Characterizing Cloud Transactions — Applying §7701(e) to Remote Access Transactions: Part I

Trust Bloomberg Tax for the international news and analysis to navigate the complex tax treaty networks and global business regulations.

By Gary D. Sprague, Esq.  

Baker & McKenzie LLP, Palo Alto, CA

One of the most actively debated issues in the area of "cloud computing" is the characterization of payments made for limited term access to content, software, or other functionality provided from a location remote from the user. As just one of many business models raising this issue, a provider may create a platform including hardware and application software that allows a user to store and process data on that platform through remote access. The characterization issues are more complex when the user is able to manipulate remotely the data or application, or even manage the operation of the hardware itself, in contrast to merely obtaining access to or receiving a download of the content. This article discusses whether §7701(e) provides useful guidance to characterize these transactions.

In most commercial cloud computing transactions falling in the category described above, the characterization choice normally is between "service" or "lease" characterization.  To get to that choice, however, other characterization possibilities first need to be eliminated. If a copy of the content is delivered to the user under use rights that allow perpetual use but no further market exploitation, the transaction is likely to be characterized as a sale of the copyrighted article.1 If the transaction grants to the user significant copyright or other intangible property rights that the user may then exploit on the market, the transaction is likely to be regarded as a license producing payments characterized as royalties for tax purposes. Where there is neither the transfer of a digital copy, nor the grant of sufficient rights to constitute a license, many remote access transactions of the sort described above seem to exhibit features of both service and lease transactions.

So what rules should apply to distinguish between those two characterizations?

Remarkably, a distinctly pre-cloud Code provision seems to address this issue. In 1984, Congress enacted §7701(e) to respond to various controversies that had arisen regarding whether the provision of equipment constituted a service or a lease of that equipment for purposes of the investment tax credit (ITC).2 By its terms, §7701(e) seems to apply to distinguish service from lease transactions for all purposes, not just for purposes of the ITC. Pre-§7701(e) cases applied the same factors that are listed in §7701(e) to distinguish service contracts and leases outside the ITC context. Subsequent to the enactment of §7701(e), the IRS specifically stated that the factors described in §7701(e) are relevant to determining whether a telecommunications provider's gross receipts are derived from providing telecommunication services or from a lease or rental of property for §199 purposes.3 Finally, the Fifth Circuit expressly concluded in the recent Tidewater case4 that the §7701(e) factors apply outside the ITC context, as it referred to those factors to determine whether a vessel charter should be characterized as a lease or service transaction for purposes of the foreign sales corporation rules. Accordingly, §7701(e) appears to be the most relevant existing statutory guidance for characterizing remote access transactions.5

Section 7701(e)(1) provides that "a contract which purports to be a service contract shall be treated as a lease of property if such contract is properly treated as a lease of property, taking into account all relevant factors." Factors indicating the existence of a lease include:

(A) the service recipient is in physical possession of the property;

(B) the service recipient controls the property;

(C) the service recipient has a significant economic or possessory interest in the property;

(D) the service provider does not bear any risk of substantially diminished receipts or substantially increased expenditures if there is nonperformance under the contract;

(E) the service provider does not use the property concurrently to provide significant services to entities unrelated to the service recipient; and

(F) the total contract price does not substantially exceed the rental value of the property for the contract period.

The statute is clear that this is a non-exclusive list of factors, and that some of these factors may not be relevant in particular cases. In addition, other factors have been mentioned, in analogous rulings or cases, which didn't find their way into the §7701(e) list, including:

(1) whether the service provider has the right to remove property from service and replace it with comparable property;

(2) the degree to which property is a component of an integrated operation in which the taxpayer has other responsibilities;

(3) who operates the equipment; and

(4) whether the fee earned by the service provider is based on the passage of time or the number of procedures performed.6

How well suited are these factors to characterizing remote access transactions? In the absence of anything more specific to e-commerce transactions, the Organisation for Economic Co-Operation and Development's (OECD's) Technical Advisory Group on Treaty Characterisation of Electronic Commerce Payments considered the §7701(e) factors and "found these factors to be useful for purposes of determining whether payments are for `the use of, or the right to use, industrial, commercial or scientific equipment'" for purposes of those treaties that separate payments for such equipment from other business profits.7 In dealing with actual cloud transactions under U.S. domestic law, the §7701(e) factors indeed have proven to be relevant, although some have been more useful to making the necessary classification distinctions than others.

Before any factors can be applied, the property that is in fact the subject of the analysis must be identified. In a remote access transaction, the property might be the computer server (in the case of the provision of data storage, for example), a software program (in the case of a program downloaded subject to a limited duration license, for example), or both hardware and software (in the case of a typical "software as a service" (SaaS) contract, for example).

The next step is to apply the appropriate factors to the commercial arrangement between the parties. In practice, the greatest difficulty in the remote access context has arisen in attempts to apply the first two factors of "possession" and "control." While there is no priority of application in the §7701(e) factors, the possession and control factors probably carried more significance than some of the others in the original context of equipment service/lease transactions, given that the property being addressed was a substantial physical item. In the context of remote access transactions, especially where the item being addressed is content, data, or software, concepts of possession and control are more difficult to apply, since a user's ability to manipulate the content, etc. is not significantly affected by whether the content is resident on a remote server owned by a third party or on the user's own equipment. In cases where the object of the contract is the provision of space on a server, it conceivably could be argued that the remote user has possession and control of the server, if the user can manipulate the equipment remotely. For most remote access transactions, however, that argument should not prevail.

Generalizing in the context of remote access transactions is very dangerous, as the terms of commercial transactions can take many forms, and the modifications from case to case can indeed cause a transaction to cross a characterization line. While acknowledging that risk, it seems that an appropriate "cloud" gloss on the §7701(e) factors would be that possession and control, in a transaction where the principal commercial value is the access to the provider's data, content, software functionality, and the like hosted on equipment not owned or maintained by the user who accesses the data, etc., normally should be regarded as exercised by the party that owns or maintains the equipment, not the user, unless the transaction involves a download of a copy of the data, content, or software to the user's computer. Transactions such as SaaS, subscription access to data, interactive gaming, data hosting, and other remote access transactions typically involve the provider making capital investments in both the hardware and software elements of the service. The provider normally uses its proprietary software to provide the SaaS or data access services, maintains the software as needed, and has the right to update and replace the software at will. The provider typically owns or leases the equipment on which the content, data, or software is hosted and is responsible for the continuous operation of the equipment.  Keeping the service operational normally requires active involvement in maintaining both the hardware and software elements of the service.  These facts all point towards a "services" characterization for these transactions.

One of the §7701(e) factors comes into its own in the remote access context: the "concurrent use" factor would seem to be a powerful element arguing for services characterization, where the provider is making the same content, data, or software available to many users through a single data center running the same application (even if the provider dedicates specific machines to host the data, etc. for a particular user, since the user normally would be agnostic as to which machine would be used for its data, and the data normally could be switched to other machines with no loss of utility for the user).

With these admittedly nebulous guideposts directing the analysis, some generalizations can be expressed about typical remote access transactions in the "cloud." For example, a limited duration license to use a copy of a digital product that has been transferred to the user through download typically should result in rental income. In this case, the property being considered is a copy of the digital product, since the equipment is nothing more than a delivery mechanism, and the user has possession and control of the digital product. The user has possession of the property (i.e., the application files reside on the user's computer), controls the product (e.g., playing the music files), has exclusive use of the product for a significant period of time (but not its entire useful life), and typically pays a fee based on the passage of time. The §7701(e) factors can be applied relatively easily here.

In contrast, remote access transactions such as SaaS, subscription access to data, and the like typically should give rise to services income, even when the user is able to manipulate the software application or the data remotely. If the argument above is accepted that the provider, not the user, should be regarded as in possession and control of the equipment and the application software or data in remote access transactions, the most problematic aspects of applying the §7701(e) factors have been resolved. The remaining §7701(e) factors would seem to point in the same direction; for example, the user should not be regarded as having "a significant economic or possessory interest" in the server or the application software, as the user has not made the sort of investment in those assets that indicates an "interest" in the properties. The fact that the provider allows access to the software concurrently for other customers should be regarded as a significant factor. Even if the customer pays a fee based on the passage of time rather than on the volume of transactions processed by the software, the §7701(e) factors overall point towards a services characterization for these transactions.

Finally, is this analysis consistent with recent judicial interpretations of §7701(e)? The only decided case to interpret that section outside the ITC context is the Tidewater case mentioned above. In that case, the Fifth Circuit concluded that a time charter would be characterized as a lease rather than a service contract, even though the vessel was provided to the customer complete with crew, on the basis that the customer's ability to direct when and where the vessel travelled constituted the more important "control" than the operational control exercised by the vessel's crew, and that the control factor was the most important criterion in that case.  Questions have been raised about whether this judicial view of "control" could lead to a conclusion that the user in a remote access transaction could be regarded as "controlling" the equipment, software, or data in the typical remote access transaction. In my next commentary, I'll describe why the Tidewater case should not affect the analysis proposed above.

This commentary also will appear in the September 2013 issue of the  Tax Management International Journal.  For more information, in the Tax Management Portfolios, see Sprague, Chesler, Reid, Cohen, and Hubbard, 555 T.M., Federal Taxation of Software and E-Commerce,  and in Tax Practice Series, see ¶2350, Losses, and ¶4020, Classification as a Partnership..

Copyright©2013 by The Bureau of National Affairs, Inc.

  1 See Regs. §1.861-18, dealing with the classification of certain transfers of computer programs, which provides for this result in context of computer programs (but not content).

  2 The ITC controversies arose because purchasers of equipment could not claim the ITC if the equipment was provided to a governmental or tax-exempt entity under a lease contract, as opposed to a service contract. See Rev. Rul. 68-109, 1968-1 C.B. 10. That context explains the peculiar wording of the introductory language of §7701(e), which refers to the characterization of "purported" service contracts.

  3 See Rev. Rul. 2011-24, 2011-41 I.R.B. 485.

  4 Tidewater Inc. v. United States, 565 F.3d 299 (5th Cir. 2009).

  5 For concordance with this view, see "Proposed Framework for the U.S. Federal Income Taxation of Cloud Computing Transactions," Los Angeles County Bar Association, Tax Analysts Document Number Doc 2013-16214.

  6 Sprague, Chesler, Reid, Cohen, and Hubbard, 555 T.M., Federal Taxation of Software and E-Commerce, II, A, 2, c, (2).

  7 Technical Advisory Group on Treaty Characterization of Electronic Commerce Payments, Tax Treaty Characterisation Issues Arising from E-Commerce, Report to Working Party No. 1 of the OECD Committee on Fiscal Affairs, ¶ 28 (2001), available at

Request International Tax