Comments on proposed regulations on the definition of research and experimental expenditures.

• Author: Ezrati, Lester D.
• Position: Filed with IRS by Tax Executives Institute on September 21, 1989

Comments on Proposed Regulations on the Definition of Research and Experimental Expenditures

1. Introduction

   On May 16, 1989, the Internal Revenue Service issued proposed regulations under section 174 of the Internal Revenue Code, relating to the definition of the term "research and experimental expenditures" and the application of section 174 to computer software development costs. The regulations are intended to provide interpretative guidance not only for purposes of section 174 (which accords taxpayers a deduction for research and experimental expenditures) but also for purposes of section 41 (which provides a credit for increasing research activities).

   The proposed regulations (PS-002-89) were published in the Federal Register on May 17, 1989 (54 Fed. Reg. 21224), and in the June 26, 1989, issue of the Internal Revenue Bulletin, 1989-26 I.R.B. 30. (1*) They represent a revision of proposed regulations which were promulgated by the IRS on January 21, 1983 (48 Fed. Reg. 2790, 1983-1 C.B. 1003). In these comments, the proposed rules that were published earlier this year are referred to as "the proposed regulations," and the initial proposed regulations are referred to as "the initial 1983 regulattions." (2)

2. Placing the Proposed

   Regulations in Historical

   Perspective

   In considering the issues raised by the proposed regulations' definition of research and experimental (R&E) expenditures, the Internal Revenue Service and Department of the Treasury should bear in mind section 174's legislative history. Enacted as part of the Internal Revenue Code of 1954, the current provision generally permits a taxpayer to deduct research and experimental costs as incurred. Costs relating to depreciable assets, however, may not be currently deducted, but rather are subject to the normal depreciation rules.

   1. Pre-1954 Code Developments

      In contrast, the Internal Revenue Code of 1939 did not authorize any specific treatment for R&E expenditures. Thus, under generally prevailing principles, R&E costs were deductible to the extent they were ordinary and necessary business expenses, whereas expenses that were capital in nature were only recoverable through deductions for amortization, depreciation, depletion, and loss. Capitalized costs that could not be attributed to a specific product or process that had a determinable useful life were deductible only as losses upon proof of either total failure or abandonment of an experimental project. Where projects were not abandoned and a useful life could not be determined, taxpayers were unable to claim any type of deduction in respect of the R&E expenses. (3)

      Under the 1939 Code, the difficulty of differentiating capital expenditures from ordinary and necessary business expenses led to conflicts over the proper tax treatment of R&E expenditures. In practice, the Bureau of Internal Revenue generally allowed taxpayers to take current deductions for expenses incurred under a continuing research program, but was more likely to require capitalization in other circumstances. (4) Because the Bureau did not permit current deduction in all cases, the issue was the subject of considerable litigation, with the courts almost invariably holding that the costs should be capitalized. (5)

      Congress was aware of the uncertainty created by the inconsistent positions taken by the Bureau, taxpayers, and the courts. In 1952, then-Commissioner Dunlap appeared before the Joint Committee on Internal Revenue Taxation and formally announced that the Bureau's policy was to allow deduction of "research and development costs in the experimental or laboratory sense" by taxpayers who currently deducted those costs under an established method of accounting. The Commissioner cited several justifications in support of this position:

      * apportioning research and development costs in advance of specific projects is very difficult;

      * such costs "usually are a necessary part of most businesses";

      * most taxpayers consistently charge these costs to expense; and

      * over time, allowing expense deductions does not create a materially different tax result from requiring capitalization with later deductions for depreciation or abandonment losses. (6)

      Although heartened by this testimony, taxpayers pressed for legislation to resolve the inconsistency between administrative practice and court decisions. Specifically, Congress was urged to enact definitive rules that could be consistently and broadly applied to R&E expenditures, thereby obviating the need for a case-by-case evaluation of R&E costs. (7)

   2. Legislative History of 1954 Code

      For these reasons, section 174 was enacted as part of the Internal Revenue Code of 1954. The House and Senate reports on the 1954 Code recognized the confusion in the existing law and stated that the purpose of section 174 was "[to] eliminate uncertainty and to encourage taxpayers to carry on research and experimentation." (8) Thus, the impetus for section 174 was a desire -- of both taxpayers and the government -- for a statutory resolution of the uncertainty over the proper treatment of research and experimental expenses.

      From an administrative standpoint, allowing current deductions for research and experimental costs makes consummate sense. The speculative nature of future benefits from R&E expenditures, as well as the frequent "cross-fertilization" of research results, makes a meaningful allocation of costs to specific assets virtually impossible. The problems of allocating costs, establishing useful lives, and justifying abandonment losses would pose serious administrative challenges -- for both taxpayers and the IRS -- in the absence of a statutory rule. If section 174 were narrowed through the regulatory process, these problems would again come to the fore. (9)

3. Specific Comments

   1. Time-Line Approach

      The proposed regulations adopt a "time-line" basis for distinguishing R&E activities from manufacturing, marketing, and other activities. Specifically, Prop. Reg. 1.174-2(a)(1) states:

      Expenditures incurred after the point that the product or property (or component of the product or property) meets its basic design spedifications related to function and performance level generally will not qualify as research and experimental expenditures under section 174 unless the expenditures relate to modifications to the basic design specifications for the purpose of curing significant defects in design, obtaining significant cost reductions or achieving significantly enhanced function or performance level.

      Thus, under this approach the character of a particular cost as a qualified R&E expenditure would seemingly depend on when the activity giving rise to the cost is undertaken. This "time line" differs from the functional approach taken in the current regulations to distinguish research from other activities. The functional approach looks to the nature of the activity conducted to determine whether that activity is research as opposed to something else.

      The preamble to the proposed regulations repeats, but does not elaborate on, the time line approach. 1989-26 I.R.B. at 31. The preamble notes, however, that the initial 1983 regulations would have excluded "the routine or periodic alteration or improvement of existing products, commercial existing products, commercial production lines, or other ongoing operations." The preamble explains that the reference to "routine" and "periodic" improvements had been deleted in response to comments filed on the initial 1983 regulations. Those comments -- including a submission filed by TEI -- argued that the initial 1983 regulations would have unreasonably restricted the definition of R&E expenditures by failing to recognize that R&E activities may in many instances be part of an evolutionary process involving a series of minor improvements. 1989-26 I.R.B. at 31.

      TEI submits that the time-line approach to R&E is flawed. Section 174 does not specify that R&E expenditures are deductible only if incurred at a certain time. To truly determine whether an expenditure is for R&E, the particular activity to which the expenditure relates must be examined -- not the time it is undertaken. Under the time-line approach, R&E costs incurred after the prescribed point will be excluded from section 174. This is inconsistent with the legislative history of section 174. The regulations should explicitly recognize that research is research, without regard to when it occurs.

      Thus, TEI recommends a return to the functional approach to distinguish R&E from other activities. Activities related to development, improvement, or testing of the functional design or specifications of the product or process are R&E. Such activities may occur before or after the point at which one could consider the basic design specifications to be met. In fact, they may occur, and often do occur, after the taxpayer is manufacturing the product or is otherwise utilizing the process. The functional approach is not only entirely consistent with the legislative history for section 174, but conforms to the manner in which taxpayers account for R&E for financial accounting purposes.

   2. Basic Design Specifications

      In general, the proposed regulations look to the point at which the "basic [functional or performance] design specifications" have been met as a guide for determining when the research effort has ended. The presence of the nebulous term "basic" in the proposed regulations is of great concern to TEI, for it has the ring of false clarity and invites debate, audit disputes, and (ultimately) litigation over the definition of R&E.

      TEI believes that all research activities up until the time the product is ready for commercial sale should qualify as R&E for purposes of section 174. Thus, assuming the time-line approach is retained in the final regulations, a product's "basic design specifications" should not be deemed to be met until that point. In the case of a process, "basic design specifications" should...