Research tax credit regulations: It's time to jettison the discovery test and start over.

INTRODUCTION

"[R]esearch is the life-blood of our economic progress and ... effective tax incentives for research and development must be a fundamental element of America's competitive strategy."(2)

Since its enactment in 1981, the research tax credit contained in section 41 of the Internal Revenue Code has provided an incentive for U.S. companies to conduct U.S.-based research to develop new and improved products resulting in safer, more efficient, and less expensive products that have new or improved functions, performance reliability, and quality. The efficacy of the credit, however, is threatened by the final section 41 regulations, which were promulgated by the outgoing Clinton Administration as Treasury Decision 8930 (66 Fed. Reg. 280) on January 3, 2001. The final regulations are fundamentally flawed and, if not substantially revised, may prompt U.S. firms to consider conducting their research outside of the United States, thereby frustrating the very purpose of the credit. An essential component of an efficient credit is an understandable and administrable definition of "qualified research" that appropriately encompasses the types of research and development activities envisioned by Congress. Regrettably, the definition of qualified research in the January 2001 is deeply flawed.

Most significantly, by retaining the misconceived "discovery" test and its underlying "common knowledge" standard (both of which are discussed below), the Treasury Department missed an opportunity to rein in IRS agents who, during the past four years, have sought to limit the reach of the credit. The continued presence of the discovery test, without clear guidance, will likely be misinterpreted by IRS agents, prompting them to challenge a wide range of product development activities that are clearly within the scope of the credit. In addition, an "uncertainty test" has been added to the process-of-experimentation test that impairs the meaningful operation of the credit. Moreover, the internal-use software rules in the January 2001 regulations arbitrarily distinguish between software that provides or facilitates computer services and that which provides or facilitates noncomputer services. Finally, the regulations contain surprise recordkeeping burdens that directly conflict with the legislative history of the Tax Relief Extension Act of 1999 (the 1999 Act). In their current form, the regulations will cause R&D departments to spend the bulk of their time complying with unnecessary red tape.

One of the most frustrating and perplexing aspects of the January 2001 regulations is their issuance in final form. History teaches that defining the term "research" is a most difficult process. For example, the regulations under section 174 relating to the definition of "research and experimental expenditures" were originally proposed in 1983, re-proposed in both 1989 and 1993, and then issued in final form in 1994. The Treasury Department should have learned from and been guided by the history of the section 174 regulations when considering whether to re-propose the section 41 regulations or to issue them in final form. In going final, Treasury minimized if not ignored the strong testimony of taxpayers and practitioners at a May 13, 1999, public hearing on the proposed section 41 regulations. It thus gave short shrift to Congress's mandate that the Secretary of the Treasury "consider carefully the comments he has and may receive regarding the proposed regulations relating to the computation of the credit under section 41(c) and the definition of `qualified research' under section 41(d), particularly regarding the `common knowledge' standard."(3)

Because the regulations are inconsistent with congressional intent, highly subjective, and not workable, many commentators urged that they be promptly withdrawn and re-proposed. Fortunately, on January 31, 2001, the new Bush Administration responded to those calls by issuing Notice 2001-19, which indefinitely postpones the effective date of the January 2001 regulations. The Notice also announced that the IRS and Treasury will undertake a complete review of the regulations and reconsider all comments previously submitted; it also solicited additional comments on all aspects of the regulations and announced that any changes to the regulations will be announced in the form of proposed regulations. Finally, the Notice states that Treasury Decision 8930 will be revised so that the provisions of the regulations will be effective "no earlier than the date when the completion of this review is announced, except that the provisions related to internal-use computer software (including any revisions) generally will be applicable for taxable years beginning after December 31, 1985." The Treasury Department has requested that comments on the January 2001 regulations be submitted by April 2. All taxpayers and practitioners affected by the regulations are urged to submit comments by this deadline.

This article discusses the background of the 20-year-old definition of "qualified research" and delves into three key areas: (1) the four-part test for "qualifying research," including the discovery test and the process-of-experimentation test; (2) internal-use software rules, and (3) a new regulatory "exclusion" for lack of recordkeeping.(4) It argues that the Treasury Department should jettison the discovery test and make other changes in the January 2001 regulations.

LEGISLATIVE BACKGROUND TO THE DEFINITION OF "QUALIFIED RESEARCH"

A. Economic Recovery Tax Act of 1981

Congress enacted the research tax credit in the Economic Recovery Tax Act of 1981 (ERTA) as a key component of a multifaceted tax-reduction package intended to ensure future economic growth of the U.S. economy.(5) The tax incentives adopted by Congress in ERTA were intended to stimulate investment both in plant and equipment and in research and development, thereby increasing the efficiency, productivity, and competitiveness of U.S. firms in the global marketplace.

Concerned about the decline in research spending by industry, Congress decided to enact a "substantial tax credit" to encourage greater private research by operating businesses.(6) Without "major new tax incentives," Congress concluded, many businesses would be reluctant to allocate sufficient funds to research and development.(7) This reluctance results from the inability of the firm performing research to capture the full profit potential of the research results. In other words, normal market forces and self-interest may not produce the optimal amount of research.

In defining "qualified research" eligible for the new credit, Congress in 1981 adopted the definition of "research or experimental expenditures" used for purposes of section 174,(8) subject to three specified credit exclusions.(9) Thus, ERTA rejected narrower definitions of credit-eligible expenditures in earlier versions of the bill.(10)

Moreover, in adopting the section 174 definition for purposes of the new credit, Congress expressly cited the 1957 regulations under section 174,(11) which defined "research or experimental" expenditures as "research and development costs in the experimental or laboratory sense" and therefore swept broadly to apply to industrial and commercial research and development costs.(12) The regulations in effect in 1981 further stated:

   [The term "research or experimental expenditures"] includes generally all
   such costs incident to the development of an experimental or pilot model, a
   plant process, a product, a formula, an invention, or similar property, and
   the improvement of already existing property of the type mentioned.(13)

Thus, under the section 174 regulations, "research or experimental expenditures" includes generally all such costs incident to the development or improvement of a pilot model, plant process, product, invention, formula, technique, patent, or similar property.

IRS rulings similarly have illustrated the application of section 174 to usual and customary industrial and commercial research and development costs. The range of activities specifically recognized in these rulings includes designing, developing, fabricating, testing, and improving, and more generally has included research and development of an "investigative nature."(14) The objects of these multifold activities have included manufacturing processes, methods, and products. The historical application of section 174 in these rulings is consistent with the purposes of enacting section 174 and the broad interpretation of research or experimental expenditures in the 1957 regulations.(15)

B. Tax Reform Act of 1986

To afford Congress the opportunity to evaluate the effectiveness of the new credit in encouraging increased U.S.-based research and development expenditures by companies, ERTA provided that the credit would terminate after 1985. In anticipation of this sunset date, the House Ways and Means Subcommittee on Oversight held extensive hearings on the credit in August 1984.(16)

The 1984 hearings reflected concerns of congressional leaders and Treasury officials that the research tax credit had been claimed by taxpayers in industries other than industrial and commercial industries, such as fast food restaurants, fashion designers, and hair stylists.(17) More specifically, there was concern that the credit might be claimed for developing a "Chicken McNugget" or a "banana that tastes like an orange."(18) Treasury urged that the credit should be targeted to "truly innovative research," and suggested amending the statute to limit credit eligibility to research designed to produce "a significant technological improvement" in a business component.(19) Treasury also supported exclusions from the credit for activities such as the development of financial products, routine product development costs, costs of reverse engineering, adaptation costs, and costs incurred after the beginning of commercial production.(20)

In the Tax Reform Act of 1986 (the 1986 Act), Congress concluded that the credit should be extended for three years. In doing so, Congress emphasized that the benefits of research expenditures to the economy as a whole frequently are greater than the rewards received by those undertaking the risks of research. According to Congress, extending the credit helps ensure that adequate amounts of research are undertaken to enhance productivity.(21)

At the same time, Congress acknowledged the need for a more explicit definition of "qualified research." Referring to testimony on use of the credit, Congress expressed the belief that the initial definition "has been applied too broadly in practice," that "some taxpayers have claimed the credit for virtually any expenses relating to product development," and that research by these taxpayers "often does not involve any of the attributes of technological innovation."(22)

The approach taken by Congress in 1986 was twofold. First, Congress reaffirmed the broad section 174 definition of "research or experimental expenditures" as the basic test for identifying expenditures eligible for the credit, but supplemented it with three definitional requirements relating to the character of the information being sought through the research activities and the process involved in seeking that information. (The section 174 definition and the three supplemental tests commonly are referred to in the aggregate as the "four-part test.") Second, Congress added a number of straightforward exclusions from credit eligibility (i.e., further clarifying the application of the general four-part test) relating to adaptation or duplication of existing products; internal-use software; surveys, studies and like activities; and post-production activities.

The statutory structure adopted in the 1986 Act focuses on the nature and potential use of the information that the researcher is studying or seeking to uncover and the process the researcher undertakes, rather than on any increase in the store of knowledge that results from the research activity. In this respect, Congress rejected a Treasury recommendation to limit credit eligibility to research designed to produce a "significant improvement" in a product,(23) as well as a 1984 Senate provision referring to developing a "technologically new or improved" product.(24) Those tests would have focused on what the researcher found out and how that result differed from existing products. Instead, the only focus in the 1986 Act on product characteristics relates to distinguishing activities with a purpose relating to product functionality from activities with a purpose relating to product style.

More specifically, the three 1986 Act provisions supplementing the basic section 174 test are, as follows:

* The information being sought must be "technological in nature." This test looks to whether the research process fundamentally relies on principles of the physical or biological sciences, engineering, or computer science.(25)

* The researcher's inquiry must relate to information that could be useful in developing a new or improved product.

* The process undertaken by the researcher must constitute a process of experimentation. In other words, the researcher must evaluate more than one alternative addressing functional (rather than stylistic) aspects of a product where the means of achieving the result "is uncertain at the start."(26)

As previously explained, these definitional tests focus on how the researcher conducts the activities and the nature of the researcher's activities, not on the result of the research. Thus, although the prefatory language leading up to the first two supplemental definitional tests includes the words "discovering information," these words simply serve to link the starting term in the rule -- "research" -- to the types of information required as the subject of the research (i.e., information that is "technological in nature").

The concept of research itself fundamentally means investigating or examining something -- seeking to identify or discover information, as opposed to other activities of the taxpayer conducting research, such as marketing. Thus, the words "discovering information," as used by Congress in the 1986 Act, do not impose another layer of analysis on the word "research" used in the same sentence, or in addition to the definition of "research or experimental expenditures" in section 174, the basic requirement for credit-eligible research.

The term "discovering information" appears in the legislative history of the 1986 Act only as part of restating the technological-in-nature," test.(27) While the committee reports explain the statutory terms "qualified research," "technological in nature," "process of experimentation," "functional purpose," "computer software," and each of the new statutory exclusions, there is no explanation or elaboration of a separate "discovery" requirement. If Congress had intended to place weight on its use of the term "discovering," the legislative history could have -- and would have -- provided an explanation of it.

"Qualified research" is research that is undertaken for the purpose of discovering information (i) that is "technological in nature," and (ii) the application of which is intended to be useful in the development of a new or improved business component of the taxpayer. A business component means any product, process, computer software, formula, technique, or invention, which is intended to be held for sale, lease, or license, or used by the taxpayer in a trade or business of the taxpayer. As confirmed in the Conference Agreement on the Tax Reform Act of 1986, to be technological in nature, the research needs to fundamentally rely on principles of the physical or biological sciences, computer science, or engineering:

   The determination of whether the research is undertaken for the purpose of
   discovering information that is technological in nature depends on whether
   the process of experimentation utilized in the research fundamentally
   relies on principles of the physical or biological sciences, engineering,
   or computer science -- in which case the information is deemed
   technological in nature -- or on other principles, such as those of
   economics -- in which case the information is not to be treated as
   technological in nature. For example, information relating to financial
   services or similar products (such as new types of variable annuities or
   legal forms) or advertising does not qualify as technological in
   nature.(28) (Emphasis added.)

Thus, in the legislative history to the 1986 Act, Congress clearly and fully addressed how taxpayers should determine whether they have met the section 41(d)(1)(B)(i) requirement that research be undertaken for the purpose of discovering information that is "technological in nature." Congress believed some taxpayers were claiming the credit for virtually any expense relating to product development and wished to clarify that the credit was not available for research relying on "non-scientific" principles, such as economics (e.g., development of a financial product) or marketing. The technological-in-nature test was added to distinguish between scientific and non-scientific principles. Under the plain language of the committee report, this test is deemed met where the process of experimentation utilized in the research fundamentally relies on principles of the physical or biological sciences, engineering or computer science.

There is no indication in the legislative history that Congress had any special meaning in mind for the term "discover," other than its plain meaning. Congress did, however, have a special meaning in mind for the term "technological in nature," which it specifically defined as fundamentally relying on the principles of the physical or biological sciences, engineering, or computer science. Thus, the manner in which Congress chose to tighten the requirements was through the technological-in-nature provision, not some special higher level definition of the term "discover."

At the time, Congress was concerned that the mere use of a computer might be viewed by some as "relying on the principles of computer science." To address this concern, footnote 3 to the Conference Report on the 1986 Act specified:

   Research does not rely on the principles of computer science merely because
   a computer is employed. Research may be treated as undertaken to discover
   information that is technological in nature, however, if the research is
   intended to expand or refine existing principles of computer science.(29)

In other words, Congress identified a continuum bordered at the base with the mere use of a computer and at the pinnacle with the intention to expand or refine the principles of computer science.

The relative importance of the 1986 changes should not be overestimated. One Treasury estimate for the changes that would have been made by the 1984 Senate version of the supplemental definitional tests(30) was only $75 million a year,(31) as compared with the $1.5 billion annual revenue loss attributable to the credit at that time.(32) Moreover, in the case of most industrial and commercial firms, the 1986 Act definitional changes seem to preclude the credit for research expenditures qualifying under section 174 only to the extent the research purpose relates to style, taste, cosmetic, or seasonal design factors. For these companies, product-related research expenditures qualifying under section 174 should qualify for the credit as long as the expenditures are related to functional aspects of a new or improved product, meet the technological-in-nature and process-of-experimentation requirements, are incurred before production, and are incurred in the United States.

Congressional intent to retain credit eligibility for such product-related research expenditures of industrial and commercial firms is confirmed by the following examples in the legislative history:(33)

* Experiments undertaken by chemists or physicians in developing and testing a new drug.

* Design by engineers of a new computer system.

* Design by engineers of improved or new integrated circuits for use in computer or other electronic products.

* Development by an automobile manufacturer of a more efficient and reliable diesel fuel injector.

* Clinical testing by a pharmaceutical manufacturer to establish new functional uses, characteristics, indications, combinations, dosages, or delivery forms of an existing drug.

By contrast, the legislative history cites research to discover information relating to financial services or similar products (such as new types of variable annuities or legal forms) or relating to advertising as activities that do not qualify under the four-part test.(34)

C. 1998 and 1999 Extensions of the Credit

Scheduled to expire after 1988, the research credit subsequently was extended on several occasions, and now is scheduled to expire June 30, 2004. In repeatedly extending the research credit since 1981, Congress repeatedly reaffirmed its belief that the credit can help promote investment in research, "so that research activities undertaken [by businesses] approach the optimal level for the overall economy."(35) For example, in 1998, Congress stated that the credit contributes to "new and better products produced at lower costs," which in turn stimulates economic growth.(36)

Congress has not modified the definition of credit-eligible research since the 1986 Act. Indeed, in legislative history to the 1998 extension of the credit, Congress expressly reaffirmed the scope of the term "qualified research":

   In extending the credit, the conferees wish to reaffirm the scope of the
   term "qualified research." Section 41 targets the credit to research which
   is undertaken for the purpose of discovering information which is
   technological in nature and the application of which is intended to be
   useful in the development of a new or improved business component of the
   taxpayer. However, eligibility for the credit does not require that the
   research be successful -- i.e., the research need not achieve its desired
   result. Moreover, evolutionary research activities intended to improve
   functionality, performance, reliability, or quality are eligible for the
   credit, as are research activities intended to achieve a result that has
   already been achieved by other persons but is not yet within the common
   knowledge (e.g., freely available to the general public) of the field
   (provided that the research otherwise meets the requirements of section 41,
   including not being excluded by subsection (d)(4)).(37)

In summary, Congress in this report language, reaffirmed that evolutionary product development activities fit squarely within the definition of qualified research. Congress in no way suggested that the scope of the credit should be restricted beyond the limitations in the 1986 act.

In the 1999 Act, Congress extended the credit for a five-year period. The Conference Report on the 1999 Act included the following language addressing the anticipated final regulations:

   In extending the research credit, the conferees are concerned that the
   definition of qualified research be administered in a manner that is
   consistent with the intent Congress has expressed in enacting and extending
   the research credit. The conferees urge the Secretary to consider carefully
   the comments he has and may receive regarding the final regulations
   relating to the computation of the credit under section 41(c) and the
   definition of qualified research under section 41(d), particularly
   regarding the "common knowledge" standard. The conferees further note the
   rapid pace of technological advance, especially in service-related
   industries, and urge the Secretary to consider carefully the comments he
   has and may receive in promulgating regulations in connection with what
   constitutes "internal-use" with regard to software expenditures. The
   conferees also wish to observe that software research, that otherwise
   satisfies the requirements of section 41, which is undertaken to support
   the provision of a service, should not be deemed "internal-use" solely
   because the business component involves the provision of a service.

   The conferees wish to reaffirm that qualified research is research
   undertaken for the purpose of discovering new information which is
   technological in nature. For purposes of applying this definition, new
   information is information that is new to the taxpayer, is not freely
   available to the general public, and otherwise satisfies the requirements
   of section 41. Employing existing technologies in a particular field or
   relying on existing principles of engineering or science is qualified
   research, if such activities are otherwise undertaken for purposes of
   discovering information and satisfy the other requirements of section 41.

   The conferees also are concerned about unnecessary and costly taxpayer
   recordkeeping burdens and reaffirm that eligibility for the credit is not
   intended to be contingent on meeting unreasonable recordkeeping
   requirements.(38)

ANALYSIS OF "QUALIFIED RESEARCH" UNDER THE JANUARY 2001 REGULATIONS

A. "Four-Part Test"

1. Section 174 Eligibility

The first prong of the four-part general test for credit eligibility, under section 41(d)(1)(A), is the requirement that expenditures must qualify as "research or experimental" expenses under section 174. Treas. Reg. [sections] 1.174-2(a) provides that expenditures qualify as research or experimental expenses if they are for activities "intended to discover information that would eliminate uncertainty concerning the development or improvement of a product." Appropriately, Treas. Reg. [sections] 1.41-4(a)(2)(i) did not modify this requirement.

2. "Discovering Information Which is Technological in Nature"

(a) "Technological in Nature." The second prong of the four-part test for credit eligibility is the section 41(d)(1)(B) requirement that research must be undertaken for the purpose of "discovering information which is technological in nature." This requirement was one of the primary means whereby Congress in the 1986 Act targeted the credit to appropriate scientific and engineering-intensive activities. The legislative history states that the requirement that research be undertaken for the purposes of discovering information that is "technological in nature" is deemed satisfied where "the process of experimentation utilized in the research fundamentally relies on principles of the physical or biological sciences, engineering, or computer science...."(39) Treas. Reg. [sections] 1.41-4(a)(4) appropriately incorporates this language from the 1986 Act.

(b) "Discovering Information." One of the most troublesome aspects of the January 2001 regulations is their isolating the words "discovering information" and treating those words as introducing a new test for credit eligibility. Treas. Reg. [sections] 1.41-4(a)(3) defines those words to mean research undertaken to "obtain knowledge that exceeds, expands, or refines the common knowledge of skilled professionals in a particular field of science or engineering [the "common knowledge" standard]." Neither the statute nor the legislative history to the 1986 Act discusses such a standard.

Effectively, the January 2001 regulations bifurcate the second prong of the four-part test under section 41(d)(1)(B) into two parts: (1) a requirement that the taxpayer undertake to "discover information," and (2) a requirement that the information be "technological in nature." The January 2001 regulations suggest that the legislative history's discussion of section 41(d)(1)(B) (deeming this test met "where the process of experimentation utilized in the research fundamentally relies on principles of the physical or biological sciences, engineering, or computer science") should be read to apply only to the latter requirement (that the information be "technological in nature"). The requirement that a taxpayer undertake to "discover information" would be determined under the regulations by reference to the new "common knowledge" standard.

(c) "Common Knowledge" Standard. The reference to "common knowledge" appears to be drawn from the legislative history to 1998 legislation extending section 41, which states in part that "research activities intended to achieve a result that has already been achieved by other persons but is not yet within the common knowledge (e.g., freely available to the general public) of the field" are eligible for the credit.(40) This legislative history makes clear that the credit is not precluded merely because the result being sought by the research has been achieved by another taxpayer, so long as the general public did not have free access to how to achieve the result.

The "common knowledge" standard in the January 2001 regulations warrants careful scrutiny and further elaboration. Unfortunately, the regulations merely define "common knowledge," as follows:

   Common knowledge of skilled professionals in a particular field of science
   or engineering means information that should be known to skilled
   professionals had they performed, before the research in question is
   undertaken, a reasonable investigation of the existing level of information
   in the particular field of science or engineering. Thus, knowledge may, in
   certain circumstances, exceed, expand, or refine the common knowledge of
   skilled professionals in a particular field of science or engineering even
   though such knowledge has previously been obtained by other persons. For
   example, trade secrets generally are not within the common knowledge of
   skilled professionals in a particular field of science or engineering
   because they are not reasonably available to skilled professionals not
   employed, hired, or licensed by the owner of such trade secrets.(41)

The examples set forth in Treas. Reg. [sections] 1.41-4(a)(8) do not fully or satisfactorily address this standard. Examples 1-3 simply state the conclusion that the capability and method for using a less expensive material (Example 1), using a material that the taxpayer had not previously used (Example 2), or building a bridge that can sustain greater traffic flow without deterioration than can existing bridges (Example 3) should (Examples 1 and 2) or should not (Example 3) have been known to skilled professionals. Examples 1-3 contain no explanation or description about how or why this information should or should not have been known. Without this information, Examples 1-3 are of little utility.

The regulations' definition of "common knowledge of skilled professionals in a particular field of science or engineering" brings new uncertainty to the qualification requirements, such as:

* How will IRS agents be expected to establish or maintain expertise regarding the common knowledge of skilled professionals in a given field of engineering, e.g., bioengineering?

* Who will determine what constitutes "common knowledge" in a particular field of science or engineering? Such an elusive standard is a recipe for disaster and will predictably lead to continuing protracted clashes between taxpayers and the IRS. Opinions can and will vary wildly and taxpayers will end up with disparate treatment depending on which field agent or engineer the IRS assigns to a case. Taxpayers and the government will likely become enmeshed in costly litigation -- employing outside "experts" -- over the state of common knowledge at a particular time.

* How is "should be known" to be applied? Will this standard be applied equally to small and large taxpayers?

* What requirements are necessary to be considered a "skilled professional"? Will an undergraduate degree be required? Will an advanced degree be required? Can a person fresh out of school be considered a "skilled professional" immediately, or must some length of employment be completed first (like an apprenticeship)? If someone is considered a "journeyman," is that sufficient to be a skilled professional or must one be a "master" or an "expert"?

* What is a reasonable investigation? How exhaustive must a search be? Will a literature search suffice? An Internet search? A patent search? Thankfully, there is only one required investigation, before a project is started, and there is no requirement to continually or periodically update the investigation.

* Can these determinations be made under a single taxpayer's activities?

* How can the determination practically be made on a project-by-project basis? Many taxpayers have hundreds, or thousands, of projects going on simultaneously with many projects overlapping each other. How can the required investigations be completed and documented in an efficient manner without interfering with the research process?

* How can companies that use cost center accounting, recognized in the IRS's research credit audit plan as acceptable and used in many industries (e.g., the high technology and pharmaceutical industries), practically make these determinations?

Thus, this broad standard, without more objective guidance, holds the potential of significantly impairing the ability of taxpayers and the IRS to administer the credit.

(d) The "Common Knowledge" Standard May Be Converted to a "Technological Advancement" Test. Absent additional clarification, the term "common knowledge" will be open to broad interpretations that could threaten the credit eligibility of many product development activities. Emboldened by the presence of this "common knowledge" standard in the January 2001 regulations, overzealous IRS agents may convert this into a "technological advancement" test for credit eligibility under section 41.(42) IRS agents likely will demand that taxpayers produce a patent or demonstrate that they have advanced the state of the art and the reasons for not applying for a patent. This would be catastrophic because there simply is no requirement that a taxpayer's activities must be intended to produce a "technological advancement" in order to qualify for the credit. The term is not part of section 41 or its legislative history.

Consider the 1994 statement of the Treasury Department to the House Ways and Means Subcommittee on Select Revenue Measures on this very point:

   ... the credit has never been expressly limited to research resulting in
   innovative products that advance the general state of technology.(43)

Indeed, a "technological advancement" standard would be inconsistent with the current requirements for credit eligibility. Under the legislative history explanation of the 1986 Act, the term "innovative" appears as one additional test that internal-use software development activities must meet in order to qualify for the credit.(44) The logical conclusion is that there is no general "innovative" or "technological advancement" requirement applicable to product development activities other than internal-use software development. Introduction of a generally applicable "innovative" test would render meaningless (or at least redundant) the "innovative" test applicable to internal-use software.(45)

A logical reason why Congress did not apply a "technological advancement" standard to all product development activities is that application of such a test would raise overwhelming tax administration issues. As the Treasury Department observed in 1994, the concept of "technological advancement" as a test for the credit would be entirely subjective, open to different interpretations by different observers:

   Such a limitation would be difficult to apply, because it would put the IRS
   in the position of evaluating the extent of the technological advancement
   achieved in the development of various products.(46)

If such a standard were applied, IRS examiners and corporate tax departments would be embroiled in continual controversy over what constitutes a "technological advancement." Proper administration of such a test would presuppose that IRS representatives and corporate tax departments are familiar with engineering practices in many industries, with the current state of science in numerous fields, and with the functional characteristics of all products to which research activities of U.S. companies relate. The result would be more protracted audits and more costly litigation.

Moreover, Congress has not discussed a technological advancement standard on any of the many occasions where it has revisited the scope of credit-eligible research. For example, neither the 1998 nor 1999 legislative history makes any reference to a technological advance as a prerequisite for credit eligibility. Indeed, the 1998 Act legislative history states that "evolutionary research activities intended to improve functionality, performance, reliability, or quality are eligible for the credit."(47) In other words, Congress views the credit as applicable in the context of evolutionary product and process development activities, which might not be viewed viscerally as constituting a technological advance.

The definition of "qualified research" under section 41 begins with the requirement that the activities must qualify as research expenditures under section 174. Treasury regulations under section 174 state explicitly that there is no requirement under section 174 that a taxpayer demonstrate a "technological advancement":

   Whether expenditures qualify as research or experimental expenditures
   depends on the nature of the activity to which the expenditures relate, not
   the nature of the product or improvement being developed or the level of
   technological advancement the product or improvement represents.(48)

(e) The Discovery Test Is Inconsistent with the Statute and Legislative Intent. By retaining the controversial discovery test in the January 2001 regulations, Treasury evinced the view that the 1986 changes to the definition of qualified research introduced a new concept of a "discovery." This view is puzzling because it overlooks the fact that section 174 itself requires the taxpayer to discover information. Specifically, Treas. Reg. [sections] 1.174-2(a)(1) states in part:

   Expenditures represent research development costs in the experimental or
   laboratory sense if they are for activities intended to discover
   information that would eliminate uncertainty concerning the development or
   improvement of a product. (Emphasis added.)

Thus, the section 41(d)(1)(B) reference to research undertaken for the purpose of "discovering information" does not introduce anew the concept of a "discovery." That concept already exists under section 174. Section 41 merely delineates the type of information that the taxpayer must seek to discover (i.e., information that is "technological in nature") and how the taxpayer must go about seeking that information (i.e., by a process of experimentation). There is no elaboration on the word "discovering."

Emphasis on the term "discovery" under the section 41 regulations is a new phenomenon. In Technical Advice Memorandum (TAM) 9346006 (August 13, 1993) -- one of the relatively few rulings on the scope of the research credit -- the IRS does not mention the term "discovery" in its analysis of the requirements for credit eligibility. Instead, the TAM discusses "requirements relating to technological information, process of experimentation, and functional purposes," which are the three statutory tests that section 174-qualified expenses must satisfy. The TAM shows that, at least in 1993, there was no emphasis placed on the term "discovering" in conjunction with section 41.

Proponents of a discovery test believe that section 41 somehow requires an across-the-board percentage "haircut" from section 174. This view would be based on a visceral presumption that section 41 must set a higher standard -- to be applied to every research activity, regardless of its nature, purpose, or methodology -- for qualifying research than section 174. While the section 41 standard may be higher, there is no set ratio applicable to all section 174 activities. Had Congress intended any such concept or approach, it simply could have provided in one sentence in the statute that, after taking exclusions into account, the taxpayer's section 41 credit equaled "X" percent of its section 174 deduction. Such an approach, of course, would have been far simpler than what Congress actually did.

While the 1986 changes eliminate from credit eligibility certain types of activities that otherwise qualify under section 174, these changes do not affect, and were not intended to affect, all industries equally. Congressional and Treasury officials(49) were concerned that the credit had been claimed by taxpayers in businesses such as fast food restaurants, fashion designers, and hair stylists.(50) In some industries that are less engineering-intensive, there may be significant differences between activities qualifying for section 174 treatment and activities qualifying under section 41; in industries where engineering and the "hard" sciences are central to a business's product and process development activities, there may be relatively little difference.

The preamble to the regulations fails to mention Tax and Accounting Software Corp. v. United States.(51) The district court in that case found no support in the statute or legislative history for the requirement that "qualified research" must be undertaken to obtain "knowledge that exceeds, expands, or refines the knowledge of skilled professionals in the particular field of technology or science." The court thus awarded summary judgment to the taxpayer, finding no basis for the discovery test and expressing concern over other aspects of the IRS's proposed regulations:

   The purpose of the "technology" requirement is to eliminate the "soft
   sciences" from contention for the credit, not to focus on the word
   "discovery."

The court ruled that the test in the proposed regulation has no support in the statute or the legislative history. The court added that the "highly structured definition" of research proffered in the IRS regulations is inconsistent with Congressional intent because the proposed rules make it "virtually impossible for commercial research to qualify" for the credit. The court concluded that the decisions in Norwest Corp. v. Commissioner(52) and United Stationers, Inc. v. United States(53) did not comport with congressional intent.

(f) Section 174 vs. Section 41. The distinction between sections 174 and 41 is clearly spelled out in the statute, the legislative history, and the section 174 regulations. The section 41 statute provides three additional requirements that must be met in order for activities meeting the section 174 standard to be credit eligible:

* Section 41(d)(1)(B)(i) provides that activities must be undertaken for the purpose of discovering information that is "technological in nature." The legislative history states that the technological-in-nature requirement is deemed satisfied where "the process of experimentation utilized in the research fundamentally relies on principles of the physical or biological sciences, engineering, or computer science."(54) This requirement targets the credit to appropriate scientific and engineering-intensive industries. For example, development of a financial product, since it does not involve the "hard" sciences, would not qualify under section 41.

* Section 41(d)(1)(C) provides that activities must involve a "process of experimentation." The legislative history states that this is a "process involving the evaluation of more than one alternative designed to achieve a result where the means of achieving that result is uncertain at the start."(55) Section 174, by contrast, does not require a process of experimentation.

* Section 41(d)(1)(B)(ii) provides that activities must be undertaken for the purpose of discovering information, the application of which is intended to be useful in the development of a new or improved business component of the taxpayer. Section 41(d)(1)(3) further clarifies that these activities must relate to a new or improved product function, performance, reliability, or quality. These requirements therefore eliminate from credit eligibility research that does not have a functional purpose, such as activities relating to styling, taste or seasonal design factors. There is no functional purpose requirement under section 174.

The specific exclusions provided under section 41(d)(4) also highlight activities that may involve section 174 expenditures but do not qualify for the credit. These include the exclusions for post-commercial production research (e.g., production "de-bugging"); duplication (i.e., reverse engineering); adaptation of an existing business component to suit the needs of an individual customer; and routine testing for quality control. Many of these exclusions are straightforward and merely highlight specific types of activities that would fail the specified tests for credit eligibility.(56) For example, the act of reverse engineering an existing business component will not involve a process of experimentation.

(g) Treasury Modifications to the Originally Proposed Discovery Test. Even if the new Administration does not ultimately withdraw the January 2001 regulations, and taxpayers are forced to live with the discovery test, there is some good news in the regulations.

First, the regulations' treatment of the means of discovery is realistic. "In seeking to obtain knowledge that exceeds, expands, or refines the common knowledge of skilled professionals in a particular field of science or engineering, a taxpayer may employ existing technologies in a particular field and may rely on existing principles of science or engineering."(57) Significantly, this regulation vitiates the adverse interpretation of footnote three of the 1986 Conference Report that the court adopted in Norwest.(58)

Second, failed projects and evolutionary product development can qualify for the credit. Treas. Reg. [sections] 1.41-4(a)(3)(i) provides that "[a] determination that research is undertaken for the purpose of discovering information does not require that the taxpayer succeed in obtaining the knowledge that exceeds, expands, or refines the common knowledge of skilled professionals in a particular field of science or engineering, nor does it require that the advance sought be more than evolutionary."

In addition, a taxpayer need not be the first one to discover information. Treas. Reg. [sections] 1.41-4(a)(3)(ii) provides that "[t]hus, knowledge may, in certain circumstances, exceed, expand, or refine the common knowledge of skilled professionals in a particular field of science or engineering even though such knowledge has previously been obtained by other persons."

If a taxpayer can demonstrate that the information sought to be discovered is a trade secret not known to the taxpayer, then the taxpayer should pass the discovery test.(59) The value of this provision, however, will most likely be limited in many, if not most, cases, since the determination whether information constitutes a trade secret is highly technical, very subjective, and hugely impractical.

The January 2001 regulations helpfully contain a patent safe harbor providing that the issuance of a patent is conclusive evidence that a taxpayer has obtained knowledge that exceeds, expands, or refines the common knowledge of skilled professionals. The issuance of a patent is not a precondition for credit eligibility, nor does it establish that the discovery requirement is met with respect to all the research associated with the patentable invention (e.g., research that relates to style, taste, cosmetic, or seasonal design factors would be precluded from credit eligibility). Treas. Reg. [sections] 1.41-4(a)(3)(iv) provides, as follows:

   For purposes of section 41(d) and paragraph (a)(3)(i) of this section, the
   issuance of a patent by the Patent and Trademark Office under the
   provisions of section 151 of title 35, United States Code (other than a
   patent for design issued under the provisions of section 171 of title 35,
   United States Code) is conclusive evidence that a taxpayer has obtained
   knowledge that exceeds, expands, or refines the common knowledge of skilled
   professionals. However, the issuance of such a patent is not a precondition
   for credit availability.

Section 171 of title 35, United States Code provides that "[w]hoever invents any new, original and ornamental design for an article of manufacture may obtain a patent therefor, subject to the conditions and requirements of this title" (emphasis added) and that its provisions relating to patents for inventions "shall apply to patents for designs, except as otherwise provided." Thus, the disallowance of the safe harbor for ornamental design patents appears consistent with section 41(d)(3)(B), which provides:

   Certain purposes not qualified. -- Research shall in no event be treated as
   conducted for a purpose described in this paragraph if it relates to style,
   taste, cosmetic, or seasonal design factors.

The IRS, however, may have created an "audit nightmare" for taxpayers by stating in the background section of the preamble that factors underlying the denial of a patent application may be relevant (read by IRS agents to mean negative evidence) to the determination whether the discovery requirement is satisfied. Because the patent application process is lengthy and complex, revenue agents could potentially use this provision to protract the examination process or burden taxpayers with information document requests that may not result in any useful information for purposes of the examination.

Finally, there is a rebuttable presumption that the information sought lies outside the common knowledge if: (1) the contemporaneous documentation required by Treas. Reg. [sections] 1.41-4(d) (which, before the regulations were suspended, would have been effective for projects begun on or after March 5, 2001) includes the basis for the taxpayer's belief that obtaining this information would exceed, expand or refine the common knowledge of skilled professionals in the particular field of science or engineering, and (2) the taxpayer demonstrates with credible evidence that research activities were undertaken to obtain the information described in the documentation. The presumption applies, however, only if the taxpayer cooperates with reasonable requests by the Commissioner for witnesses, information, documents, meetings, and interviews.

Regardless of whether taxpayers accept the burdensome documentation requirements set forth in the January 2001 regulations, they may want to establish this rebuttable presumption for each and every research project. Accordingly, as long as the taxpayer cooperates with reasonable requests by the Commissioner for witnesses, information, documents, meetings, and interviews, the burden will be on the IRS to demonstrate that the information described in the taxpayer's documentation was within the common knowledge of skilled professionals, or that the research activities were not undertaken to obtain the information described in the taxpayer's documentation. This, of course, depends to some extent on what is considered "reasonable" in terms of IRS requests and may place additional pressure on taxpayers to comply with requests that are borderline.

3. "New or Improved Business Component"

The third prong of the test for credit eligibility, under section 41(d)(1)(B)(ii), is the requirement that the taxpayer must undertake activities for the purpose of discovering information, the application of which is intended to be useful in the development of a new or improved business component of the taxpayer. Treas. Reg. [sections] 1.41-4(a)(2)(ii) does not expand on this requirement.

4. "Process of Experimentation"

(a) "Uncertainty Test"

(i) The New Provisions. The final element of the four-part test is the process-of-experimentation requirement. Treas. Reg. [sections] 1.41-4(a)(5) relies heavily on language from the legislative history to the research credit in defining the term "process of experimentation," but introduces a new uncertainty requirement.

The January 2001 regulations begin by looking to section 174 regarding the type of uncertainty necessary to meet the process-of-experimentation requirement. Specifically, Treas. Reg. [sections] 1.41-4(a)(5) focuses on uncertainty about the "capability or method" of achieving a particular result. (Emphasis added.) This language is consistent with the statement in the 1998 Act legislative history that confirms that it is not necessary for a taxpayer to be uncertain about its technical ability (i.e., capability) to achieve a result:

   Activities constitute a process of experimentation, as required for credit
   eligibility, if they involve evaluation of more than one alternative to
   achieve a result where the means of achieving the result are uncertain at
   the outset, even if the taxpayer knows at the outset that it may be
   technically possible to achieve the result. Thus, even though a researcher
   may know of a particular method of achieving an outcome, the use of the
   process of experimentation to effect a new or better method of achieving
   that outcome may be eligible for the credit (provided that the research
   otherwise meets the requirements of section 41, including not being
   excluded by subsection (d)(4)).(60)

Regarding the required level of uncertainty, the January 2001 regulations replace "means" from the proposed regulations with "method." Although no definition of "method" is provided, an analysis of dictionary definitions

indicates that "method" is synonymous with "design." Webster's Third New International Dictionary defines "method" as "a procedure or process for attaining an object"; "a systematic procedure, technique, or mode of inquiry employed by or proper to a particular science, art, or discipline"; and, "orderly arrangement, development or classification."

Without support, the January 2001 regulations seek to make a distinction between a "process of experimentation" under section 41 and "research and development in the experimental or laboratory sense" under section 174. For purposes of section 41(d), a "process of experimentation" is a process to evaluate more than one alternative designed to achieve a result where the capability or method of achieving that result is uncertain at the outset. (Emphasis added.) Under the January 2001 regulations, an evaluation of alternatives to establish the appropriate design of a business component when there is no uncertainty about the capability and method for developing or improving the business component does not constitute a process of experimentation as required for credit eligibility.

For purposes of section 174, expenditures represent "research and development in the experimental or laboratory sense" if they are for activities intended to discover information that would eliminate uncertainty concerning the development or improvement of a product. Treas. Reg. [sections] 1.174-2(a)(i) provides:

   Uncertainty exists if the information available to the taxpayer does not
   establish the capability or method for developing or improving the product,
   or the appropriate design of the product. (Emphasis added.)

(ii) Background to the Uncertainty Issue. Invoking the legislative history as support, the regulations proclaim that "a process of experimentation does not include the evaluation of alternatives to establish the appropriate design of a business component, if the capability and method for developing or improving the business component are not uncertain." This new carve-out simply does not comport with the 1998 legislative history, which is directly on point. In 1998, Congress provided that "even though a researcher may know of a particular method of achieving an outcome, the use of the process of experimentation to effect a new or better method of achieving that outcome may be eligible for the credit...."(61) Thus, Congress clarified that uncertainty about the best method. (i.e., appropriate design) of achieving a result is credit eligible. This 1998 language reflects the 1986 legislative history which provides that the costs of developing a new or improved business component are eligible for the credit if the method of reaching the desired objective is not readily discernable and applicable as of the beginning of the research activities. The 1986 language further explains that true experimentation takes place when an engineer develops tests and chooses among viable alternatives. What is critical here is the process employed, not some degree of uncertainty.

In keeping with an apparent goal of enlarging significantly the gap between section 174 eligibility and the scope of section 41, the "appropriate design" language likely will be interpreted broadly to exclude many product and process development research activities. This new "exclusion" constitutes a clear departure from the statutory and interpretative definition of research and experimental expenditures. Development of new or improved products or processes generally requires both "basic research," such as that resulting in patents, and also "applied research," such as the development of design specifications, integration of components, and testing of new products/processes to ensure functionality, safety, reliability, durability, performance, and various quality standards. "Basic research," eligible for a separate basic research credit under section 41(e), is defined as "any original investigation for the advancement of scientific knowledge not having a specific commercial objective" (i.e., not specifically tied to a product). The legislative history describes the intent underlying the basic research credit:

   By contrast to other types of research or product development, where
   expected commercial returns attract private investment, basic research
   typically does not produce sufficiently immediate commercial applications
   to make investment in such research self-supporting. Because basic research
   typically involves greater risks of not achieving a commercially viable
   result, larger-term projects, and larger capital costs than ordinary
   product development, the Federal Government traditionally has played a lead
   role in funding basic research, principally through grants to universities
   and other nonprofit scientific research organizations.(62)

Both "basic research" and "applied research" are essential to the development of a new or improved business component and are eligible for research credits. Both types of research are undertaken for the purpose of discovering information that is "technological in nature" and constitute elements of a process of experimentation. These are precisely the types of activities that Congress intends to foster in its continual reenactment of section 41.

Treasury recognized these fundamental parallels regarding "basic research" and "applied research" when it finalized the section 174 regulations. The term "appropriate design" was discussed in the preamble to the final section 174 regulations, as follows:

   Several commentators expressed concern that the uncertainty test provided
   in the proposed amendments could be construed to limit unduly the
   definition of research or experimental expenditures. In particular, the
   commentators argued that the test could exclude expenditures incurred to
   determine the appropriate design of a product if the taxpayer knows at the
   outset that the procedure will be successful. In such a case, the taxpayer
   is not uncertain as to either its capability to develop the product or the
   method by which it will develop the product.

The Treasury Department and the IRS agree that a taxpayer's knowledge that a product development project will be successful does not preclude the process of determining the appropriate design of the product from qualifying as research. The Treasury Department and the IRS also agree that the language of the uncertainty test provided in the proposed amendments could be construed to reach a result contrary to that intended. The final amendments thus clarify the uncertainty test. Under the final amendments, the requisite uncertainty exists if the information available to the taxpayer does not establish either (i) the capability or method for developing or improving the product, or (ii) the appropriate design of the product.(63)

(iii) Application of the January 2001 Regulations. The "appropriate design" language could be read broadly by IRS agents to exclude applied research activities that are not themselves aimed at new technological breakthroughs. Such a reading would clearly contravene statutory law and its legislative history. An example in the 1986 legislative history provides that "engineers who design a new computer system, or who design improved or new integrated circuits for use in computer or other electronic products, are engaged in qualified research because the design of those items is uncertain at the outset and can only be determined through a process of experimentation relating to specific design hypotheses and decisions."(64) (Emphasis added.)

Many taxpayers that benefit from the above example are rightly alarmed by the introduction of the appropriate design "exclusion." Owing to the experience of their skilled engineers, a taxpayer may have little doubt that it has the capability of incorporating a new component into an existing product (e.g., a new printed circuit board into an existing computer). The taxpayer also will likely know how to incorporate the new circuit board into the existing computer (e.g., by reconfiguring the location of other printed circuit boards, modifying the racks holding the circuit boards, changing the path of internal wiring, altering capacitors and resistors, etc.). What the taxpayer does not know is what corrective actions are necessary to resolve the engineering problems discovered via testing (e.g., over-heating, power surges, electromagnetic incompatibility, etc.). If these applied research activities involve appropriate design uncertainty, and thus are ineligible for the credit, the result is absurd.

(iv) What Is the Appropriate Level of Uncertainty? There should be no difference in the level of uncertainty attributable to a "process of experimentation" for section 41 and "research and development in the experimental or laboratory sense" for section 174. For purposes of section 41(d), a "process of experimentation" should be defined as "a process to evaluate more than one alternative designed to achieve a result where the capability or method for achieving that result, or the appropriate design of a product, is uncertain at the outset." In other words, the researcher must evaluate more than one alternative addressing functional (rather than stylistic) aspects of a product where the means of achieving the result "is uncertain at the start." Such a level of uncertainty is consistent with the statute and legislative history.

(b) Four-Step Process of Experimentation. Consistent with legislative history, the January 2001 regulations outline four steps that "may involve" a process of experimentation:

* Developing one or more hypotheses to achieve the intended result.

* Designing an experiment (that, where appropriate to the particular field of research, is intended to be replicable with an established experimental control) to test and analyze those hypotheses (through, for example, modeling, simulation, or a systematic trial and error methodology).

* Conducting the experiment.

* Refining or discarding the hypotheses as part of a sequential design process to develop or improve the business component.

The January 2001 regulations provide that a taxpayer conducting a process of experimentation may, but is not required to, engage in the four-step process specified in the January 2001 regulations. Thus, a failure to conform precisely to the four-step process is not fatal to claiming the credit. If a taxpayer demonstrates that it follows these four steps, however, the taxpayer should enjoy the benefit of a safe harbor regarding the process-of-experimentation test. Significantly, there is also no requirement that a taxpayer record the results of its experiments. This requirement that was in the proposed regulations came under a firestorm of criticism from taxpayers and Congress, and the January 2001 regulations wisely dropped the requirement.

(c) "Shrinking-Back" Concept. Under Prop. Reg. [sections] 1.41-4(b), if the requirements of section 41(d) are not met for an entire product, then the credit may be available with respect to the next most significant subset of elements of that product. This shrinking-back continues until either a subset of elements of the product that satisfies the requirements is reached, or the most basic element of the product is reached and such element fails to satisfy the test.

The January 2001 regulations clarify that this shrinking-back rule applies only if the taxpayer incurs some research expenses with respect to the overall business component that would constitute qualified research expenses with respect to that business component but for the fact that less than substantially all of the research activities with respect to that component constitute elements of a process of experimentation that relates to a new or improved function, performance, reliability, or quality. In cases where the substantially all test is satisfied with respect to the overall business component, those research expenses with respect to the overall business component that are qualified research expenses are credit eligible, and there is no need for a taxpayer to shrink back to apply the tests with respect to subsets of elements of the business component.

The January 2001 regulations also clarify that, if the original product is not eligible for the credit, the application of the shrinking-back rule may result in credit eligibility for multiple business components that are subsets of the original product. The regulations clarify that the shrinking-back rule may not itself be applied as a reason to exclude research activities from credit eligibility.

(d) Integration Risk. Significantly, the shrinking-back example in Treas. Reg. [sections] 1.41-4(b)(3)(i) recognizes a critical concept: integration risk. The example recognizes that "the method of developing the improved cooling mechanism and of incorporating the improved mechanism into the widget would not have been known to skilled professionals had they conducted a reasonable investigation of the existing level of information in the particular field of science or engineering." (Emphasis added.) Other integration risks that might be associated with the cooling mechanism in the example may include concerns over the durability or functionality (e.g., insulation) of the widget's casing material when housing the new cooling mechanism and concerns that the electrical properties of the widget will change as a result of the improved cooling mechanism. Moreover, further testing of the redesigned widget will probably be necessary to meet government safety regulations.

B. Additional "Three-Part Test" for Internal-use Software

1. High-Threshold-of-Innovation Test

A number of changes have been made to the 1997 proposed internal-use software regulations. Under the proposed regulations, research with respect to software developed primarily for a taxpayer's internal use is "qualified research" only if it satisfies both the general requirements for credit eligibility under section 41 and three additional conditions for eligibility. Except for certain software developed for use in conducting "qualified research" or for use in a production process, and for certain software created as part of a package of hardware and software developed concurrently, the additional condition for eligibility is a requirement that the taxpayer satisfy a high-threshold-of-innovation test, requiring that the internal-use software be innovative, that its development involve significant economic risk and that it not be commercially available.

The January 2001 regulations clarify how the high-threshold-of-innovation test supplements the discovery requirement. Specifically, the January 2001 regulations provide that several aspects of the three-part test (e.g., the determination of whether the software is intended to result in an improvement that is substantial and economically significant, and the extent of uncertainty and technical risk) also must be applied with respect to the common knowledge of skilled professionals. In essence, the common knowledge of skilled professionals rather than the knowledge base of the taxpayer's employees is treated as the baseline with respect to which the intended software must satisfy the innovative prong and other prongs of the three-part test. Stated differently, under the January 2001 regulations, research with respect to internal-use software is credit eligible only if it is intended to exceed, expand, or refine the common knowledge of skilled professionals to a degree that is substantial and economically significant.

2.Computer vs. Noncomputer Services

(a) New Regulatory Exception. The January 2001 regulations clarify that the determination whether software is internal-use software depends on the nature of the service provided by the taxpayer. Software that is intended to be used to provide noncomputer services to customers is internal-use software, whereas software that is to be used to provide computer services is not developed primarily for internal use. Computer services are services offered by a taxpayer to customers who do business with the taxpayer primarily for the use of the taxpayer's computer or software technology. Noncomputer services are services offered by a taxpayer to customers who do business with the taxpayer primarily to obtain a service other than a computer service, even if such other service is enabled, supported, or facilitated by computer or software technology.

The regulations claim that the conclusion that software used to provide noncomputer services is internal-use software is consistent with the legislative history to the 1986 Act, which defined internal-use software as software used in general administrative functions and software used in providing noncomputer services (such as accounting, consulting, or banking services).(65)

The January 2001 regulations contain a new exception under which a taxpayer is not required to establish that internal-use software used to provide noncomputer services containing features or improvements that are not yet offered by a taxpayer's competitors satisfies the three-part test. Software that is intended to be used to provide noncomputer services is described within this exception if:

* the software is designed to provide customers a new feature with respect to a noncomputer service;

* the taxpayer reasonably anticipated that customers would choose to obtain the noncomputer service from the taxpayer (rather than from the taxpayer's competitors) because of those features of the service that will be provided by the software; and

* those features are not available (at the time the research is undertaken) from any of the taxpayer's competitors.

(b) Why the Distinction Between Computer and Noncomputer Services? One problem with the "computer vs. noncomputer services" approach of the regulations is that it fails to acknowledge that services offered by many companies today are so extensively integrated with computer systems and processes that they have become indistinguishable. Just as power steering on automobiles has become one integrated feature of the automobile that customers cannot now do without, computer systems and processes have similarly become integrated features of services offered which customers cannot do without. Without question, the enhancement of a company's service offering with computer applications will be factored into a customer's decision whether to do business with a particular company. Indeed, in some industries (such as the financial industry, the securities brokerage industry, and the air transportation industry), companies are selected not only for the services that they offer, but more important, for the convenience, efficiency, and value added by the company's on-line or other computer-based capabilities (e.g., bill paying capabilities, after-market trading capabilities, air transportation package tracking capabilities, etc.).

In such instances, it is at best difficult, and often simply impossible, to determine whether the customer's primary purpose is related to a noncomputer service "enabled, supported or facilitated by computer or software technology" or, alternatively, to a computer service relying on "computer or software technology." To be sure, the noncomputer service is enhanced or improved by the computer or software technology, but these factors do not fit neatly into the two tests set out in the regulations. The integration of technology and computer services by traditionally "noncomputer" companies will blur the lines of what a customer's primary use is for selecting one company over another.

Although the exception for "bells and whistles" software may be an attempt to strike a balance, it is arguably just as onerous as the high-threshold-of-innovation test, and in any event, will be difficult or impossible to apply in practice. For instance, how exactly does one go about ascertaining whether the new features of the software were available from competitors at the time the research is undertaken? Suppose one company offers software supporting its noncomputer service that is much the same as one of its competitors except that it is simpler, involving one less key stroke to execute. Would this qualify under the "bells and whistles" exception? The reliance by the regulations on these types of broad, misguided concepts is guaranteed to increase taxpayer-IRS disputes in this area and, consequently, further increase the business community's uncertainty over what can qualify for the section 41 credit.

Other questions and impracticalities arise with respect to the "bells and whistles" exception. For example., how does a taxpayer "reasonably anticipate" that customers would choose its noncomputer services because of the new software features? The regulations appear to be inconsistent with their own preamble on this account, in that the preamble uses language indicating the taxpayer must prove a "demonstrable competitive advantage," while the regulations make no mention of such. Indeed, computer systems, processes, software, etc. are often developed to eliminate or mitigate a competitor's advantage, but not necessarily to provide a company with a competitive advantage per se. The new feature, combined with all other features of the company's service, may tip customers to select the company's service, that clearly ultimately results in a competitive advantage, but is that competitive advantage solely attributable to the new feature? And in a service product involving numerous features, with thousands, hundreds of thousands, or millions of supporting computer software code, much of which is changing constantly, how can it be determined which change made the difference that resulted in the competitive advantage? The answer is it cannot be, or if it could be, a determined IRS agent could stonewall the allowance of the tax credit simply by demanding never-ending proof from the taxpayer that a certain change or set of changes to certain lines of code actually generated a "demonstrable competitive advantage."

The January 2001 regulations fail to recognize that the banking industry, the telecommunications industry, and other service-providing industries have radically changed since 1986. The public has dramatically shifted away from walking into a bank and transacting business with tellers. Now, almost all banking is done electronically through direct deposits, electronic transfers, and automated teller machines. Thus, banks increasingly combine hardware and software to deliver services to customers. This new banking technology fundamentally relies on principles of engineering and computer science.

The January 2001 regulations also do not fully acknowledge this mandate from the 1999 legislative history:

   The conferees further note the rapid pace of technological advance,
   especially in service-related industries, and urge the Secretary to
   consider carefully the comments he has and may receive in promulgating
   regulations in connection with what constitutes "internal-use" with regard
   to software expenditures. The conferees also observe that software
   research, that otherwise satisfies the requirements of section 41, which is
   undertaken to support the provision of a service, should not be deemed
   "internal-use" solely because the business component involves the provision
   of a service.(66)

C. New Recordkeeping "Exclusion"

The new regulations at first blush appear to temper the recordkeeping burden required by the proposed regulations concerning the process of experimentation. Unfortunately, they impose a possibly even more onerous contemporaneous recordkeeping requirement. The January 2001 regulations provide that taxpayers must prepare and maintain written documentation before or during the early stages of the research project that describes the principal questions to be answered and that describes the information the taxpayer seeks to obtain exceeds, expands, or refines the common knowledge of skilled professionals in the relevant field of science or engineering. Treas. Reg. [sections] 1.41-4(d), which introduces a new regulatory "exclusion" from credit eligibility, provides, as follows:

(d) Documentation. No credit shall be allowed under section 41 with regard to an expenditure relating to a research project unless the taxpayer --

(1) Prepares documentation before or during the early stages of the research project, that describes the principal questions to be answered and the information the taxpayer seeks to obtain to satisfy the requirements of paragraph (a)(3) of this section, and retains that documentation on paper or electronically in the manner prescribed in applicable regulations, revenue rulings, revenue procedures, or other appropriate guidance until such time as taxes may no longer be assessed (except under section 6501(c)(1), (2), or (3)) for any year in which the taxpayer claims to have qualified research expenditures in connection with the research project; and

(2) Satisfies section 6001 and the regulations thereunder.

The recordkeeping "exclusion" contained in Treas. Reg. [sections] 1.41-4(d) should be immediately withdrawn for the following reasons:

1. The new rule is phrased "No credit shall be allowed under section 41...." This is the functional equivalent of a new "exclusion" for lack of recordkeeping. There is no authority for the new recordkeeping "exclusion."

2. The new documentation requirement was not contained in the proposed regulations and this is wholly improper. Taxpayers were not provided notice, nor given the opportunity to comment regarding this significant new burden. Taxpayers have conducted and documented research projects for two decades without such a recordkeeping requirement and there has been no compelling reason put forth for forcing the new requirements on taxpayers now. The recordkeeping requirement threatens to hobble researchers with costly and time-consuming paperwork.

3. The new recordkeeping requirement is inconsistent with the 1999 committee reports: "The conferees also are concerned about unnecessary and costly taxpayer recordkeeping burdens and reaffirm that eligibility for the credit is not intended to be contingent on meeting unreasonable recordkeeping requirements."(67)

4. The documentation "exclusion" could result in paradoxical results. If a taxpayer develops a product that results in a patent (which would be conclusively presumed to satisfy the discovery requirement) at the back end of a research project, that project may be denied the credit because the taxpayer did not complete the newly required documentation at the front end of the project.

5. The annual estimated burden per respondent varies from .5 hours to 2.5 hours, depending oil the circumstances, with an estimated average of 1.5 hours. This estimated average of 1.5 hours per taxpayer in the regulations is a disgraceful attempt to mask the true burden. The estimated total annual recordkeeping burden for Treas. Reg. [sections] 1.41-4(d) of 18,000 hours (for an estimated 12,000 recordkeepers) will probably be consumed by companies in the Fortune 10 alone.

Specific issues that Treasury will need to consider include:

* If a taxpayer uses the cost center method of accounting (recognized in the IRS's research credit audit plan as acceptable), the required project documentation cannot be prepared.

* If the required documentation is prepared at the overall business component level and the IRS applies the shrinking-back rule to a smaller business component, will the taxpayer be penalized for not keeping documentation at the smaller business component level?

CONCLUSION

The January 2001 regulations are inconsistent with congressional intent underlying section 41, and they present an excellent opportunity for the Bush Administration to confirm its commitment to providing American business with a meaningful incentive to conduct research activities in the United States. The Treasury Department can act to liberate taxpayers from highly complex and strained rules that threaten to undermine congressional intent.

Given the significance of the research tax credit, the January 2001 regulations should be withdrawn and reproposed at the earliest possible date. The section 174 regulations addressing deductions for research or experimental expenditures were promulgated in final form only after being proposed three separate times (during a period spanning 11 years). The open and frank exchange between Treasury and taxpayers that took place in respect of the section 174 regulations helped produce a set of rules that are at once administrable and consistent with congressional intent. The constructive dialogue that took place in finalizing the section 174 regulations provides an excellent model for development of the final section 41 regulations.

Notes

(1) This article is adapted from a paper presented at a symposium sponsored by the Tax Council Policy Institute on February 15-16, 2001, in Washington, D.C.

(2) H.R. Rep. No. 100-1104 (1988) (Conference Report on the Technical and Miscellaneous Revenue Act of 1988), at 88.

(3) Id. at 132.

(4) The regulations contain various other provisions regarding the general exclusions from the credit, the definition of gross receipts, the Alternative Incremental Research Credit (AIRC), etc., which are beyond the scope of this article.

(5) JOINT COMMITTEE ON TAXATION, 97TH CONG., 1ST SESS., GENERAL EXPLANATION OF THE ECONOMIC RECOVERY TAX ACT OF 1981, at 17-19 (Comm. Print 1981.) (hereinafter referred to as "ERTA General Explanation").

(6) Id. at 120.

(7) Id. at 19, 120.

(8) Id. at 121.

(9) The three exclusions enacted in ERTA were for (1) expenditures made for research conducted outside the United States, (2) research in the social sciences or humanities, and (3) research funded by another person.

(10) These versions, for example, would have limited the credit to costs of developing a business item or a "significant improvement" to an existing business item of the taxpayer. See H.R. Rep. No. 97-201, 97th Cong., 1st Sess. 113-115, 327-28 (1981).

(11) See ERTA General Explanation, at 123-24.

(12) For an analysis of the legislative history of section 174, as we]l as revenue rulings and case decisions interpreting that provision, see McConaghy & Ruge, Congressional Intent, Long-standing Authorities Support Broad Reading of Section 174, 58 TAX NOTES 639 (February 1, 1993) (hereinafter referred to as "McConaghy & Ruge"). The current regulations under section 174, adopted in 1994, define "research or experimental expenditures" as expenditures that "represent research and development costs in the experimental or laboratory sense."

(13) Treas. Reg. [sections] 1.174-2(a)(1)(1957). This sentence basically was restated as two sentences, with some variations in wording, in the 1994 amendments to the section 174 regulations. Treas. Reg. [sections] 1.174-2(a)(1) (1994).

(14) See Rev. Rul. 69-484, 1969-2 C.B. 38; Rev. Rul. 73-20, 1973-1 C.B. 133; Rev. Rul. 73-275, 1973-1 C.B. 134; Rev. Rul. 73-275, 1973-1 C.B. 134; Rev. Rul. 73-324, 1973-2 C.B. 72; Rev. Rul. 74-67, 1974-1 C.B. 63; Rev. Rul. 75-122, 1975-1 C.B. 87.

(15) See analysis in McConaghy & Ruge.

(16) Research and Experimentation Tax Credit, Hearings Before the Subcommittee on Oversight of the House Committee on Ways and Means, 98th Cong., 2d. Sess. (1984) (hereinafter referred to as "1984 Hearings").

(17) Id. at 8, 46.

(18) Id. at 9.

(19) Id. at 25, 26, 29 (statement of Assistant Treasury Secretary Ronald A. Pearlman).

(20) Id. at 29-30.

(21) JOINT COMMITTEE ON TAXATION, 100TH CONG., GENERAL EXPLANATION OF THE TAX REFORM ACT OF 1986, at 10 (Comm. Print 1987) (hereinafter referred to as "1986 Act General Explanation").

(22) See ERTA General Explanation, at 130.

(23) See 1984 Hearing (statement of Assistant Treasury Secretary Ronald A. Pearlman).

(24) Senate Finance Committee (98th Cong., 2d Sess.), Deficit Reduction Act of 1984: Explanation of Provisions Approved by the Committee on March 21, 1984, Vol. I at 906, Vol. II at 1159 (S. Print 98- 169) (hereinafter referred to as "DEFRA Explanation"). Provisions relating to the research credit were not included in the final version of DEFRA.

(25) See 1986 Act General Explanation, at 133.

(26) Id.

(27) Id. at 130, 132, 133.

(28) H.R. Rep. No. 99-841, 99th Cong., 2d Sess. II-71 to II-72 (1986).

(29) Id. at II-72 n.3.

(30) 1984 Hearings at 25, 26, 29 (statement of Assistant Treasury Secretary Ronald A. Perlman).

(31) See id. at 53.

(32) Id.

(33) See 1986 Act General Explanation, at 133-36.

(34) Id. at 133.

(35) JOINT COMMITTEE ON TAXATION, 105 CONG., 1ST SESS., GENERAL EXPLANATION OF TAX LEGISLATION ENACTED IN 1997, at 85 (Comm. Print 1997); JOINT COMMITTEE ON TAXATION, 104TH CONG., 2D. SESS., GENERAL EXPLANATION OF TAX LEGISLATION ENACTED IN THE 104TH CONGRESS 105 (Comm. Print 1996); House Committee on Ways and Means, 103rd Cong., 1st Sess., Fiscal Year 1994 Budget Reconciliation Recommendations of the Committee on Ways and Means 161 (Comm. Print 1993).

(36) JOINT COMMITTEE ON TAXATION, 105TH CONG., 2D SESS., GENERAL EXPLANATION OF TAX LEGISLATION ENACTED IN 1998, at 236 (Comm. Print 1998) (hereinafter referred to as "1998 General Explanation").

(37) Id. at 236; H.R. Rep. No. 105-825, 105th Cong., 2d Sess. 1548-49 (1998).

(38) H.R. Rep. No. 106-478, at 132.

(39) See 1986 Act General Explanation, at 133; H.R. Rep. No. 99-841, at II-71, II-72.

(40) H.R. Rep. No. 105-825, at 1548.

(41) Treas. Reg. [sections] 1.41-4(a)(3)(ii).

(42) See, e.g., the reference to the term "advance" in Treas. Reg. [sections] 1.41-4(a)(3)(i).

(43) Statement of Glen A. Kohl, U.S. Department of the Treasury Tax Legislative Counsel, Hearing before the Subcommittee on Select Revenue Measures of the Committee on Ways and Means, U.S. House of Representatives, 103rd Cong., 2d Sess. (October 6, 1994), Serial 103-99 (hereinafter referred to as "Statement of Glen A. Kohl").

(44) H.R. Rep. No. 99-841, at II-73.

(45) Congress, in defining "qualified research" in the 1986 Act, rejected a recommendation to limit credit eligibility to research designed to produce a "significant improvement" in a product (Research and Experimentation Tax Credit, Hearings Before the Subcommittee on Oversight of the House Committee on Ways and Means, 98th Cong., 2d. Sess., at 25, 26, 29 (1984) (statement of Assistant Treasury Secretary Ronald A. Pearlman)) and a 1984 Senate provision referring to developing a "technologically new or improved" product (DEFRA Explanation, at II-1159). The Finance Committee provisions relating to the research credit were not included in the final version of DEFRA or in the Tax Reform Act of 1986.

(46) See Statement of Glen A. Kohl.

(47) H.R. Rep. No. 105-825, at 1548-49.

(48) Treas. Reg. [sections] 1.174-2(a).

(49) Research and Experimentation Tax Credit, Hearings Before the Subcommittee on Oversight of the House Committee on Ways and Means, 98th Cong., 2d. Sess. (1984).

(50) Id. at 8, 46.

(51) 86 AFTR2d [paragraph] 2000-5194 (N.D. Okla., July 31, 2000).

(52) 110 T.C. 454 (1998).

(53) 982 F. Supp. 1279 (N.D. Ill. 1997), aff'd, 163 F.3d 440 (7th Cir. 1998).

(54) H.R. Rep. No. 99-841, at II-71, II-72.

(55) Id. at II-72.

(56) Others, such as the exclusion for internal-use software, are considerably more complex.

(57) Treas. Reg. [sections] 1.41-4(a)(3)(ii).

(58) H.R. Rep. No. 99-841, at II-72 n.3.

(59) See Treas. Reg. [sections] 1.41-4(a)(3)(ii) and Example 4 of Treas. Reg. [sections] 1.41-4(a)(8).

(60) See 1998 General Explanation, at 236; H.R. Rep. No. 105-825, at 1548-49.

(61) H.R. Rep. No. 105-825, at 1548-49.

(62) 1986 Act General Explanation, at 131.

(63) T.D. 8562 (59 Fed. Reg. 50159, October 3, 1994).

(64) H.R. Rep. No. 99-841.

(65) See H.R. Rep. No. 99-841, at II-73.

(66) H.R. Rep. No. 106-478, at 132.

(67) Id.

JAMES R. SHANAHAN, JR. is a partner in the Washington National Tax Services Office of PricewaterhouseCoopers LLP. He serves as the National Director of Tax Services for the firm's Technology Industry Group. He is also the tax technical review partner for high technology issues for the firm and is the firm's R&D issue specialist. Mr. Shanahan thanks James Carlisle (PwC-Washington National Tax Services), who extensively researched the 14-year history of the regulations, and Tyrone J. Montague (PwC-New York), who assisted in interpreting the regulations, for their help in writing this article. Mr. Montague was one of the IRS litigating attorneys in the case of Norwest Corp. v. Commissioner, 110 T.C. 454 (1998). Copyright [C] 2001, PricewaterhouseCoopers LLP.