Tax Policy Implications of Global Electronic Commerce, November 1996.

2. Prop. Reg. Search7RH1.861-18.

3. Communication services (e.g., telephone, broadcast, and network service providers), permit end users to access and use electronic networks. Since the function of communication companies is to allow users to communicate, and because the substance of those communications is not relevant to these service providers, these companies may someday attract a special tax on the quantity of information sent on their lines (e.g., a "bit" tax), or on their gross revenue (e.g., from advertisers), for the same reasons that utilities are commonly subjected to excise or gross receipts taxes. The proper taxation of communications companies is not addressed in this article.

4. Cf. SDI Netherlands B.V. v. Commissioner, 107 T.C. 161 (1996).

5. 57 F.2d 65 (2d Cir. 1932). See also, Boulez v. Commissioner, 83 T.C. 584 (1984), cert. denied, 484 U.S. 896 (1987); Karrer v. United States, 152 F. Supp. 66 (Ct. Cl. 1957).

6. See also Rev. Rul. 74-555, 1974-2 C.B. 202.

7. See SDI Netherlands B.V. v. Commissioner, 107 T.C. 161 (1996) (the Tax Court rejected application of the cascading royalty theory of Rev. Rul. 80-362, 1980-2 C.B 208.)

8. See, e.g., section 1235 (Sale of patents).

9. Contrast this to the typical treaty rule that excepts from treatment of a permanent establishment any activities of a "preparatory or auxiliary" nature. See, e.g., 1992 OECD Model Convention, Article 4. There are also some broader exceptions to the treaty rules, including the fairly typical exclusion for a building site or construction or installation project that lasts twelve months or less.

10. As a corollary, when personal services are provided, such as legal advice, there might be a tangible aspect as well if a written memorandum is provided. Traditionally, the writing is viewed as ancillary to the services. No separate consideration is attributed to the "sale" of the document.

11. In most cases, such apportionment will be done on a time basis, where the amount included in U.S. income will be that amount which bears the same relation to total compensation as the number of days of performance of services in the U.S. bears to the total number of days of performance of services for which such payments are made. Treas. Reg. Search7RH1.861-4(b).

The standard for allocating expenses to U.S. business income for inbound investors is found in sections 873(a) and 882(c)(1)(A) (items must be "connected" with effectively connected income) and for foreign tax credit purposes in the regulations under section 861. Treas. Reg. Search7RH1.861-8 provides rules on how to allocate expenses incurred by a taxpayer for services rendered to a related corporation, as well as to stewardship, and supportive functions.

12. Under the transfer pricing regulations, functional analysis is applied and advance written agreements are required. Treas. Reg. Search7RH1.482-1(a)(3).

13. Treas. Reg. Search7RH1.482-2(b)(3).

14. A representative from the Internal Revenue Service stated that it is unlikely that new regulations on the allocation of services income under section 482 will be proposed in 1997. She also stated that when such regulations are eventually proposed, consideration will be given to profit-based methods and guidelines recently promulgated by the OECD.

15. Treas. Reg. Search7RH1.482-2(b)(7)(i).

16. Treas. Reg. Search7RH1.482-2(b)(7)(ii).

17. Treas. Reg. Search7RH1.482-2(b)(7)(iii).

18. Treas. Reg. Search7RH1.482-2(b)(7)(iv).

19. That case, Hospital Corp. of America v. Commissioner, 81 T.C. 520 (1983), involved a U.S. corporation that had created a Cayman Islands subsidiary to manage a hospital in Saudi Arabia. While the subsidiary performed some administrative functions under the contract, all of the substantive work was performed by the U.S. parent corporation, which had a reputation as an expert in the field of hospital administration. The management contract also referred to the experience and resources of the U.S. parent corporation. The IRS had argued that all the income earned by the Cayman Islands subsidiary should be allocated to the U.S. parent. See also Nat Harrison Assocs., Inc. v. Commissioner, 42 T.C. 601 (1964).

20. 1994-1 C.B 351.

21. Cf. India - U.S. Convention, Article 12, which provides identical tax treatment to royalties and fees for included services.

22. The U.S. has claimed all sales in space and on the high seas as U.S. source income. Section 863(d). Congress reasoned that no foreign jurisdiction had the right to tax this income and generally did not do so, and by allowing U.S. taxpayers to characterize such income as foreign-sourced, taxpayers had been able to maximize use of their foreign tax credits. See General Explanation of the Tax Reform Act of 1986, H.R. 3838, 99th Cong. 2d Sess. at 935. Residence generally now also governs the source of sales of noninventory property under section 865.

23. Entities can qualify for benefits under numerous modern treaties even if they are not "qualified residents" as long as they are engaged in an active trade or business in such State. See e.g., Luxembourg-U.S. Income and Capital Convention, April 3, 1996 at Article 24(3); Mexico - U.S. Income Convention, September 18, 1992 at Article 17(1)(c); France-U.S. Income and Capital Convention, August 31, 1994 at Article 30 (d). In the Treasury Department Technical Explanation of the U.S. Model Income Tax Convention (September 20, 1996), the commentators suggest that an active trade or business will be determined under the laws of the affected country, such as section 367(a).

24. Section 482 provides that the term intangible property has the meaning given in section 936(h)(3)(B). Section 936(h)(3)(B) states that "the term 'intangible property' means any (i) patent, invention, formula, process, design pattern, or know how; (ii) copyright, literary, musical, or artistic composition; (iii) trademark, trade name, or brand name; (iv) franchise, license, or contract; (v) method, program, system, procedure, campaign, survey, study, forecast, estimate, customer list, or technical data; or (vi) any similar item, which has substantial value independent of the services of any individual. See also new U.S. Model Treaty, Art. 12(2)(a) which defines royalties as "consideration for the use of, or the right to use, any copyright of literary, artistic, scientific or other work (including computer software, cinematographic films, audio or video tapes or disks, and other means of image or sound reproduction), any patent, trademark, design or model, plan, secret formula or process, or other like right or property, or for information concerning industrial, commercial, or scientific experience."