Yellow Freight Systems, Inc. v. United States, 24 Cl. Ct. 804 (1991)



This tax case comes before the court on plaintiff’s motion for partial summary judgment and on defendant’s cross-motion for partial summary judgment. Plaintiff, Yellow Freight System, Inc. of Delaware and Subsidiaries (“YFI”), is a Delaware corporation engaged in the business of transporting property by motor vehicle. In 1983 and 1984, plaintiff paid wages to its employees for work performed to develop or modify computer software programs that would assist plaintiff in its business dealings. Believing that these wage expenditures qualified for a tax credit as research expenses under 26 U.S.C. § 44F (“§ 44F”), plaintiff filed a claim for tax refunds for both 1983 and 1984. The Internal Revenue Service (“IRS”) denied these claims.

Before this court, plaintiff claims that the wages paid on nine particular data services requests, which it alleges are representative of all such requests, were improperly disallowed by the IRS. Plaintiff claims that the work done was experimental and as such qualifies as a § 44F research credit. Defendant claims that plaintiff was properly denied a refund because the computer software work completed did not involve research, and thus § 44F does not apply.

This court finds that it is unable to resolve this case on either plaintiff’s motion for partial summary judgment or defendant’s cross-motion for partial summary judgment because genuine issues of material fact remain in dispute.


Plaintiff is a motor carrier engaged in the transportation of general commodity freight. In order to remain competitive in the face of a growing emphasis by manufacturers on “just-in-time” inventory control practices, plaintiff developed certain software during the 1983–1984 time period. Plaintiff incurred expenses for the development of computer software by its Systems & Programming Department.

Plaintiff paid wages to its employees to formulate or modify various computer software programs. Plaintiff regarded the wages it paid for developing the computer software as experimental expenses deductible under § 174 of the Internal Revenue Code of 1954 and deducted them on its federal income tax returns for 1983 and 1984. On May 31, 1988, plaintiff filed for a refund on its 1983 and 1984 tax payments with the Internal Revenue Service claiming it deserved § 44F research credits. The IRS denied those claims.

The computer software resulting from the work in question consists of a very large number of commands directing the computer’s central processing unit to read, analyze, and sort pieces of information and then compile it into a desired format. The commands, which were written in a computer language, require the computer to make a long series of choices. If a single choice is omitted from the program, the program will fail. Therefore, extensive testing, correcting, and retesting were required to achieve a working end product. Yellow Freight submitted evidence pertaining to nine distinct computer software projects. Many of these fall into the category of Management Information Systems (“MIS”) that employ a database and allow examination of the data, report generation, data entry and updating. An MIS is a computer-based system that combines database, decision support, report generation, and telecommunications to support management analysis and decision making. The database system component supports the organization, storage, and retrieval of information. The report generation component presents the results of analysis. The telecommunications component allows for on-line data input and output. On-line implies that the data is input where generated and output where required. Thus, the data may be input from stations at break terminals and the report output returned to that same location. Plaintiff alleges that the work performed on the computer software programs in question was experimental under § 174 and thus became a qualified research expense for purposes of the § 44F credit. The work was experimental, plaintiff alleges, because it involved the hypothesis/trial-and-error process.

Defendant contends that the projects at issue were not research from a scientific perspective and do not constitute “qualified research expenses” eligible for the § 44F credit. Defendant relies on a report prepared by its proposed expert witness, Dr. John H. Carson, showing some contemporary applications which are similar to those utilized by Yellow Freight. Defendant postulates that in the development of computer software, as in any field of endeavor, there is a line of demarcation, a way to distinguish research from both engineering and constructive non-research activities. Defendant’s expert described the projects as “typical of the 1983–84 time frame” and no more than “what should be expected of a company in Yellow Freight’s field.” Defendant contends that an activity involving experimentation cannot be equated with an activity involving research. Defendant applies a “novelty” test arguing that if there is nothing innovative, risky or challenging about a system, it should not receive a credit. Applying its test to the nine separate projects presented in the exhibits, defendant finds no evidence of research qualifying for the credit.


Pursuant to the order entered on October 19, 1990, the sole issue to be decided at this time is whether the work performed on software programs was qualified research such that the wages paid for the work qualify for the tax credit claimed by YFI.

In order to determine whether the wages paid for the software which was developed or modified qualify for treatment under the research credit provision, several statutes are pertinent. Our starting point is § 44F, entitled “Credit for increasing research activities,” which reads as follows:

(a) General rule

There shall be allowed as a credit against the tax imposed by this chapter for the taxable year an amount equal to 25 percent of the excess (if any) of—

(1) the qualified research expenses for the taxable year, over

(2) the base period research expenses.

(b) Qualified research expenses

For purposes of this section—

(1) Qualified research expenses

The term “qualified research expenses” means the sum of the following amounts which are paid or incurred by the taxpayer during the taxable year in carrying on any trade or business of the taxpayer—

(A) in-house research expenses, and

(B) contract research expenses.

26 U.S.C. § 44F (1982). 1 Plaintiff claims it is due a tax refund for wages paid to its employees to develop computer software programs in-house. For the refund to be due, the expenditures must be deemed “in-house research expenses.” According to the statute, in-house research expenses include wages paid to an employee for qualifying services. Thus, the computer software work completed by plaintiff’s employees must fall into the category of “qualified services.” Qualified services include engaging in qualified research. § 44F(b)(2)(B)(i). Qualified research is described in § 44F(d): “For purposes of this section the term ‘qualified research’ has the same meaning as the term research or experimental has under section 174 . . .”

1 Section 44F was added to the Code on August 13, 1981 by Pub.L. 97–34, title II, § 221(a), 95 Stat. 241.

Section 174 entitled “Research and Experimental Expenses” states:

(a) Treatment as Expenses

(1) In general.—A taxpayer may treat research or experimental expenditures which are paid or incurred by him during the taxable year in connection with his trade or business as expenses which are not chargeable to capital account. The expenditures so treated shall be allowed as a deduction.

Accordingly, qualified research as used in § 44F(d) encompasses both research and experimental expenditures. Plaintiff will have engaged in qualified research under § 44F(d) if, in carrying on its trade or business, it incurred research or experimental expenditures during 1983 and 1984, the years for which it disputes the IRS’s disallowance of the credit.

Defendant concedes that there is no bright-line test for determining whether software development costs are research or experimental. At the same time, defendant attempts to formulate just such a test by citing to its expert’s scientific idea about what constitutes research. The scientific definition of research which defendant urges us to accept is not applicable in this case. Words in a statute are to be given their plain and ordinary meaning. United States v. Moody , 923 F.2d 341, 347 (5th Cir.1991) (not giving the plain wording and meaning [of the section] the strained, hypertechnical reading urged by [defendant]); United States v. Chen , 913 F.2d 183, 189 (5th Cir.1990); Matthews v. Commissioner of Internal Revenue , 907 F.2d 1173, 1175 (D.C.Cir.1990) (absent an indication that Congress intended a novel meaning, words of a statute are presumed to have their ordinary and usual meaning.)

As previously stated, qualified research as used in § 44F has the same meaning as research or experimental in § 174. Since § 174 does not define research or experimental, this court looks to a dictionary to determine the plain and ordinary meaning of the terms. Research is defined as the “critical and exhaustive investigation or experimentation having for its aim the discovery of new facts and their correct interpretation, the revision of accepted conclusions, theories or laws in the light of newly discovered facts, or the practical applications of such new or revised conclusions, theories or laws.” Webster’s Third New International Dictionary (1976). The definition of experimental is considered with that of research to determine what qualifies for the § 44F credit. 2

2 Experiment is defined as “1 a: TEST, TRIAL b: a tentative procedure or policy c: an operation carried out under controlled conditions in order to discover an unknown effect or law, to test or establish a hypothesis, or to illustrate a known law 2 obs: EXPERIENCE 3 the process of testing: EXPERIMENTATION.”

Experimental is defined as follows: “1 of, relating to, or based on experience: EMPIRICAL 2 founded on or derived from experiment 3 a: serving the ends of or used as a means of experimentation b: relating to or having the characteristics of experiment: TENTATIVE.” Webster’s New Collegiate Dictionary (1981).

The legislative objective of Congress in enacting section 174 may prove helpful in determining what Congress intended the terms research and experimental to encompass. Congress sought to provide an economic motivation for businesses to develop new products and new inventions. Snow v. Commissioner of Internal Revenue , 416 U.S. 500, 94 S.Ct. 1876, 40 L.Ed.2d 336 (1974) (overturning decision disallowing petitioner his distributive share of the net operating loss of a partnership engaged in research).

What Congress intended research expenditures to encompass may be elucidated by examining what was deemed non-research expenditures.

The provision does not allow the new credit for such expenditures as the costs of construction of copies of prototypes after construction and testing of the original model(s) have been completed; of pre-production planning and trial production runs; of engineering follow-through or trouble-shooting during production; or of adaptation of an existing capability to a particular requirement or customer’s need as part of a continuing commercial activity. For example, the costs of adapting existing computer software programs to specific customer needs or uses, as well as the modifications of previously developed programs, are not eligible for the credit.

Joint Committee, 97th Cong., 1st Sess., General Explanation of the Economic Recovery Act of 1981, at 125 (“Blue Book”). If any of plaintiff’s expenditures fall into the items specifically excluded above, the research credit will not be available for those expenditures.

To clarify the intended scope of the terms research or experimental, a look at the legislative objective of Congress in enacting § 44F is appropriate. The report of the Committee on Finance of the United States Senate sets out its reasons for supporting § 44F:

The committee believes that a substantial tax credit for incremental research and experimental wage expenditures will overcome the resistance of many businesses to bear the significant costs of staffing which must be incurred in initiating or expanding research programs. While such costs bear characteristics of investment activity, the relationships between the investment and the subsequent earnings often are less directly identifiable, and many businesses are reluctant to allocate scarce investment funds for uncertain rewards.

Research and experimentation are basic activities that must precede (1) the development and application of new techniques and equipment to production and (2) the development and manufacture of new products. The committee believes that its multifaceted approach in the bill, designed to stimulate a higher rate of capital formation and to increase productivity, appropriately includes incentives for greater private activity in research.

In recent years, spending for these purposes has not been adequate . . . The committee believes that the decline in this country’s research and development activities has adversely affected economic growth, productivity gains, and our competitiveness in world markets.

S.Rep. No. 97–144, 97th Cong., 1st Sess., 76–77 (1981), U.S.Code Cong. & Admin.News 1981, pp. 105, 181, 182.

Regulations exist which further speak to the meaning of research. A Treasury regulation elucidates the meaning of research and experimental as used in § 174. This Treasury regulation, § 1.174-2(a)(1), states:

The term “research or experimental expenditures,” as used in section 174, means expenditures incurred in connection with the taxpayer’s trade or business which represent research and development costs in the experimental or laboratory sense. The term 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. The term does not include expenditures such as those for the ordinary testing or inspection of materials or products for quality control or those for efficiency surveys, management studies, consumer surveys, advertising, or promotions.

While computer software, as property, can obviously fall within the Treasury regulation’s definition of “research and experimental expenditures” if all the prerequisites set out in the regulation are met, a specific pronouncement on the costs of developing computer software is available. Revenue Procedure 69–21, 1969–2 Cum.Bull. 303, pronounces guidelines for the treatment on federal income tax returns of costs incurred to develop computer software:

For the purpose of this Revenue Procedure, “computer software” includes all programs or routines used to cause a computer to perform a desired task or set of tasks, and the documentation required to describe and maintain those programs. Computer programs of all classes, for example, operating systems, executive systems, monitors, compilers and translators, assembly routines, and utility programs as well as application programs are included.

* * * * * *

[a]ll the costs properly attributable to the development of software by the taxpayer are consistently treated as current expenses and deducted in full in accordance with rules similar to those applicable under section 174(a) of the Code . . . 3

3 Congress, for purposes of the § 44F credit, “intended that otherwise qualifying types of expenditures (for example, direct wage expenditures) which are part of the costs of otherwise qualifying research for the development of new or significantly improved computer software are to be eligible for the credit to the extent that such expenditures (1) are treated as similar to costs, incurred in product research or experimentation, which are deductible as research or experimental expenditures under section 174; (2) satisfy the requirements of new section 44F which apply to research expenditures, including the trade or business requirement; and (3) do not fall within any of the specific exclusions in new section 44F.” Blue Book at 124.

The Joint Committee explains that Congress enacted the research credit provision to encourage additional research and development believing that the nation’s edge in those areas had declined during the preceding years. Blue Book at 19. Plaintiff cites to the Blue Book for the proposition that Congress intended expenditures for development of computer software deductible under Revenue Proc. 69–21 to qualify for the research credit under § 44F. Blue Book at 124. Defendant cites to another passage, which states:

For this purpose, the cost of developing computer software means costs incurred in developing new or significantly improved programs or routines that cause computers to perform desired tasks (as distinguished from other software costs where the operational feasibility of the program or routine is not seriously in doubt.)

Id . Under Revenue Proc. 69–21 cited by plaintiff, “all the costs properly attributable to the development of software by the taxpayer” would be eligible for the research credit. Defendant defines the “costs properly attributable to the development of software” as those utilized in “developing new or significantly improved programs” only. Thus, in order for a research credit to be available, new or significantly improved programs must have been produced by plaintiff.

Since the wages paid in this case were direct wage expenditures, they are qualified expenditures. They will be allowed research credit status under § 44F if they were: (1) spent for qualified research; (2) properly attributable to the development of new or significantly improved software; (3) incurred by the taxpayer in carrying on his trade or business; and (4) not specifically excluded under § 44F.

It is uncontroverted that plaintiff incurred the expenses while conducting trade or business, and the costs are not specifically excluded under § 44F; therefore, the expenditures will qualify for the credit if the wages were paid for qualified research and if the software produced was either new or a significant improvement from that which existed previously.

To determine if summary judgment is appropriate for either of the parties, no genuine issues of material fact should exist concerning whether the wages were research related or whether the software programs were new or substantially improved.

Under Rule 56(c), summary judgment is proper “if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law.” In our view, the plain language of Rule 56(c) mandates the entry of summary judgment, after adequate time for discovery and upon motion, against a party who fails to make a showing sufficient to establish the existence of an element essential to that party’s case, and on which that party will bear the burden of proof at trial.

Celotex Corp. v. Catrett , 477 U.S. 317, 322, 106 S.Ct. 2548, 2552, 91 L.Ed.2d 265 (1986).

To determine whether there are material facts in dispute, this court will look separately at each of the nine data services systems which plaintiff submits as representative of all the software work for which it claimed research credits on its 1983 and 1984 tax returns.

The Profitability and Cost System (“ PACS ”)

PACS was designed to provide true actual costs; specific shipment costs; standard costs and variance analyses; management information to assist in making pricing decisions and to analyze profitability and performance efficiency; and to show the simulated effect of “what-if” type questions providing credible and sophisticated simulation relative to changes in operating conditions. Plaintiff purchased a nonexclusive license from A.T. Kearney, Inc., a management consultant group in Chicago, to use a PACS which Kearney owned. This PACS was to be modified by Kearney with the assistance of plaintiff’s Systems and Programming staff to suit plaintiff’s needs. Defendant contends that it is not clear that plaintiff worked on this program, or if it did, what percentage of the work was performed by plaintiff.

The Break Freight Assistance System

Through various formulae and algorithms, this system assists the freight terminals in selecting the ‘best’ trailer out of a large number of trailers for the terminal to unload. Plaintiff alleges that significant improvements were made to this program in the tax years 1983 and 1984 and that when the project began, it was unknown if the desired results could be achieved. Defendant’s expert, Dr. Carson explained that “the improved version involved gathering additional data, adding this data into the current system, generating screen and perhaps hard copy reports for this data.” He stated that this “requires changes to existing programs and the addition of new programs to the current system.” Carson claimed that the computer portion of the system “seem[ed] conventional without substantial risk” since he found no documentation in the exhibits noting any doubt about the potential effectiveness of the system.

Buick City Program

This program was developed to allow the computer at the Buick Motor Division of General Motors to obtain data regarding freight shipments to Buick City from plaintiff’s computer. Plaintiff’s computer was to notify the appropriate freight terminal that a shipment was ready to be picked up at the supplier and delivered to Buick City by a certain date and to provide the Buick City computer with information regarding the status of the shipment. Plaintiff alleges that this software was novel and “leading edge” in 1984. Defendant alleges that there is no evidence that work on this project actually started in 1984, the year for which the credit was filed. Defendant argues that even if work was being done in 1984, “[t]he novelty of this system . . . is doubtful” since another company commercially obtained a system at that time which seemed to accomplish much of what the Buick City project provided.

DuPont Electronic Freight Billing

This system was designed to enable plaintiff’s computer to bill DuPont’s computer for freight charges. Plaintiff alleges that this computer software, which allowed electronic data interchange between the computers of trading partners, was novel in 1984. Defendant alleges that it was not novel, and that the data for DuPont corresponded to the Transportation Data Coordinating Committee standard (“TDCC”) which already existed. The defendant describes the programming effort as probably consisting of changes and additions to an existing system with a new data format.

Trailer Dispatch Time to Make Service Program (“TTMS”)

Formulating TTMS involved designing and programming a system that assists personnel at freight terminals to decide which trailer for the terminal to dispatch. Plaintiff alleges that the system is sophisticated, uses various formulae and algorithms, and that it was unknown at the inception of the project whether the desired results could be achieved. Dr. Carson, defendant’s expert, states that integrating this facility into existing systems is a fairly complex task since in many cases, a single change would require a number of actions to different programs. Overall, he characterized it as a straightforward report generation system. Since a user ultimately makes the decisions when presented with data reduced by the system, he did not believe that any challenges involving operation research techniques were presented.

Customer Master System

This program was used to identify costs by customer and relate costs to prices charged that customer. Plaintiff alleges that the system uses a sophisticated algorithm to establish customer identification. Soundex 4 keys were used to help users locate customers. Defendant argues that the use of a customer master file had been standard practice for a long time so that nothing novel or research-based was involved in the development of its particular customer file.

4 The Soundex is an algorithm that transcribes names into standard codes and clusters names that sound alike into the same codes and allows one to locate a name based on its sound, rather than its exact spelling.

Computer-Assisted Cashiering System

This software was created to allow clerks at a central location to apply cash, on computer terminals, to open accounts receivables maintained on the computer. Plaintiff maintains that this system replaced an antiquated practice wherein clerks in freight terminals applied cash to open account receivables maintained in file ‘tubs.’ Because of this system, plaintiff believes its effort to stay competitive in the trucking industry was improved. Dr. Carson contradicts plaintiff and states that the cashiering system seems straightforward with “no significant algorithms, software configurations, data organizations or user dialogs that would indicate anything novel.” He states “the ‘extraordinary measures’ cited by [plaintiff] for testing are a responsible and typical reaction to any system dealing with financial matters. ‘Extraordinary’ by no means implies research.”

City Licensing Inquiry Development

This request apparently involved the development of a program that would produce a report showing all location changes for city equipment. Defendant characterizes this as a “very small, simple effort” completed in just five days. Although conceding that the project did involve software development, defendant asserts that “there is nothing challenging or novel about this project.”

Competitive Rate/Quotation

This system allows the plaintiff to provide an automatic class rate quote and to compare its rates with major competitors. The system was necessitated by “an outgrowth of deregulation of the trucking industry.” Defendant argues that this data processing system was straightforward and did not entail any novel challenges.


To determine that summary judgment is appropriate, this court would have to conclude that no genuine issue exists as to any material fact. This court is unable to make that determination because the parties have offered conflicting evidence concerning the software programs at issue. The parties have not addressed the test this court set out above. Facts necessary to the determination of whether the nine software projects are new or significantly improved are contested. This court is unable to make a legal determination without a clear knowledge of the underlying material facts. Some of these facts are either contested or have not been submitted to the court and are therefore in dispute. A trial is necessary to resolve the factual matters in dispute. Accordingly, plaintiff’s motion for partial summary judgment is denied, and defendant’s cross-motion for partial summary judgment is denied.