CORPORATION NOT ENTITLED TO QUALIFIED RESEARCH CREDIT FOR SOFTWARE SYSTEM.

UNITED STATES DISTRICT COURT EASTERN DISTRICT OF WISCONSIN

DECISION AND ORDER

[1] Wicor, Inc., and its subsidiaries bring this action against the United States to receive a federal income tax refund under 26 U.S.C. sections 41 and 1341. The section 41 and section 1341 issues were bifurcated and, on September 30, 1999, this court granted the defendant's motion for partial summary judgment, wherein the court denied the plaintiff's request for an income tax credit under section 1341.

[2] On April 19, 1999, a trial to the court was conducted on the plaintiff's claim that it is entitled to a research credit under section 41. Following the trial, the parties submitted their respective proposed findings of fact, conclusions of law and supporting memoranda. The following constitutes this court's findings of fact and conclusions of law. The parties have consented to this court's complete jurisdiction. See 28 U.S.C. section 636(c)(1).

[3] WICOR and the Wisconsin Gas Company ("WGC") are accrual basis taxpayers and maintain their books and records on a calendar- year basis. WGC is a wholly-owned subsidiary of WICOR, and is hereinafter included with references to WICOR filed consolidated federal income tax returns on a calendar-year basis and timely filed income tax returns for 1990 and 1991. WICOR also timely filed for a tax refund; $332,119 for 1990 and $336,637 for 1991. Its requests for refunds were denied. The parties stipulated that WICOR complied with all jurisdictional prerequisites for bringing a lawsuit to dispute the government's denial of WICOR's request for a tax refund. For tax years 1990 and 1991, WICOR claims a credit for qualified research activities under Internal Revenue Code section 41(26 U.S.C. section 41) and for expenditures incurred to develop a computer-based information system.

I. FINDINGS OF FACT

[4] Numerous lay and expert witnesses testified during the course of this trial, and a substantial number of documents were received in evidence. Since credibility is not determinative of the issues in this case, a general overview of the chronology of Wicor's project will be presented instead of a detailed analysis of each witness' testimony. The issues in dispute can be resolved with a determination of whether the generally undisputed facts fit under the Internal Revenue Code's definition of what constitutes "qualified research." During the discussion of each of the issues, specific reference will be made to the more relevant testimony.

[5] WICOR, in conjunction with its contract consultants, developed a computerized integrated customer information system ("CIS"), which resulted in improved customer service and company performance in the areas of automated meter reading, computer and dispatch and online cash posting. This CIS was the end product of a process that began in 1984, when WICOR formed an internal customer project team to identify business and information needs, develop conceptual design, and review and evaluate cost/benefit feasibility and make recommendations. The initial CIS devised by the project team was to provide a greater level of service to customers while decreasing the amount of time and expense required to perform day-to- day customer service functions: specifically, the CIS was designed to improve service order processing, customer billing, cash processing and maintenance. During 1986, the team evaluated various CIS packages but were unable to locate a system that performed the functions required by WICOR. WICOR eventually selected a package by Computer Systems & Associates ("CS&A") entitled the Customer Information System/Data Base ("CIS/DB") a software package designed for the gas and utility industries.

[6] WICOR and CS&A entered into a service agreement, wherein CS&A agreed to provide systems consultants, system analysts and programmers to help manage, customize, modify and develop the CIS. In July 1989, the original CIS cost estimate of $5.4 million was increased to $10.9 million based on additional functional, technical and staffing needs.

[7] Eventually, WICOR concluded that CS&A could not meet its technological needs at an acceptable cost and concluded that in-house custom development was necessary. To facilitate such internal development, WICOR retained the Andersen Consulting Division of Arthur Andersen & Co. to review the CIS plan. Andersen had previously performed customer information system work with other utilities, including Madison Gas & Electric and Wisconsin Public Service. In its report dated August 18, 1989, James Drayer of Andersen Consulting summarized the findings and recommendations from its review of the CIS project. Drayer stated that the overall project scope "was well- defined and included the functions and features normally found in a CIS." Drayer noted that the online architecture appeared well- structured. However, Drayer identified four areas of concern: project approach, database, project planning and control, and functional. With regard to project approach, Drayer noted that the design for two of the largest and most critical areas (billing and conversion) had not been started. With regard to the data base, Drayer stated that the primary concern was the performance worthiness of the new system and that the existing prototype database would not perform satisfactorily and would have to be modified. As for project planning, Drayer said that the overall project plan did not establish a baseline against which project status could be tracked and that the project's link to company management was neither well-defined nor understood. (Drayer Letter to Paul Petry, Ex. 65).

[8] In December 1989, WICOR's contract with CS&A was terminated and CS&A phased out of the project, while Andersen Consulting was phased in. In January 1990, a consulting services agreement was entered between WICOR and Andersen Consulting, which documented Andersen's arrangement to assist WICOR with completing the CIS project.

[9] In May 1990, the IDMS database, which was acquired during the CS&A phase, was abandoned and WICOR purchased a DB2 database to use as the platform for its new CIS, Andersen commenced work on the CIS project in the summer of 1990. WICOR's agreement with Andersen estimated 19,000 WICOR labor days and 16,000. Andersen labor days. A 29-month project Schedule was proposed with a November 1, 1990, start date. The time line called for software development to be completed in early 1993 with the system released and available for production sometime between February and April 1993. The CIS, as envisioned in January 1990, included service order processing, meter reading, billing, billing adjustments, rate maintenance, cash processing, meter management, interface to a third-party credit and collection system and partial non gas-user billing. 2

[10] Anderson Consulting embarked on the project by establishing a structural format which included several components. Method/1 was a project management methodology, which was a guideline checklist developed by Andersen with respect to tasks, documents and decisions that needed to be made throughout the project. Manage/1 was a time-tracking tool, which tracked the actual experience against the estimates. Design/1 was an electronic filing repository system for creating, storing and manipulating all of the documentation produced as part of the system project. Install/1 was a tool that allowed a programmer to establish codes for use with different programs or applications, which were then used by other programmers.

[11] As time went on, the scope of the project expanded and in early 1991, WICOR decided to integrate three additional systems into the CIS project -- a computer aided dispatching system ("CAD"), a remote meter reading system ("RMR"), and on-line cash posting. Each used software provided by other vendors, although each required an interface to the new system. Before these systems were integrated into the project, the CAD system was not directly linked to the CIS. [C]ustomer orders were received and given to the dispatcher and relayed to service vehicles, whereupon the information was downloaded and batched processed overnight. With the new system, all orders were transmitted in real time to a mainframe and then transmitted via radio from the AS400 to field personnel, where information was conveyed via real time from the field personnel back to the CIS system. The key objectives of the CAD systems were real time updating, interfacing to the network existing at the time, use of smart terminals and interfacing with the CIS.

[12] With regard to the RMR, the primary objective was to get timely, accurate readings. Prior to integrating the RMR into the CIS, porta processors were used by meter readers, which required a WICOR employee to approach customer meters and either manually enter the meter reading or enter the information in a porta processor. Under the new system, a personal computer is installed in the meter reader's vehicle and, as the vehicle moves down the street, it sends out a signal, which activates a radio device installed on the meter and transmits meter reads to the service vehicle. With respect to cash posting, customer payments that were made in person were processed immediately and linked back to the other CIS subsystems. Previously, a single day's payments were taken during the day, batched together and processed during a nightly batch run.

[13] While developing the CIS project, the project team performed unit testing, string testing, integration testing and stress testing. Unit testing was followed by string testing, which was followed by integration testing. During the week of February 1993, the so-called "Big Bang" conversion occurred, whereby information from more than 500,000 customer accounts were converted from the old system to the new CIS. The new CIS resulted in a multiple-function integrated system that improved customer service and resulted in labor and cash flow savings by contributing to a reduction in staff, including reducing information technology positions and meter readers. For example, automated meter reading eliminated the 30 percent error rate incurred under the old system. The CIS project required 19,000 employee labor days and 16,000 consultant labor days. WICOR eventually spent more than $30 million on the project, and claimed a research tax credit under Internal Revenue Code Section 41 for costs incurred in the development.

II. ANALYSIS

[14] The taxpayer bears the burden to show that it is entitled to a tax credit. United Stationers, Inc. v. United States 163 F.3d 440, 443 (7th Cir.), cert. denied, 527 U.S. 1023 (1999). Section 41 provides a tax credit for "qualified research expenses." "Qualified research" is defined as research:

     (A) with respect to which expenditures may be treated as

[15] Section 41 establishes four tests that must be satisfied for the taxpayer to obtain a tax credit for qualified research. When section 41 is combined with a 1986 House of Representatives Conference Report, three additional tests arise that a taxpayer must satisfy. See, Northwest Corp., v Commissioner of Internal Revenue, 110 T.C. 454 (T.C. 1998). The parties agree that all seven tests apply in this case, but further agree that not all are at issue. The government does not dispute that the plaintiff satisfied the first test which is that the research must qualify as a business deduction under section 174; in addition, the government does not dispute that the information gained Was technological in nature, that application of the information was intended by WICOR to be useful in a new or improved business component, and that the CIS software was not commercially available. The government disputes whether the plaintiff's research was for the purpose of discovering information which is technological in nature, whether substantially all of the research constituted a process of experimentation, and whether the plaintiff satisfied the "innovativeness" or "significant economic risk" tests for internal computer software. These are the tests that remain at issue, and the court will discuss each one in turn.

1. DISCOVERY TEST

[16] In order to qualify for the research credit, research must be undertaken to discover information that is technological in nature. The House Conference Report accompanying the discovery test states:

     The determination of whether the research is undertaken for the

[17] The discovery test is intended to limit the form of discovery to the "process of experimentation." Id. "The fact that the information is new to the taxpayer, but not new to others, is not sufficient for such information to come within the meaning of discovery for purposes of this test." Id. "[D]iscovery demands something more than mere superficial newness; it connotes innovation in underlying principle." United Stationers, 163 F.3d at 444.

[18] With regard to computer science in particular, the House and Senate Conference Report accompanying the 1986 amendments to the Internal Revenue Code state that "' [r]esearch 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.'" H.R. Conf. Rep. No. 99-841, at II-71 n.3 (1986) (quoted in United Stationers, 163 F.3d at 444)). Congress intended that "qualifying research pass a high threshold of innovation and be of broad effect." See United Stationers, 163 F.3d at 444.

[19] The plaintiff cites expert testimony offered by Donald O'Neill, who stated that the CIS project "applies the principles of computer science and software engineering in the integration of accounts receivable, billing, credit, meters and orders along with the high value changes of On-Line Cash, Computer Aided Dispatch, Automatic Meter Reading, and Advanced Function Planning." O'Neill also stated that, in conducting the CIS project, "technological information was discovered in determining the best mix of algorithms and data and rules of construction needed to balance execution traces, path lengths, user response times, and integration effects." O'Neill concluded that the research needed to complete the CIS project "expanded the application of computer science principles within the gas utility customer information system application."

[20] In addition, Dr. Walter Scacchi, another of plaintiff's experts, concluded that the CIS team "was successful in its repeated efforts to discover information that was technological in nature through prototyping, reviews, inspections and testing. Specifically, Dr. Scacchi stated that integrating the automated meter reading system and the computer aided dispatch into the CIS, which made the CIS a distributed networking system, constituted a discovery in the field of software information systems engineering. Finally, Mr. Steven Dick testified that, based on his knowledge of software development techniques and the utility industry in 1990, the concepts WICOR explored in its CIS project exceeded that which was known in the utility field.

[21] The government concedes that the CIS project was "technological in nature;" however, the government maintains that the project was not research undertaken for the PURPOSE OF DISCOVERING technological information. In other words, the government submits that WICOR's purpose was not to discover information that was technological in nature, but to install a cost-effective and cost- reducing business information system that would automate in a single integrated information system a number of company information functions.

[22] In addition, the government contends that, in order to be eligible for a research credit, the plaintiff must show that it gained knowledge that exceeded what was already known in the computer science field and that the plaintiff's computer system was not sufficiently unprecedented in the utility industry. Moreover, even if the plaintiff's software was unprecedented in the utility industry, the government maintains that the relevant field is the computer science field, not the utility industry. The government cites the opinion offered by one of its experts, Dr. Jaideep Srivastava, who stated that WICOR's software development practices were "a fairly standard procedure that had been in existence at least a decade or more before this period." In addition, Dr. Srivastava testified that he was not aware of any new algorithms that were created through the CIS project.

[23] As noted above, one of the plaintiff's experts, Donald O'Neill, testified that the CIS project satisfied the discovery test because it expanded knowledge of computer science principles within the gas utility customer information system application. Mr. O'Neill's testimony is revealing for he does not assert that the CIS project expanded the range of knowledge in the computer science field itself, only in the utility industry's utilization of computer science. According to O'Neill, it is the essence of discovery which satisfies the test, even if the knowledge discovered is only new to the taxpayer.

[24] Contrary to O'Neill's interpretation, it has been held that the research tax credit is not intended for projects that only expand the universe of knowledge in a particular industry. As stated by the Norwest court, "[t]he legislative history of section 41 dictates that the knowledge gained from the research and experimentation must be that which exceeds what is known in the field in which the taxpayer is performing the research and experimentation -- in this case, the computer science field." Norwest, 110 T.C. at 493. Of note is that the court did not limit the relevant field to computer science as applied to a particular industry -- in Norwest, the banking industry; rather, the court mandated that the knowledge gained from the research and experimentation add to the existing base of knowledge in the computer science field as a whole. In this case, there is no evidence to show that the CIS project exceeded that which was already known in the computer science field. As indicated above, O'Neill testified only that the CIS project expanded the universe of computer science knowledge as applied to the utility industry. Mr. O'Neill's testimony, coupled with Dr. Srivastava's uncontradicted testimony that the CIS project did not contribute additional information to the field of computer science as a whole, that the project did not involve new software writing techniques and did not utilize a proven methodology with respect to its use of Method/1, leads the court to conclude that the CIS project did not constitute research undertaken to discover information that is technological in nature for purposes of the research credit.

2. PROCESS OF EXPERIMENTATION TEST

[25] Section 41(d)(1)(C) requires that qualified research involve "substantially all of the activities which constitute elements of a process of experimentation." The Conference Report explains the meaning of the process of experimentation test:

     The term process of experimentation means a process involving

[26] The plaintiff contends that it followed the engineering method of experimentation, whereby engineers develop or test a solution to a hypothesis based upon the results of the test and improve the solution until it requires no further improvements to engineering development. Dr. Walter Scacchi distinguished the engineering method from the scientific, empirical, and analytical methods of experimentation. The plaintiff notes that its software engineers developed testing solutions to a variety of hypotheses, which appeared as lessons learned that were collected and internally distributed, were transformed into case studies and presented to customers, management, and consultants. The government responds that the WICOR project did not involve developing a new computer system in which the design was uncertain at the outset.

[27] The plaintiff failed to present evidence that the means for achieving the final result of the WICOR project were uncertain at the outset. For example, Paul Kowalski, the WICOR credit team leader, testified that the WICOR project team followed a tried and true data processing approach -- the DB-2 system. Perhaps most telling is Kowalski's acknowledgment that the data processing system utilized by WICOR was not new, although it was new to WICOR (Kowalski Dep. at pp. 28-29). However, a method does not constitute a process of experimentation simply because the method is experimental to a specific taxpayer. Rather, the method must be experimental to the relevant field -- in this case, computer science. Kowalski's testimony belies the plaintiff's assertion that the WICOR project was experimental to the field of computer science.

[28] The court is persuaded by the testimony offered by the government's experts -- Dr. Srivastava and Thomas Niccum -- that the project did not involve identification of models or theories of computer science using a process of experimentation and that there was not a systematic research-oriented process by which data was collected for analysis of competing hypotheses. Although WICOR developed "lessons learned" reports, these reports were informal and only identified miscellaneous programming errors. The court concludes that the WICOR project did not constitute a process of experimentation.

3. INNOVATIVENESS TEST

[29] The third element disputed by the parties with respect to entitlement to the research credit is the innovativeness test; i.e., whether WICOR developed innovative software. Software is "innovative" if it "results in reduction in cost, or improvement in speed, that is substantial and economically significant." Prop. Treas. Reg. section 1.41-4(e)(5)(I), 62 Fed. Reg. 83 (1997). "[T]he extent of the improvements required by Congress with respect to internal use software is much greater than that required in other fields." Norwest, 110 T.C. at 499. Only a "high threshold of innovativeness" will satisfy this threshold. Id.

[30] The plaintiff contends that its project was "innovative" because it included a streamlined, powerful integrated system including computer aided dispatch, remote meter reading, and online cash posting that served as a model for other utility companies. The government responds that, although the plaintiff's software development activities involved a reduction in WICOR's cost and an improvement in efficiency, WICOR failed to prove that any such reduction or improvement substantially contributed to technological advances in computer science.

[31] Dr. Scacchi and Mr. O'Neill opined that the "innovative test" was satisfied because the streamed integrated system developed by WICOR -- such as the use of computer aided dispatch, remote meter reading, and online cash posting -- served as a model for what other utilities lacked but for what could ultimately be achieved in the industry. Mr. Dick testified that the CIS project was one of the few software packages of its time to address customer services such as cash management, which had been previously performed manually. In addition, Mr. Dick stated that the majority of meter reading was not automated until the CIS system was developed.

[32] Again, the plaintiff's case is hindered by the fact that its experts focused on the utility industry in a vacuum and did not expand their analysis to the computer field as a whole. which its necessary to meet the "high threshold of innovativeness" required for the research credit. Of relevance is Dr. Srivasta's uncontradicted testimony that the CIS project did not result in any new discoveries with respect to computer science principles. Intuitively, a computer software project cannot meet the "high threshold of innovation" without discovering some new principle of computer science. Also for consideration is the fact that WICOR did not seek a patent for its project. Although a research tax credit need not meet the same standard of innovativeness as is required for issuance of a patent, the fact that WICOR did not consider seeking a patent raises a reasonable inference that WICOR personnel did not consider this project highly innovative to the field of computer science. More to the point, the plaintiff offered no evidence that its software has been marketed to or used by other utility companies. Even though the project resulted in lower operating costs for WICOR, this is not sufficient to satisfy the test for innovativeness, which requires a high degree of innovation, not just a reduction in operating expenses. The court therefore concludes that the CIS project does not meet the innovative test for purposes of the research tax credit.

[33] At this juncture, it is appropriate to comment on a recent case brought to the court's attention by plaintiff, Tax and Accounting Software Corporation, Tim E. Kloehr and Sheryl Kloehr v. United states of America, 2000 U.S. Dist. Lexis 12124 (N.D. OK., July 31, 2000). The district court in this case distinguishes its application of section 41 to computer software programs from that found in Norwest and United Stationers. The Kloehr Court concludes that section 41 should be construed more strictly for internal use" software than it is for commercial" software. Id. at *12. The taxpayer in that case created and sold software programs designed for use by tax and accounting professionals; the programs were a commercial success and the taxpayer had thousands of customers. In the district court's opinion, the research credit should be used "as an incentive to encourage taxpayers to incur the cost of research in developing new products and stimulating the economy." Id at *23. As a result, the court concluded that the IRS had applied too strict a standard in regard to the discovery and experimentation tests.

[34] First, this court does not believe that Kloehr is applicable to the facts of this case. Nowhere in the record is there any indication that the development of the CIS system was a "commercial" success, i.e. a system that would benefit the public. It is accurate to say that the system was of benefit to WICOR customers, and while improving service to customers is a laudable goal, that remains an internal benefit. Market competition, not the qualified research tax credit, should motivate companies to reduce their costs, become more efficient and improve service. In that way, they will gain more customers. So even if this court were to agree with the Kloehr court that commercial research of computer software should be viewed differently under section 41, that is not the case here. Second, United Stationers states the law in this circuit, and the Seventh Circuit adopted the approach taken in Norwest. For example, the Seventh Circuit cited Norwest when it stated, "Congress clearly intended, then that qualifying research pass a high threshold of innovation and be of broad effect." United Stationers, 163 F.3d at 444.

4. SIGNIFICANT ECONOMIC RISK TEST

[35] Finally, the parties dispute whether the plaintiff's software project constituted a significant economic risk for purposes of section 41. "The significant economic risk test requires that the software development involve significant economic risk 'as where the taxpayer commits substantial resource to the development and also there is substantial uncertainty, because of technical risk, that such resources would not be recovered within a reasonable period.'" Norwest, 110 T.C. at 499 (quoting H. Conf. Rept. 99-841 (Vol. II)). With regard to internal use software, a higher threshold of technological advancement is required than in other fields. Id.

[36] Anderson's Dennis Terry acknowledged that the project did not involve an inordinate amount of technical risk because Anderson had prior experience with customer information systems from its work on such systems with other utility companies. Terry stated that he knew the customer information system could be completed because Anderson had successfully completed a similar project with Wisconsin Public Service. Terry testified at trial that the risk he envisioned for the project revolved around scheduling concerns, not technical concerns:

     A. The risk in my mind was whether or not we would be able to do

[37] In addition, the government's experts -- Srivastava and Niccum -- testified that the primary risk presented to the CIS project was managerial risk, such as the risk that WICOR management would reduce or eliminate funding for the project. Had WICOR developed the CIS project without Andersen Consulting's input, then WICOR may very well have incurred technical risk with respect to integrating the various aspects of the CIS. However, Andersen's prior experience with similar software projects, including software development for other utilities, sufficiently decreased the technical risk presented to WICOR such that WICOR fails to show that it incurred significant economic risk as a result of the technical risks presented by the CIS project.

[38] To conclude, the court finds that the plaintiff is not entitled to a research credit under 26 U.S.C. section 41 because its computer software project did not pass the discovery test, did not constitute a process of experimentation, was not sufficiently innovative, and did not constitute a significant economic risk.

[39] Based upon the court's conclusion on this issue, and its prior decision granting the defendant's motion for partial summary judgment, judgment will be entered in favor of the government and against the plaintiff. The plaintiff's complaint and this action shall be dismissed on the merits.

[40] SO ORDERED.

[41] Dated at Milwaukee, Wisconsin this 28th day of September, 2000.

                                   Aaron E. Goodstein