2008 WI 88
|
Supreme Court of Wisconsin |
|
|
|
|
Case No.: |
2004AP3239 |
|
Complete Title: |
|
|
|
Wisconsin Department of Revenue, Petitioner-Respondent-Petitioner, v. Menasha Corporation, Respondent-Appellant. |
|
|
|
|
|
|
|
|
REVIEW OF A DECISION OF THE COURT OF APPEALS 2007 WI App 20 Reported at: 299 (Ct. App. 2007-Published) |
|
|
|
|
Opinion Filed: |
July 11, 2008 |
|
Submitted on Briefs: |
||
Oral Argument: |
November 29, 2007
|
|
|
|
|
Source of Appeal: |
|
|
|
Court: |
Circuit |
|
County: |
Dane |
|
Judge: |
Steven D. Ebert
|
|
|
|
Justices: |
|
|
|
Concurred: |
CROOKS, J., concurs (opinion filed). PROSSER and ROGGENSACK, JJ., join the concurrence. |
|
Dissented: |
ABRAHAMSON, C.J., dissents (opinion filed). BRADLEY and BRADLEY, J., dissents (opinion filed). |
|
Not Participating: |
|
|
|
|
Attorneys: |
|
For the petitioner-respondent-petitioner the cause was argued by F. Thomas Creeron III, assistant attorney general, with whom on the briefs was J.B. Van Hollen, attorney general.
For the respondent-appellant there were briefs by Andrew L. Nelson, Michael B. Van Sicklen,
and Foley & Lardner LLP,
An amicus curiae brief was filed by Rebecca Kathryn Mason, Brady Williamson, and
2008 WI 88
notice
This opinion is subject to further editing and modification. The final version will appear in the bound volume of the official reports.
REVIEW of a decision of the Court of Appeals. Affirmed.
¶1 ANNETTE KINGSLAND ZIEGLER, J. This is a review of
a published court of appeals' decision,[1]
which reversed the decision of the
¶2 This appeal presents the following two issues: (1) what is the proper level of deference that this court should give to the Commission's decision; and (2) did the Commission reasonably conclude that the R/3 System was a custom program and therefore not subject to sales and use tax.
¶3 As to the first issue, we consider the level of deference that this court must give to the Commission's interpretation of the statute and the administrative rule. As a part of that determination, we also consider whether the Commission was required to give deference to the DOR's interpretation of Wis. Admin. Code § Tax 11.71(1)(e). We conclude that the Commission's statutory interpretation of Wis. Stat. § 77.51(20) is entitled to due weight deference and that its rule interpretation of Wis. Admin. Code § Tax 11.71(1)(e) is entitled to controlling weight deference. We further conclude that when a DOR decision is appealed by the taxpayer to the Commission, the Commission is not required to give deference to the DOR's interpretation of Wis. Admin. Code § Tax 11.71(1)(e) when deciding that appeal.
¶4 The second issue this court must decide is whether the Commission reasonably concluded that the R/3 System was a custom program and therefore not subject to sales and use tax. We conclude that when applying the controlling weight deference standard to the Commission's interpretation of Wis. Admin. Code § Tax 11.71(1)(e), the Commission reasonably interpreted the rule and concluded that the R/3 System was custom.
¶5 This decision has great import to the average taxpayer in this
state. More typically, it is the
individual taxpayer who seeks a fair and neutral hearing before the Commission
when that person believes that he or she has been taxed incorrectly by the
DOR. If the Commission must defer to the
DOR, the average taxpayer does not receive a fair hearing before a neutral
tribunal. Although the Commission is subject
to judicial review, the legislature specifically charged the Commission as
"the final authority for hearing and determination of all questions of law
and fact" under the tax code. We
must not second guess that act of the legislature.
I. UNDISPUTED FACTS
¶6 The Commission made the following findings of fact in the
underlying dispute:[5] Menasha is a Wisconsin corporation with
headquarters in
¶7 In April of 1995, Menasha began discussions with SAP regarding its R/3 System and made it clear that "a critical factor in its selection of a software system was one that could be customized to fit its business needs." The initial R/3 System consists of more than 70 software modules each of which is designed to "provide a rudimentary business and accounting computer software system for a segment of the client's business" such as "accounting and finance" or "personnel." The client, in this case Menasha, can select the modules it wishes to use. SAP provides a company, such as Menasha, with CD-ROMs containing all of the R/3 System modules, but SAP provides the client with the access codes only for modules that the client licenses for its particular business operation.
¶8 "As provided, the basic modules of the R/3 System contain a business and accounting system that must be customized to fit a client's business operations." "It is only after this customization process is completed that the client has a usable software system that will serve its business and accounting needs." In other words, the system is not usable to a client as sold and must be modified to fit a client's business operations. By use of a computer language——ABAP/4——the R/3 System is customized. However, "the ABAP programming is part of a larger development environment within the R/3 System called the Development Workbench, which offers many tools for customizing the R/3 System to fit a customer's business needs." SAPScript is used, within the Development Workbench, to create client-specific forms. Data Dictionary Objects "is used for changing and adding fields to tables provided by the basic R/3 System modules," and Data Dictionary Objects "is also used to create search helps" and "to create indices to make table access times more efficient."
¶9 "The two most common uses for ABAP are to permit the design of custom reports and to develop custom interfaces for the R/3 System." However, the ABAP programming is also used in the "creation of conversion programs that change data into a format usable by the R/3 System, and the creation of custom programs to run parallel to the R/3 System to fulfill business functions not provided by the R/3 System." By virtue of the licensing of the R/3 System, SAP customers "almost always retain either SAP or SAP's designated consultants to assist" in customizing the basic R/3 System.
¶10 When arguing to the Commission, the parties disagreed over how to characterize the sophistication of the initial or basic R/3 System modules that are provided to all customers. Menasha argued that the R/3 System provided "only a very basic" business and accounting system. The DOR, on the other hand, argued that the modules provided a "sophisticated" business and accounting system. The DOR cited to two treatises on implementing the R/3 System for its assertions regarding the sophistication of the R/3 System. The Commission, however, concluded that excerpts from the treatises that the DOR cites to "appear to describe the overall R/3 System and do not specifically address the [initial or basic] R/3 System modules." Furthermore, the Commission concluded that the "sophistication of the modules is largely immaterial because both parties agree that the R/3 System modules are only usable once the software system has been customized."
¶11 On April 20, 1995, SAP conducted a demonstration of its R/3 System. Following the initial demonstration, SAP was asked to conduct a demonstration with sample data from Menasha. Prior to preparing the demonstration, SAP spent a number of days at Menasha collecting information regarding Menasha's business operations, and SAP had extended conversations with Menasha's officers and employees regarding its diverse operations. It took SAP four weeks to prepare the demonstration. Menasha subsequently requested additional explanations and clarifications on the modification tools and techniques available within the R/3 System and a demonstration of the ABAP programming tools used to make modifications to the system.
¶12 Menasha understood that the requisite customization process could take years to complete and would cost tens of millions of dollars. Menasha's budget for purchasing the R/3 System included the costs that it expected to pay both SAP and SAP's designated consultants for the configuration, modification, and customization of the R/3 System. In addition, Menasha understood that "without this customization, the system would be of no value to its operations." "The customization was necessary to justify any amount spent on the licensing of the basic R/3 System modules." Menasha's board of directors approved the licensing of the R/3 System acknowledging that the projected cost of implementation would be approximately $46,575,000.
¶13 Menasha purchased the R/3 System from SAP on September 27, 1995, for $5.2 million. The licensing agreement contained no provision for customization of the R/3 System by SAP. However, SAP had advised Menasha during the course of demonstrations that because of the complexities of the system and substantial customization necessary to make the R/3 System usable, Menasha would be required to retain either SAP consultants or a consultant designated by SAP. Because SAP was unable to supply all of the necessary consultants for the installation and customization of the R/3 System, Menasha would have to work with one of the SAP designated consultants. Menasha chose ICS Deloitte (ICS) as its designated SAP consultant.
¶14 From September 1995 until March 1996, Menasha worked with ICS to analyze Menasha's systems, prepare Menasha's hardware for installation, and begin introducing Menasha's technology team to the intricacies of customizing the R/3 System. During the pre-installation phase, SAP representatives provided training and other support such as: providing a single point of contact for all questions, concerns, and communications; keeping abreast of project plans and status; providing SAP issue resolution and escalation; providing input as an implementation consultant; scheduling of professional services from SAP; assisting in creating a training curriculum; assisting with the planning and execution of the project plans; and communicating SAP product information.
¶15 The R/3 System was delivered to Menasha on multiple CD-ROM disks. Initial installation began on March 25, 1996, and downloading was complete on March 27, 1996. A former employee of SAP loaded the R/3 System onto a new computer purchased by Menasha. This former SAP employee was retained for the specific purpose of providing support in the initial installation of the system. During this time, SAP's former employee had access to SAP's online support system.
¶16 The R/3 System, upon delivery, did not provide adequate processing for Menasha. Each subsidiary put together an implementation team consisting of Menasha's information support personnel, SAP representatives, ICS representatives, and other third-party consultants. "The members of the implementation team worked under the direction of SAP and ICS [] to determine the operational and functional needs of each of [Menasha's] subsidiaries in order to configure and customize the basic modules of the R/3 System to fit the needs of each subsidiary." If the implementation team was not able to configure and modify the R/3 System, the team would refer the problem——described as a "functional gap"——to the ABAP programming team. "The programming team's job was to draft ABAP code to fill the functional gaps or to integrate the R/3 System with other systems that could provide the functionality required by the subsidiary." The ABAP programming team was directed by SAP and ICS and included people from Menasha and third-party consultants. Menasha contracted with SAP to provide an on-site programmer who worked with the ABAP programming team and became a member of a subsidiaries programming team.
¶17 The implementation and ABAP programming team members worked to customize the R/3 System from March 27, 1996, to January 1, 1997, in order to meet Menasha's functional needs. "[T]he ABAP programming team created codes for hundreds of user exits to the R/3 System to integrate external programs with the R/3 System." In addition, "the ABAP programming team created new subsystems to run parallel to the R/3 System for operations that were not available within the R/3 System, but were critical to [Menasha's] business." The ABAP programming team customized fields and reports within the R/3 System.
¶18 In total, more than 3,000 modifications were made to Menasha's R/3 System by the implementation and programming teams. Throughout, SAP representatives provided both off- and on-site technical and functional support to Menasha directly and through ICS. SAP also provided "patches" for the R/3 System to correct functional gaps. "Some of these patches included new source code written specifically for [Menasha's] system, to address the shortfalls of the R/3 System as it applied to [Menasha's] business."
¶19 Once customization was completed, Menasha worked with SAP and ICS to test the R/3 System to ensure it complied with Menasha's operations requirements. Testing lasted three to four months and included running real data through the system to determine whether it was operational in accordance with Menasha's required specifications. During testing, SAP, either directly or through ICS, provided off- and on-site technical support to Menasha and assisted in trouble-shooting problems during testing.
¶20 After testing was complete, all relevant employees from all subsidiaries were required to attend two- to five-day classes provided by ICS and Menasha's information support staff. ICS and Menasha's information support staff prepared extensive written materials for training Menasha's employees. Information support staff was also required to attend additional training classes for maintaining and making modifications to trouble-shoot problems once the R/3 System was in use. SAP provided Menasha and ICS with extensive material for the running and maintenance of the R/3 System.
¶21 Once training was completed, Menasha began to switch to the new R/3 System. Because further customization for each subsidiary was required, Menasha elected to implement the R/3 System on a subsidiary-by-subsidiary basis. The implementation of the R/3 System took nearly seven years. SAP representatives and ICS provided support to Menasha and assisted in solving any operational problems that occurred. At least 23 different SAP consultants were involved in the on-site installation and implementation. Menasha continues to contact SAP on a weekly basis for assistance and support. In addition, SAP provides Menasha with upgrades, new releases, and patches to the R/3 System on at least a quarterly basis.
¶22 The installation and customization of the R/3 System cost more than $23 million of which $5.2 million was for the core R/3 System. To customize the system for Menasha's business, Menasha paid SAP $2.5 million, ICS approximately $13 million, and third-party consultants approximately $775,000.
¶23 In 1998, the DOR audited SAP. In that audit, the DOR determined that the R/3 System was non-custom and thus taxable. SAP did not dispute that determination.[6] Separately, SAP and Menasha entered into an agreement whereby Menasha would pay $342,614.45 in sales tax for the R/3 System but that Menasha would dispute that payment and file a claim for a refund. The Commission found the following undisputed facts regarding that audit:
65. During 1998, [DOR]
audited the American subsidiary of SAP for liability under
66. In the course of the
audit, SAP and [DOR] agreed that SAP's sales of R/3 software in
67. As a consequence of
the agreement by SAP that sales of the R/3 program modules were taxable, SAP
paid [DOR] more than $1.9 million in tax and interest for sales to
DOR v. Menasha Corp., 2007 WI App 20, ¶17, 299 Wis. 2d 348, 728 N.W.2d 738 (citing Menasha Corp. v. DOR, Wis. Tax Rptr. (CCH) ¶400-719, at 32,847-48 (WTAC 2003)).
II. PROCEDURAL FACTS
¶24 Although
the date is unknown from the record before us, Menasha initially paid sales and
use tax for the R/3 System.
However, on June 30, 1998, Menasha filed a refund claim with the DOR in
the amount of $342,614.45 for the sales and use tax it had paid.[7] On April 28, 1999, the DOR denied Menasha's
refund claim. Menasha then petitioned
the DOR for a redetermination. On March
1, 2001, the DOR denied that request, and as a result, on April 27, 2002,
Menasha requested that the Commission review the DOR's determination. Appearing before the Commission, both Menasha
and the DOR filed separate motions for summary judgment. Accordingly, both parties submitted briefs,
proposed findings of fact, affidavits, and exhibits. On December 1, 2003, the Commission granted
Menasha's motion for summary judgment and denied the DOR's motion for summary
judgment. The Commission, thereby,
reversed the DOR's denial of Menasha's petition for redetermination and granted
Menasha's refund claim. The Commission
concluded that the R/3 System was custom and not subject to sales and use tax. The DOR then petitioned the Dane County Circuit
Court for review pursuant to Wis. Stat. §§ 73.015(2),
227.52 and 227.53. Both parties again
moved for summary judgment and submitted briefs, affidavits, and exhibits. The
A. Tax Appeals Commission's Decision
¶25 Menasha petitioned the Commission for review of the DOR's determination. The DOR had determined that the R/3 System was not custom and, therefore, it was subject to sales and use tax.[8] Both parties moved for summary judgment and submitted briefs, proposed findings of facts, affidavits, and exhibits. Thus, both parties, Menasha and the DOR, had the opportunity to present their respective arguments to the Commission. The Commission concluded that no genuine issues of material fact existed, and as a result, decided the matter at summary judgment. The Commission concluded that the R/3 System was "a custom program because of the significant investment [Menasha] made in presale consultation and analysis, testing, training, written documentation, enhancement, and maintenance support, and because it [was] not a prewritten program." As a result, it found the R/3 System exempt from sales and use tax. The Commission turned to the definition of "custom" in Wis. Admin. Code § Tax 11.71(1)(e). That definition instructs one to consider "all the facts and circumstances," and it recites seven factors to be considered. The Commission concluded that every factor need not be met for the program to be deemed custom, but rather, it determined that the factors in the rule are to be weighed along with all the other facts and circumstances.[9]
¶26 The Commission considered each factor. Under the first factor, regarding presale consultation, the Commission found the DOR had conceded that significant presale consultation and analysis occurred. Under the second factor, regarding loading and testing of the software, the Commission found that a former SAP employee loaded the software and thus, "[a]rguably, this part of the factor is a draw," but "the fact that a former SAP employee loaded the software weighs in favor of finding that the software at issue is custom software." Factor two also includes testing as a consideration and the Commission concluded that the testing——regardless of who conducted the testing——took three to four months, and thus, the second factor supported the conclusion that the software was custom.
¶27 Under the third factor, regarding training and written documentation, the Commission found that the DOR conceded that substantial training and written documentation was required. Under the fourth factor, regarding enhancement and maintenance support, the Commission found that the DOR conceded that the R/3 System needed enhancement and maintenance support. Under the fifth factor, regarding a rebuttable presumption that a program is not custom if it cost $10,000 or less, the Commission concluded that factor five did not apply because the cost greatly exceeded $10,000.
¶28 The Commission determined, with regard to factor six, that
"[t]his case hinge[d] on whether the R/3 System [was] a prewritten
program." "The other factors
set forth in the rule either lead to a conclusion that the R/3 System is custom
software or are neutral." The
Commission concluded that the R/3 System was not a prewritten program under the
rule's definition of "prewritten."
"Prewritten programs", often referred to as "canned programs", means programs prepared, held or existing for general use normally for more than one customer, including programs developed for in-house use or custom program use which are subsequently held or offered for sale or lease.
(Emphasis added.)
¶29 The Commission utilized the rule's definition of "prewritten" and reasoned that the R/3 System was not prepared, held or existing for general use. The DOR urged the Commission to consider the dictionary definition of "prewritten," but the Commission rejected this argument because the definition of "prewritten" was provided by the rule.
¶30 The DOR also utilized "numerous descriptions of the R/3 System as providing standard solutions or being a standard package" (emphasis omitted). The Commission, however, concluded the DOR "appears to equate the rule's phrase 'general use' with standard. This misses the point." The Commission reasoned that "[t]he issue is not whether the end result is a program that provides standard business applications, but rather the obstacles one must overcome to get to apply the software." Thus, the Commission declined to construe the phrase "general use" with the word "standard" as the DOR repeatedly described the R/3 System as "providing standard solutions or being a standard package." "What matters[, the Commission concluded,] is the process by which the application is accomplished, regardless of whether the application is standard or not."
¶31 Under the DOR's analysis, the Commission stated, "a vendor
could develop a software program completely from scratch . . . , and if the resulting program
provided standard business applications, the software would not be custom
software." This, however, the
Commission concluded "does not make sense."
¶32 The
Commission stated that "even the [DOR] concedes the following" facts:
1. The basic modules of the R/3 System must be subject to a certain degree of customization, and it is only after this customization process is complete that the client has a usable software system;
2. As delivered, the R/3 System was inadequate for petitioner's use;
3. Members of petitioner's implementation team working with SAP and ICS Deloitte determined the operational and functional needs of each subsidiary in order to configure and customize the system;
4. The implementation team worked to configure and modify the R/3 System to adapt the system to each subsidiary's identified needs;
5. The implementation and ABAP programming teams worked to customize the R/3 System to meet petitioner's functional needs;
6. The ABAP programming teams created codes for hundreds of user exits to integrate external programs with the R/3 System, so that petitioner was able to realize the functionality needed for its unique business while preserving the functional efficiencies of the R/3 System;
7. The ABAP programming teams created new subsystems to run parallel to the R/3 System for operations not available within the R/3 System more critical to petitioner's business;
8. The ABAP programming teams customized fields and reports within the R/3 System to insure it produced output to be useful to petitioner's business;
9. SAP provided petitioner with patches to correct functional gaps identified during implementation, some of which included new source code written to address shortfalls of the R/3 System;
10. In total, more than 3,000 modifications were made to petitioner's R/3 System.
Menasha Corp. v. DOR, Wis. Tax Rptr. (CCH) ¶400-719, at 32,855-56 (WTAC 2003).
¶33 Therefore, the Commission concluded, it took a significant amount of time, effort, and resources to make the R/3 System usable for Menasha. "The distinction between custom and prewritten programs hinges on the amount of effort necessary to get the software operational for the particular customer's needs." Because of the substantial amount of resources, time, and effort needed to make the R/3 System usable, the Commission stated, "we cannot conclude that the software at issue is prewritten." The Commission reasoned that if a program exists for general use, it will take little effort for it to be put in place for any user, whereas, if a program is useful only after a significant investment in "planning, testing, training, enhancement, and maintenance, then the software cannot be said to be prepared, held or existing for general use."
¶34 Under factor seven, regarding significant modification to an existing program, the Commission concluded that this factor did not apply in this case because the R/3 System was a custom program rather than an existing program. The Commission stated, "[t]his factor only makes sense in the overall scheme presented by [Wis. Admin. Code] section Tax 11.71(1)(e) if one concludes that 'existing program' means an 'existing program for general use' as that phrase is used in the definition of 'prewritten programs' in [Wis. Admin. Code] section Tax 11.71(1)(k)."
¶35 Finally, because all facts and circumstances must be considered, the Commission considered the cost of the software. It reasoned that because there is a presumption that anything under $10,000 is not custom, cost could be considered. The Commission concluded that the significant cost, $5.2 million, favored a conclusion that the R/3 System was custom.
B.
¶36 The DOR petitioned the Dane County Circuit Court for review of the decision by the Commission, and both parties again filed cross-motions for summary judgment. "Because the [circuit] court [found] error by the commission, the court vacate[d] the commission's decision and reinstate[d] DOR's determination" that sales and use tax must be paid on the R/3 System. The circuit court concluded that the Commission erroneously interpreted and applied the law to the facts, "as those facts are set forth in the commission's ruling."
¶37 The circuit court applied a due weight deference standard to the Commission's interpretation and application of Wis. Admin. Code § Tax 11.71(1)(e).[11] Thus, the circuit court stated it would "uphold the commission's reasonable interpretation unless the [circuit] court deems another interpretation more reasonable." The circuit court concluded that due weight deference was appropriate because the Commission had not developed expertise that would necessarily place it in a better position than the court to make judgments about the administrative rule.
¶38 The circuit court also concluded that because the DOR, and not the Commission, promulgated the administrative rule, the DOR's interpretation should be afforded controlling weight deference because "[a]n administrative agency's interpretation of its own rule is controlling unless plainly erroneous or inconsistent with that rule." The circuit court, thus, disagreed with Menasha's argument that the Commission was under no obligation to give deference to the DOR's construction of Wis. Admin. Code § Tax 11.71(1)(e).
¶39 While the circuit court found no error with the Commission's
interpretation of the rule's introduction, it disagreed with the Commission's
interpretation and application of some factors.
The circuit court
concluded:
The [circuit] court finds that [the R/3 System] was existing and prewritten when SAP sold it to Menasha because 1) while there may have been significant presale consultation and analysis, and significant testing of the customized system, the fact that a former SAP employee installed it rather than SAP weighs in favor of deeming R/3 prewritten; 2) while DOR conceded the third, fourth, and fifth factors, R/3 fits the definition of a prewritten program because a) it was already prepared and available for general consumption prior to the sale thereof to Menasha, b) it was held by SAP to be licensed to thousands of world-wide customers as requested, and c) it was not written solely for Menasha upon Menasha's request therefore; and finally 3) given that R/3 was an existing program, the seventh factor does apply, and the facts as set forth by the commission do not show that SAP performed the significant modification of R/3 that was required to make it useful to Menasha.
. . . .
. . . Where the software selected for purchase is an already existing program available for general use, significant modification must be made by the vendor for the software to be deemed custom. Thus, Menasha's purchase of R/3 did not involve the purchase of custom software. . . .
C. Court of Appeals' Decision
¶40 Menasha
appealed the circuit court's decision, and the court of appeals concluded that
the program was custom under Wis. Admin. Code § Tax 11.71(1)(e).
Citing to Caterpillar,[12]
the court of appeals stated, "'[b]ecause the commission is the final
administrative authority that reviews the decisions of the DOR, any deference
that might be due to the decision of an administrative agency is due to the
commission, not to the DOR.'" Menasha,
299
[B]oth Wis. Stat. § 73.01(4) and the [] case law plainly establish that the commission is the final authority on all the facts and questions of law regarding the tax code; the DOR is not. Any deference due to an administrative agency's decision in this context is given to the commission, not the DOR. Our deference is based on the commission's extensive expertise achieved by discharging its legislative duty as the final authority on the facts and questions of law in resolving tax disputes. . . .
Menasha, 299
¶41 The court of
appeals concluded that the Commission's decision was entitled to due weight
deference. It stated:
In addition to being designated the final authority on all questions of law involving taxes, the commission has generated and employed its substantial experience discharging its duty in construing the rules governing the taxability of tangible property since Wis. Stat. § 77.51(20) was enacted.
¶42 Applying
the due weight deference standard, the court of appeals concluded that the
Commission's interpretation was reasonable and consistent with the plain
language of the rule, and the DOR failed to offer a more reasonable
interpretation of the rule.[13] The court of appeals reasoned that the DOR
"appears to ignore the first five factors" of the rule and instead
focuses solely on factors six and seven.
The DOR argued that factors six and seven "are mandatory and provide
no exceptions, and therefore, [they] trump the other factors set forth in [Wis.
Admin. Code] § Tax
11.71(1)(e)1-5." However, the court
of appeals, citing to the Commission, concluded that "all the facts and
circumstances" must be considered under Wis. Admin. Code § Tax 11.71(1)(e). "Aside from its narrow and unreasonable
interpretation of [the rule], the DOR offers no reason for why [one] should
ignore the commission's determination that the tax rule shall be considered in
its entirety in determining whether a computer program is customized."
¶43 The
court of appeals concluded:
We agree with the commission's conclusion that § Tax 11.71(1)(e)6. should not be considered in isolation of the other factors in the rule. When the factors are read together, the most reasonable construction of the rule is that the definitions of "custom" and "prewritten" software in § Tax 11.71[(1)](e) and (k) are informed in large part by consideration of the first five factors of § Tax 11.71(1)(e). The commission reasonably concluded that the more effort required making the software usable, the more likely it is custom software. This conclusion is consistent with the plain language of the rule. . . .
III. STANDARD OF REVIEW
¶44 This
case requires interpretation of both Wis. Stat. § 77.51(20) and
¶45 "[W]hen
interpreting administrative regulations, we use the same rules of
interpretation as we apply to statutes."
DaimlerChrysler, 299
¶46 In
a case such as this, we review the Commission's decision rather than the
decision of the circuit court. Racine
Harley-Davidson, Inc. v. Wis. Div. of Hearings & Appeals, 2006 WI 86,
¶8 n.4, 292 Wis. 2d 549, 717 N.W.2d 184; DOR v. Caterpillar, Inc.,
2001 WI App 35, ¶6, 241 Wis. 2d 282, 625 N.W.2d 338. While we are not bound by an agency's——like
the Commission's——conclusions of law, we may defer to the Commission's legal
conclusions.
A. Deference to the Commission's Interpretation of the
Statute
¶47 This
court has articulated three possible levels of deference for an agency's
interpretation of a statute: great weight, due weight, and no deference. Racine Harley-Davidson, 292
Wis. 2d 549, ¶12; DOR v. River City Refuse Removal, Inc., 2007
WI 27, ¶¶32-35, 299 Wis. 2d 561, 729 N.W.2d 396.
¶48 Great
weight deference is given to the agency's statutory interpretation when each of
the following requirements are met: (1) the agency is charged by the
legislature with the duty of administering the statute; (2) the agency
interpretation is one of long standing; (3) the agency employed its expertise
or specialized knowledge in forming its interpretation; and (4) the agency's
interpretation will provide uniformity and consistency in the application of
the statute. Racine Harley-Davidson,
292
¶49 Due
weight deference is given to the agency's statutory interpretation when the
agency has some experience in an area but has not developed the expertise that
necessarily places it in a better position than a court to make judgments
regarding the interpretation of the statute.
¶50 No
deference is given to the agency's statutory interpretation when the issue is
one of first impression, the agency has no experience or expertise in deciding
the legal issue presented, or the agency's position on the issue has been so
inconsistent as to provide no real guidance.
¶51 In
the case at hand, we afford the Commission due weight deference on its
statutory interpretation of Wis. Stat. § 77.51(20). The legislature designated the Commission as
the final authority on all questions of law and fact arising under the tax
statutes——although, the Commission's decision is still subject to judicial
review.
¶52 While
we afford the Commission due weight deference on its interpretation of the
statute, this case centers on the Commission's interpretation of the regulation
or rule, Wis. Admin. Code § Tax 11.71(1)(e). The statute in question defines tangible——and
thus taxable——personal property as including "computer programs except custom
computer programs."
B. Deference to Commission's Interpretation of the
Regulation
¶53 "An
administrative agency's interpretation of its own rules or regulations is
controlling unless 'plainly erroneous or inconsistent with the
regulations.'" DaimlerChrysler,
299
¶54 When
applying controlling weight deference, we ask "whether the agency's
interpretation is reasonable and consistent with the meaning or purpose of the
regulation."
¶55 Because
the legislature has subjected the Commission's decision to judicial review, it
is the Commission's decision, rather than the DOR's decision, to which we must
give deference. Kamps v. DOR,
2003 WI App 106, ¶26, 264
¶56 While
the DOR concedes that courts must review the Commission's decision and not the
DOR's decision, it argues that the Commission should at least give deference to
or address the DOR's construction of its rule.
Because the Commission did not defer or specifically refer to the DOR's
construction, the DOR argues that this court should conduct a de novo
review of the Commission's rule interpretation.
We decline that request.
¶57 The
DOR asserts several arguments in favor of its conclusion that the Commission
must give deference to the DOR. While we
address those specific arguments below, in short, the DOR fails to articulate
any reason that persuades us to conclude that the Commission must defer to the
DOR's interpretation of its rule.
¶58 The
legislature designated the Commission as the final authority on all tax
questions. Requiring the Commission to
grant deference to the DOR would undermine the Commission's authority to review
tax questions. By virtue of creating the
Commission and identifying it as the agency charged with interpreting questions
that arise under our tax code, the legislature has divided the agencies' duties
with regard to the tax code. The DOR may
promulgate rules and even issue private letter rulings or interpretations that
give guidance to taxpayers, but the Commission is the final authority. Thus, it is for the Commission to interpret
those rules. The legislature has
designated the Commission, not the DOR, as the final authority on all tax
questions. It is within the province of
the legislature to enact the law.
¶59 Requiring
the Commission to give deference to the DOR not only undermines the
Commission's authority, it is troubling for the taxpayer who appeals the DOR's
decision. The Commission hears and
decides disputes between the DOR and individual taxpayers or entities. If the Commission, which serves a
quasi-judicial function, must defer to one party——namely, the DOR——how does the
taxpayer receive a fair hearing? The
taxpayer brings his or her appeal to the Commission at a significant disadvantage
if the Commission must defer to the taxpayer's opponent, the DOR.
¶60 While
the Commission may find it useful to consider the DOR's construction of its
rule by reviewing such things as published tax releases, technical information
memorandums, and private letter rulings, it need not ultimately
"defer" to the DOR's interpretation, nor must it specifically refer
to the DOR's construction in its decision.
In this case, while the Commission did not specifically refer by name to
the DOR's published tax releases, technical information memorandums, and
private letter rulings, these materials were submitted to the Commission. Moreover, the DOR submitted briefs, proposed
findings of fact, affidavits, and exhibits.
The Commission states in its written decision that "[b]ased upon
the submissions of the parties and the entire record in this matter, the
Commission hereby rules and orders as follows[]." Therefore, the Commission did consider the
DOR's submissions since the DOR was a party.
¶61 No
authority exists for requiring the Commission to defer to the DOR's
construction of its rule. The
legislature does not mandate such a requirement, and we decline the opportunity
to create one. The DOR's deference argument
is intriguing because the DOR promulgated this rule, and thus, it is, at least
arguably, reasonable to argue that the Commission should give the DOR
deference. However, this would ignore
the boundaries that the legislature created when it gave the Commission final
authority over all tax questions.
¶62 The
DOR, citing to Hillhaven,[16]
argues that an agency's interpretation of its own rule is controlling unless
plainly erroneous or inconsistent with the rule's language. Thus, according to the DOR, the Commission
should afford the DOR deference because the DOR promulgated the rule at issue. As a general rule, that is correct. However, the DOR fails to acknowledge a
critical difference between this case and Hillhaven. In Hillhaven, no adjudicative agency
body was charged with reviewing the Department of Health and Family Services'
decision. The department's decision was
appealed directly to the circuit court.
In the case at issue, the Commission reviews the DOR's decisions, and it
is the Commission's decision, not the DOR's, that is appealed to the circuit
court. It is significant that the
legislature, in this case, designated an "intermediary"——the
Commission——to review the DOR's decision and to be the final authority on tax
questions. Again, it is the Commission's
decision that is subject to judicial review and to which we must give
deference, if deference is to be given.
¶63 The
DOR, citing to State ex rel.
¶64 The
DOR also relies on Griffin to further argue that by declining to require
the Commission to give deference to the DOR, the Commission is given more power
than Wisconsin courts because courts must defer to an agency's interpretation
of its rule unless the agency's interpretation is plainly erroneous or
inconsistent with the regulations.
Courts give deference to an agency, either a line agency or an
adjudicative agency, because the agency is in a better position to interpret
the rule than the court; the agency has significant and continual experience
with the area of law in question.[18]
¶65 However,
when, as in this case, the Commission reviews the DOR's decision, the
Commission need not rely on the DOR's expertise because the Commission itself
is experienced with interpreting and applying the tax code, and is, in fact,
the final authority on all tax questions.
See DOR v. Heritage Mut. Ins. Co., 208
¶66 The
DOR, citing to Pfeiffer,[19]
argues that because it "performed all of the[] legislative functions, it
necessarily 'knows the specific purposes of the regulations it has promulgated'
and 'is in the best position to interpret its own regulations in accordance
with their underlying purposes.'"
However, Pfeiffer and other cases cited by the DOR do not
consider how the quasi-judicial function of the Commission alters this
analysis. While it remains true that the
agency that promulgates the rule should be given controlling weight deference,
this is altered here because the legislature designated an agency, the
Commission, with the task of reviewing decisions and being the final
authority. In DaimlerChrysler, we
stated that "[a]n administrative agency that regularly works with the
rules and regulations of another agency, whose actions it is authorized by the
legislature to review, is in the best position to interpret such rules and
regulations because the agency knows the specific purposes of the rules and
regulations that have been promulgated, and has expertise in the area the
agency is called upon to review." DaimlerChrysler,
299
¶67 The
DOR argues that "[n]owhere does DaimlerChrysler state that
LIRC would have reasonably interpreted DWD's rule if LIRC had either failed to
accord controlling weight to DWD's construction of DWD's own rule or had failed
to address DWD's prior official constructions of its
rule. . . ."
However, nowhere in DaimlerChrysler does the court suggest that
giving deference to or addressing the DWD's construction was critical to its
conclusion that the LIRC reasonably interpreted the regulation in
question. In fact, the LIRC's
interpretation was afforded controlling weight deference "because the LIRC
reasonably interpreted a rule adopted by the DWD, the LIRC's interpretation was
not inconsistent with the language of the rule or clearly erroneous, and the
LIRC was charged by the legislature with the duty of reviewing decisions . . . of the DWD, and does so
frequently."
¶68 The DOR, citing to River City Refuse Removal,[20]
argues that the Commission should not have been afforded any deference because
it had not issued prior decisions construing the substantive provisions of the
DOR's rule. However, in so arguing, the
DOR utilizes the standards of review for statutory interpretation instead of
rule interpretation. While this argument
may apply to statutes, as articulated in River City Refuse Removal, it does not necessarily
follow that it applies to interpretation of rules. No deference is due when an agency's statutory
interpretation is one of first impression.
River City Refuse Removal, 299
¶69 In any event, the Commission does have some experience with the
factors articulated in Wis. Admin. Code § Tax
11.71(1)(e). See, e.g.,
IBM Corp. v. DOR, Wis. Tax Reptr. (CCH) ¶202-854 (WTAC 1987); Health Micro Data Sys., Inc. v.
DOR,
¶70 Finally, the DOR argues that the acquiescence and nonacquiescence
provisions of Wis. Stat. §§ 73.01(4)(e)[21]
and 73.015(2)[22]
apply to the case at hand.[23]
In short, the DOR argues that when the DOR does not petition for judicial
review and does not file a notice of nonacquiescence after the Commission's
decision, the DOR has determined that the Commission has correctly construed the
statutes or rules at issue. On the other
hand, the DOR argues that when it petitions for judicial review or files a
notice of nonacquiescence, the DOR has determined that the Commission has
incorrectly construed the statutes or rules at issue. As a result, the DOR argues that in such
cases, the Commission's decision should be subject to de novo review.
¶71 We,
however, are persuaded by Menasha's argument that the acquiescence and
nonacquiescence provisions of Wis.
Stat. §§ 73.01(4)(e) and
73.015(2) do not apply to the case at bar.
The acquiescence rules, in general, expand the binding effect of the
Commission's decision. While the
Commission's decision is always binding on the parties of a particular dispute,
the Commission's decision binds the DOR on other potential disputes only if the
DOR does not appeal or file a notice of nonacquiescence. The acquiescence rules ensure stability in
tax matters because they allow taxpayers to rely on the Commission's decisions
and interpretations unless notice to the contrary is given——by appeal or
nonacquiescence——that the DOR does not intend to follow the Commission's
interpretations in subsequent taxpayer disputes. See DOR v. U.S. Shoe Corp., 158
¶72 Accordingly, we give the Commission's statutory interpretation due
weight deference, and we give the Commission's rule interpretation controlling
weight deference. While the statute uses
the word "custom," that word is defined only by the rule. Therefore, it is the rule interpretation that
controls in this case. Under controlling
weight deference, we ask whether
the agency's interpretation is reasonable and consistent with the meaning or
purpose of the regulation. If the
Commission's interpretation is reasonable and consistent with the meaning or
purpose of the rule, then we uphold the Commission's decision rather than
substitute our judgment for that of the Commission's.
IV. APPLICATION OF WIS. ADMIN.
CODE § TAX 11.71(1)(e)
TO THE R/3
SYSTEM
¶73 Tangible
personal property is subject to
"Custom programs" mean utility and application software which accommodate the special processing needs of the customer. The determination of whether a program is a custom program shall be based upon all the facts and circumstances, including the following:
1. The extent to which the vendor or independent consultant engages in significant presale consultation and analysis of the user's requirements and system.
2. Whether the program is loaded into the customer's computer by the vendor and the extent to which the installed program must be tested against the program's specifications.
3. The extent to which the use of the software requires substantial training of the customer's personnel and substantial written documentation.
4. The extent to which the enhancement and maintenance support by the vendor is needed for continued usefulness.
5. There is a rebuttable presumption that any program with a cost of $10,000 or less is not a custom program.
6. Custom programs do not include basic operational programs or prewritten programs.
7. If an existing program is selected for modification, there must be a significant modification of that program by the vendor so that it may be used in the customer's specific hardware and software environment.
¶74 The Commission concluded that Wis. Admin. Code § Tax 11.71(1)(e) instructs one to consider "all the facts and circumstances," including the seven factors listed in the rule. The factors, the Commission concluded, are not elements that must be met for a program to be deemed custom, but rather, the factors in the rule are to be weighed along with the other facts and circumstances. As a result, when considering all seven factors and all the facts and circumstances, the Commission concluded that the R/3 System was "a custom program" because of the significant investment Menasha made in "presale consultation and analysis, testing, training, written documentation, enhancement, and maintenance and support, and because [the R/3 System] is not a prewritten program."
¶75 The DOR disagrees with the Commission's interpretation of Wis. Admin. Code § Tax 11.71(1)(e). It argues that the R/3 System "was an existing set of prewritten, pre-engineered noncustom program modules" that was in general use in 1995. Thus, under the mandatory language of factor six of § Tax 11.71(1)(e), the R/3 System is prewritten and thus not custom. The DOR further argues that factor seven cannot be ignored because it contains mandatory language regarding existing programs. Therefore, the DOR argues that, under factor seven, the R/3 System was an existing program that SAP did not modify for Menasha prior to or even subsequent to delivering the R/3 System. Menasha, on the other hand, argues that the Commission's consideration of "all the facts and circumstances" was appropriate, and the DOR's interpretation excludes factors one through five from consideration.
¶76 Under the standard of review governing this case, we conclude that the Commission's interpretation is reasonable and consistent with the rule's language and the purpose of the regulation. The rule is instructive; a determination of whether a program is custom "shall be based upon all the facts and circumstances including" the seven factors listed in the rule. Because all the facts and circumstances must be considered, the factors are not elements to be satisfied but are factors that should be weighed along with all the facts and circumstances. Under the standard of review, we must now turn to the Commission's consideration of the seven factors and all the facts and circumstances.
¶77 Under the first factor, regarding presale consultation, the Commission concluded that the DOR conceded that significant presale consultation and analysis had occurred. Presale consultation began in 1993 with the analysis of Menasha's then current software, continued through 1995 when Menasha began discussions with SAP in April, and purchasing the license for the R/3 System in September of 1995. Presale consultation with SAP included SAP evaluating and collecting data from Menasha and SAP performing a sample demonstration for Menasha. Thus, factor one weighs in favor of the R/3 System being a custom program.
¶78 Under the second factor, regarding loading and testing of the R/3 System, the Commission concluded that even though the vendor did not load the software, a former SAP employee, who was retained by Menasha specifically for providing support during the installation, loaded the software. As a result, the Commission concluded that this weighs in favor of the software being custom. While we acknowledge that the rule's language in factor two specifically refers to whether the vendor loaded the software, it was reasonable——even if it is not conclusive——for the Commission to consider under all the facts and circumstances that a former SAP employee installed the software.
¶79 The second factor also requires consideration of whether the newly installed program must be tested against the program's specifications. The Commission concluded that the DOR conceded that after installation and customization was complete, the R/3 System was tested for three to four months. As a result, the Commission reasonably concluded that factor two weighs in favor of the R/3 System being a custom program.
¶80 Under the third factor, regarding whether substantial training and written documentation was required to use the software, the Commission concluded that the DOR conceded that substantial training and written documentation was needed. All employees were required to attend two- to five-day classes. Extensive written materials were prepared by ICS and Menasha's support staff. Additionally, Menasha's information support staff attended additional training focused on maintaining, modifying, and trouble-shooting problems once the R/3 System was in use. Thus, the Commission reasonably concluded that factor three weighs in favor of the R/3 System being a custom program.
¶81 Under factor four, regarding the extent to which the software requires enhancement and maintenance support by the vendor, the Commission concluded and the DOR conceded that the R/3 System needed enhancement and maintenance support. Menasha continues to contact SAP on a weekly basis for assistance and support. In addition, SAP provides Menasha with upgrades, new releases, and patches to the R/3 System on at least a quarterly basis. Thus, the Commission reasonably concluded that factor four weighs in favor of the R/3 System being a custom program.
¶82 Under factor five, regarding a rebuttable presumption that any program is not custom if it costs $10,000 or less, the Commission concluded this factor did not apply because the cost greatly exceeded $10,000. However, after its discussion of the seven factors, the Commission did consider the cost of the R/3 System when considering "all the facts and circumstances." In a footnote, the court of appeals' opinion expressed concern because no other portion of the rule indicates that cost of the program should be a factor in determining whether a computer program is a custom program. It, however, makes sense that cost can be evaluated when considering all the facts and circumstances. As addressed in factor five, cost is a factor, but not determinative. The Commission concluded that factor five should not apply in this case because "the R/3 System greatly exceeded [the] threshold" of $10,000 or less, but it reasonably concluded that cost could be considered when evaluating all the facts and circumstances. In the end, the Commission reasonably considered cost.
¶83 The DOR's argument relies on factors six and seven. In fact, the DOR not only argues that these two factors weigh in its favor, but it argues that these factors contain mandatory language that could render a program non-custom regardless of the other factors. Under factor six, regarding basic operational or prewritten programs not being custom, the Commission concluded that the R/3 System was not a prewritten program. "Prewritten" is defined as:
"Prewritten programs", often referred to as "canned programs", means programs prepared, held or existing for general use normally for more than one customer, including programs developed for in-house use or custom program use which are subsequently held or offered for sale or lease.
¶84 The Commission concluded that "[t]he distinction between custom and prewritten programs hinges on the amount of effort necessary to get the software operational for a particular customer's needs." Because of the substantial amount of resources, time, and effort needed to make the R/3 System usable, the Commission stated it could not conclude that the R/3 System was prewritten. A prewritten program would "require relatively little effort to be put in place for any user."
¶85 The Commission's conclusions based upon factor six and Wis. Admin. Code § Tax 11.71(1)(k), which defines "prewritten programs," are reasonable and consistent with the rule's language. The Commission reasoned that the first four factors "hinge[] on the degree to which the software is ready for use off-the-shelf," and as a result, this supported the Commission's distinction between prewritten and custom as determined by the amount of effort needed to bring software online for a particular customer.[24] The DOR takes issue with the Commission's reasoning and argues that factor six contains mandatory language that if a program is prewritten then it is not custom regardless of the effort required to make it usable.
¶86 We conclude that the Commission's overall conclusions regarding factor six are reasonable. We agree that the R/3 System was not a prewritten program. A program is prewritten and thus perhaps not custom if the program is "prepared, held or existing for general use." In other words, a prewritten program is ready to be used by those who purchase it. Thus, this factor "hinges" on whether a program is available for general use right off-the-shelf. While the purchaser of a prewritten program may make some minor changes in an effort to make it more efficient or to simply set it up, a prewritten program does not require significant modifications in order for one to make any use out of the program. In this case, the DOR concedes that the R/3 System must be subjected to a certain degree of customization before it is even usable by a client and it concedes that the implementation teams worked to identify and meet the specific needs of each subsidiary, which included:
· Members of petitioner's implementation team working with SAP and ICS Deloitte determined the operational and functional needs of each subsidiary in order to configure and customize the system;
· The implementation team worked to configure and modify the R/3 System to adapt the system to each subsidiary's identified needs;
· The implementation and ABAP programming teams worked to customize the R/3 System to meet petitioner's functional needs;
· The ABAP programming teams created codes for hundreds of user exits to integrate external programs with the R/3 System, so that petitioner was able to realize the functionality needed for its unique business while preserving the functional efficiencies of the R/3 System; and
· The ABAP programming teams created new subsystems to run parallel to the R/3 System for operations not available within the R/3 System more critical to petitioner's business.
¶87 Moreover, the DOR conceded that more than 3,000 modifications were made to the R/3 System. The R/3 System is not prewritten because it is not available for general use; in fact, it could not be used unless and until it was customized.
¶88 What makes factor six arguable is that the R/3 System exists and can be sold to everyone in the same form, i.e., all purchasers seem to receive disks, and they select specific modules and make other changes in order to make the R/3 System fit their needs. We acknowledge the DOR's arguments that the basic R/3 System was existing rather than created and that it was to some extent already encoded rather than built entirely from scratch. These considerations could weigh in the DOR's favor in that at least this initial program is non-custom and thus its $5.2 million cost should be taxable.
¶89 However, when we consider that this initial or basic R/3 System is
useless until modified, that customization is time-consuming and expensive and
thus significant, and that every purchaser must have the R/3 System modified in
order to make use out of it, we cannot conclude that this is a prewritten
program under factor six. It is not available for general use——in fact
no one can use it without modifications——and thus, it is not canned or
prewritten. Rather, the basic R/3 System
modules are available to become a custom program. Thus, the Commission's conclusion was
reasonable.
¶90 This is unlike a program such as Microsoft Excel. While a purchaser may wish to modify Excel so as to make it more efficient or enhance its capabilities, the program can and is used "right out of the box" by a number of purchasers. The R/3 System is not such a prewritten program.
¶91 The DOR argues that factor six contains mandatory language, which is controlling over the more general language of factors one through five. However, this is not consistent with other language in the rule. The rule instructs that all facts and circumstances shall be considered including the seven factors in the rule. It does not say that all factors may be considered but one or even two of the factors can be dispositive. The DOR argues that the Commission's interpretation renders factor six surplusage, but the DOR's interpretation could render the other factors and portions of the rule surplusage except for factor six. The reasonable approach is to consider all of the factors and all the facts and circumstances. The Commission did just that, and it is not for this court to override that decision if the Commission's actions were reasonable.
¶92 Additionally, because we agree with the conclusion that, under subsection (k) of factor six, this is not prewritten or canned, this factor alone does not control the analysis. It is worth noting, however, that under the current language, it would be difficult to conclude that a program is custom even when factors one through four weigh in favor of finding it custom but it is clearly prewritten or canned under factor six. Perhaps then, factor seven becomes crucial to the analysis. Today, however, this scenario is not before us.
¶93 The DOR also argues that the Commission's interpretation ignores the rule-making process. One definition of "prewritten" that was suggested but rejected was prewritten is "intended for general use and mass distributions as prepackaged ready-to-use programs." The DOR rejected this definition because there are situations where prewritten programs may not be mass produced. Presumably, the DOR takes issue with the following sentences in the Commission's decision:
Our conclusion that the distinction between "prewritten programs" and "custom programs" hinges on the amount of effort needed to bring software online for a particular customer is consistent with the first four factors and consistent with the definition of "prewritten programs." With respect to the latter, if a program is prepared, held or existing for general use normally for more than one customer, then the program will require relatively little effort to be put in place for any user. If, on the other hand, a program like the R/3 System is useful only after a significant investment of resources in planning, testing, training, enhancement, and maintenance, then the software cannot be said to be prepared, held or existing for general use.
Menasha Corp. v. DOR, Wis. Tax Rptr. (CCH) ¶400-719, at 32,856 (WTAC 2003) (footnotes omitted).
¶94 The DOR seems to overstate the Commission's position. The Commission simply reasons that the more effort it takes to get a program up and running, the more likely the program is not prewritten or canned and thus not available for general use. It does not, however, naturally follow from this proposition that the Commission has concluded that prewritten programs must be mass produced. In fact, a prewritten program could be utilized by only a few customers. If a program is available for general use, however, it is more likely——even if it is not dispositive——that a program does not require substantial effort, cost, and modifications to get a program up and running. While some of the Commission's reasoning regarding factor six is confusing, its overall conclusion was reasonable. The R/3 System is not a prewritten program because it is not available for general use.
¶95 Under factor seven, regarding an "existing program [] selected for modification," the Commission concluded that factor seven did not apply "given the Commission's conclusion that the R/3 System is a custom program." The Commission stated, "[t]his factor only makes sense in the overall scheme presented by [Wis. Admin. Code §] Tax 11.71[(1)](e) if one concludes that 'existing program' means an 'existing program for general use' as that phrase is used in the definition of 'prewritten programs' in [Wis. Admin. Code] section Tax 11.71(1)(k)." While the Commission could have used different language rather than stating that factor seven does not apply because it has determined the software is custom, the Commission's overall reasoning regarding factor seven is reasonable.
¶96 Instead of stating that factor seven does not apply because the program is custom, perhaps the Commission should have stated that factor seven does not apply because the program was not prewritten and thus not an existing program. "Existing program" under factor seven must refer to prewritten or canned programs because otherwise "existing programs" would include custom programs. The DOR states that "the sale of an 'existing' program is a sale of tangible personal property" and thus taxable, which means it is not custom. Prewritten or canned programs are available for general use. Thus, it makes sense that "existing program" means existing program for general use.
¶97 Factor seven should not be completely disregarded merely because other factors suggest that a program is custom. Factor seven, just like all the other factors and the facts and circumstances, should be used to determine whether a program is custom. While the Commission may not have selected the same language that this court would have selected, the Commission's subsequent statements and reasoning are sound. Factor seven considers existing programs for general use, and because the Commission reasonably concluded that this program did not exist for general use, factor seven does not apply.
¶98 The DOR, however, argues that factor seven does apply and that the R/3 System was prewritten and not modified by SAP before shipping it to Menasha, which the DOR determines to be critical to the application of factor seven. Because the Commission reasonably concluded that factor seven does not apply and we must be mindful of the deference we owe to the Commission, we reject the DOR's arguments. The Commission reasonably concluded that the R/3 System was not available for general use and therefore not canned or prewritten. Factor seven applies only when a program is an existing program, i.e., canned or prewritten, and thus existing for general use. The R/3 System did not exist for general use even though it did "exist," and thus, factor seven does not apply.
¶99 The DOR also argues that factor seven contains mandatory language and that the Commission's interpretation renders factor seven surplusage. For the reasons stated in our discussion of factor six, we reject the DOR's mandatory language and surplusage arguments. See ¶¶85-91.
¶100 Lastly, under all the facts and circumstances, the Commission considered the cost of the R/3 System, and it concluded that this weighed in favor of the R/3 System being a custom program. Under all the facts and circumstances, it is reasonable for the Commission to consider the cost of the R/3 System. While we acknowledge that a canned or prewritten program available for general use could be very expensive, cost is still a factor that may be considered by the Commission. Cost alone, however, is not dispositive. The Commission reasonably concluded that under the rule's language, all the facts and circumstances along with all seven factors must be considered and thus cost, even when over $10,000 and thus not relevant under factor five, may be considered.
¶101 We must afford the Commission's decision controlling weight
deference if it reasonably interprets the rule at issue. DaimlerChrysler, 299
¶102 The Commission concluded that the R/3 System was "a custom program because of the significant investment [Menasha] made in presale consultation and analysis, testing, training, written documentation, enhancement, and maintenance support, and because it [was] not a prewritten program."
¶103 We too are persuaded by the initial cost, the costs for modifications, the presale consultations over the span of a few years, the testing required once installed, the requisite training, the requisite enhancement and maintenance, and that the R/3 System cannot be used until modified——in this case some 3,000 modifications. As a result, we conclude that the Commission's conclusion that the R/3 System was custom is reasonable. Accordingly, we will not substitute our judgment for that of the Commission's.
¶104 While the program was not loaded by the vendor, and the vendor itself was not obligated pursuant to the licensing agreement to modify the R/3 System here, we still conclude that the R/3 System is a custom program. Under the rule's current language, it is irrelevant whether the vendor or an independent consultant carries out the modifications, and it is irrelevant whether those modifications take place before or after it leaves the vendor. The rule's language simply does not identify those as factors to be considered. While who modifies an existing program, under factor seven, is relevant, this does not appear, under the rule's language, to extend to situations that do not involve prewritten or canned programs.
¶105 The Commission reasonably concluded that all the facts and circumstances and all seven factors must be considered when determining whether a program is custom. Applying this construction to the particular facts of this case, the Commission reasonably concluded that the R/3 System was a custom program.
¶106 Finally, a brief response to the dissents is warranted. Once one sifts through the unnecessary rhetoric of the dissents, it is clear that the dissents are based on certain fundamental flaws. For example, the dissents undertake a project to rewrite the law. The legislature specifically designated the Commission as "the final authority for hearing and determination of all questions of law and fact" under the tax code. The Commission's decision is entitled to deference. The Chief Justice's dissent dissects Wis. Admin. Code § Tax 11.71(1)(e) in a manner that ignores the plain language of the entire rule, and instead, it creates a never before recognized or argued two-prong test. See Chief Justice Abrahamson's dissent, ¶¶180-182. Neither the DOR nor Menasha use this first sentence of the rule as the first prong of a two-prong test. Despite that dissent's assertions, the rule does not end with the first sentence.[25] Rather, as Attorney McGinnity so aptly stated at the court of appeals' oral argument on November 17, 2005: "The preface language in Sub (1)(e) talks about [a] custom program as being one that accommodates the special processing needs of the customer. And then the rule goes forward with a number of factors as to how you figure out if it is a program that is accommodating the special processing needs of the customer." This correct interpretation is also in conflict with Justice Bradley's interpretation of the rule. See Justice Bradley's dissent, ¶¶217-225. However, because the plain language of the rule specifically states that "[t]he determination of whether a program is a custom program shall be based upon all the facts and circumstances, including the following" seven factors, we disagree with Justice Bradley's interpretation. As to the remaining arguments raised by the dissents, they have been otherwise addressed in the majority opinion.
V. CONCLUSION
¶107 Accordingly, we affirm the court of appeals' decision. We conclude that the Commission's statutory interpretation of Wis. Stat. § 77.51(20) is entitled to due weight deference, and that its rule interpretation of Wis. Admin. Code § Tax 11.71(1)(e) is entitled to controlling weight deference. We further conclude that when a DOR decision is appealed by the taxpayer to the Commission, the Commission is not required to give deference to the DOR's interpretation of Wis. Admin. Code § Tax 11.71(1)(e) when deciding that appeal.
¶108 Finally, we conclude that when applying the controlling weight deference standard to the Commission's interpretation of Wis. Admin. Code § Tax 11.71(1)(e), the Commission reasonably interpreted the rule and concluded that the R/3 System was custom.
By the Court.—The decision of the court of appeals is affirmed.
¶109 N. PATRICK CROOKS, J. (concurring). While I join the majority opinion, I write separately to emphasize that resolving the issue of deference is key to a correct decision in this case. Rules regarding deference are important limits that respect the different roles of the branches of government. Granting the appropriate level of deference to the Commission's interpretation of Wis. Admin. Code § Tax 11.71(1)(e) and (k)[26] and Wis. Stat. § 77.51(20),[27] I concur that the Commission's decision should be upheld.
¶110 The threshold question on the issue of deference is whether it is the DOR's interpretation or the Commission's interpretation of the statute and the rules that is entitled to deference. I concur that it is the Commission's interpretation which is entitled to deference, while recognizing that the DOR's interpretation certainly must be carefully reviewed as well.
¶111 In DaimlerChrysler v. LIRC, 2007 WI 15, 299 Wis. 2d 1, 727
N.W.2d 311, we treated a rule promulgated by the DWD and reviewed by LIRC as
the functional equivalent of LIRC's own rule for purposes of analyzing the
level of deference due.
¶112 Further, granting deference to the Commission is consistent with the analysis in Caterpillar, which, like this case, concerned a conflict between the DOR and the Commission. There the court of appeals found that "[b]ecause the commission is the final administrative authority that reviews the decisions of the DOR, any deference that might be due to the decision of an administrative agency is due to the commission, not to the DOR." DOR v. Caterpillar, Inc., 2001 WI App 35, ¶6, n.3, 241 Wis. 2d 282, 625 N.W.2d 338, review denied, 2001 WI 43, 242 Wis. 2d 545, 629 N.W.2d 784. A fair reading of the DaimlerChrysler and Caterpillar cases leads to the conclusion that the agency whose interpretation is entitled to deference here is the Commission.
¶113 The deference afforded to the Commission's interpretation of the
administrative rules is significant in this case, since such rules are the
functional equivalent of the Commission's rules. "An administrative agency's
interpretation of its own rules or regulations is controlling unless 'plainly
erroneous or inconsistent with the regulations.'" DaimlerChrysler,
299
¶114 I concur with the majority that the Commission's interpretation of the relevant administrative rules——which contain the language crucial to resolving this case——is entitled to controlling weight deference. The Commission applied all seven factors in Wis. Admin. Code § Tax 11.71(1)(e), as well the definition of "prewritten programs" in Wis. Admin. Code § Tax 11.71(1)(k) in determining whether the R/3 system here was a custom program or a prewritten program. I am satisfied that the Commission’s decision is neither plainly erroneous nor inconsistent with the rules, and that it is reasonable.
¶115 In their dissents, Chief Justice Abrahamson and Justice Bradley
decline to defer to the Commission's interpretation on the grounds that the
interpretation is inconsistent with the administrative rules. Chief Justice Abrahamson's dissent, ¶143; Justice Bradley's
dissent, ¶214. That conclusion rests on an alternative
interpretation that carves just eight words from a single sentence, taken from the definition of custom programs,
designates those eight words as the relevant definition, and seemingly ignores
the surrounding text. Part of
¶116 The error in both dissents comes in taking the last eight words of a
sentence from the definition and saying it is the
definition. See Chief Justice
Abrahamson's dissent, ¶148;
Justice Bradley's dissent, ¶214. The punctuation denotes the end of the
sentence, not the end of the definition. Even within
¶117 It
was hardly "plainly erroneous or inconsistent with the regulations"——the
standard we apply here concerning interpretation of these administrative
rules——for the Commission to interpret the rules by reading the full text of
the subsection and examining and applying each factor to the facts of this
case. The court of appeals rightly noted
in its opinion,
As the commission pointed out, Wis. Admin. Code § Tax 11.71(1)(e) requires that the commission consider "all facts and circumstances" in determining whether a computer program comports with the definition of a custom program, including factors 1.—7. of the tax rule. We observe that the plain language of § Tax 11.71(1)(e) imposes this requirement. . . . Aside from its narrow and unreasonable interpretation of § Tax 11.71(1)(e), the DOR offers no reason for why we should ignore the commission’s determination that the tax rule shall be considered in its entirety in determining whether a computer program is customized.
DOR v. Menasha Corp., 2007 WI App 20, ¶49, 299 Wis. 2d 348, 728 N.W.2d 738.
¶118 Controlling
weight deference (similar to great weight deference applied to statutory
interpretation (DaimlerChrysler, 299 Wis. 2d 1, ¶15)) is appropriate
here, because the Commission reasonably interpreted a rule adopted by the DOR,
the Commission’s interpretation was not inconsistent with the rule's language
or clearly erroneous, and the Commission was charged by the legislature with
the responsibility of reviewing decisions of the DOR.
¶119 Finally,
under controlling weight deference, proposing an equally reasonable alternative
interpretation of the administrative rules, or even an alternative
interpretation that is more reasonable (Caterpillar, 241
¶120 The
statutory interpretation necessary in this case is at best perfunctory. The relevant statutory language merely
establishes that an exemption from tax exists for "custom computer
programs." Under the statute, "'[t]angible personal property,'"
i.e., taxable property, "includes . . . computer programs except custom
computer programs." Wis. Stat. § 77.51(20) (emphasis added). Tax liability or lack thereof is certainly the
heart of this case, but the dispute here is not centered on whether custom
computer programs are exempted from tax under the statute. The parties agree on that. Thus, while it is necessary to acknowledge the
statutory framework, the statute does not contain any language that is helpful
to the resolution of the disputed matter. As counsel for the DOR said in oral argument
before this court, "The outcome of this case depends entirely upon the
application of an administrative rule promulgated by the department. That rule determines whether or not a software
sale by a software vendor is subject to sales and use tax."
¶121 To
the extent that statutory interpretation comes into play it does so because
"[w]hen an administrative agency promulgates regulations pursuant to a
power delegated by the legislature, we construe those regulations 'together
with the statute to make, if possible, an effectual piece of legislation in
harmony with common sense and sound reason.'" DaimlerChrysler, 299
¶122 The
majority discusses the three levels of deference——great weight, due weight, and
no deference——accorded an agency's interpretation of a statute, as well as the
circumstances under which each is appropriate.
Majority op., ¶¶47-50. I note
that the majority applies due weight deference to the Commission's statutory
interpretation, and I see no reason to take issue with that holding, since I
believe such a determination is not required to resolve this case. I recognize, however, that "[w]hen
applying due weight deference, we will not overturn a reasonable agency
interpretation that is consistent with the purpose of the statute, unless there
is a more reasonable interpretation." DaimlerChrysler, 299
¶123 Because
the Commission's decision was consistent with the purpose of the statute——to
exempt custom computer programs from tax——and because I do not think a more
reasonable interpretation was set forth by the DOR, there certainly is no
reason to overturn the decision based on the Commission's interpretation of the
statute.
¶124 As
Chief Justice Abrahamson's dissent notes, the tax code is designed to ensure
that taxpayers pay what they owe and no more.
Chief Justice Abrahamson’s dissent, ¶127. What is owed is determined by
¶125 For
these reasons, I respectfully concur.
¶126 I
am authorized to state that Justices DAVID T. PROSSER and PATIENCE DRAKE
ROGGENSACK join this concurrence.
¶127 SHIRLEY
S. ABRAHAMSON, C.J. (dissenting). Each taxpayer should pay the taxes that he or
she owes under the tax laws--no more, no less.
The majority opinion rewrites the plain language of the governing law
and creates a new exemption in the tax laws that the legislature did not see
fit to enact.
¶128 The
majority opinion states that its decision "has great import to the average
taxpayer in this state."[29] I agree.
The fiscal implications of the new tax exemption created by the majority
opinion are substantial: According to the Legislative Fiscal Bureau's January
30, 2007, revenue and expenditure projections, the state's projected loss in
revenue as a result of the erroneous decision in the present case will exceed
$277.6 million prior to the end of the 2007-09 biennium and $28.3 million
annually thereafter.[30]
¶129 A
court should not effectively override an enacted tax statute by imposing the
court's views of economic policy or of the wisdom of a tax law. With this principle in mind, I examine the
text of the applicable statute and administrative rule.
¶130 The
legal question presented, and the facts, are really rather simple when stripped
to their essence.
¶131 Here's a snapshot of the law:
¶132 A
custom computer program is defined by the Department of Revenue's rule, Wis.
Admin. Code § Tax 11.71(1)(e), (k), and (m), which everyone agrees governs
the instant case. According to the
administrative rule, the words "custom
computer program" mean "utility and application software which
accommodate the special processing needs of the customer."[33] The administrative rule also provides that custom programs do not include
"prewritten programs" (sometimes also called "canned
programs"),[34]
which are "programs prepared, held or existing for general use normally
for more than one customer . . . ."[35] A program is either a (taxable) prewritten
program or a (tax-exempt) custom program; it cannot be both.[36]
¶133 Here's a snapshot of the facts: In 1995 SAP leased its R/3 System to Menasha
Corporation. The R/3 System consists of
some 70 software modules, each providing a rudimentary business and accounting
software system for a customer's business.
The R/3 System is not designed specifically for any customer. R/3 is a product that SAP leases in
unmodified form to many customers. In
other words, the R/3 System is a platform that an individual business uses for
creating its own customized business and accounting software. SAP's customers must undertake their own
modification of the R/3 platform to customize it for their own business
needs.
¶134 By
1998 SAP had leased this same software to more than 20,000 customers across the
world.[37]
In 1997 Business Week noted that "SAP's R/3 runs the back
offices of half of the world's 500 top companies——scheduling the manufacture of
washing machines at General Electric Co. and shipping soda pop on time at
Coca-Cola Co."[38]
¶135 The
leasing arrangement between SAP and its customer (here Menasha Corporation) is
separate and distinct from any arrangements the customer thereafter makes with
any entity for modifying the R/3 System to meet the customer's particular
software needs.
¶136 The
lease of the R/3 software by SAP to Menasha Corporation for $5.3 million (which
did not oblige SAP to provide customization services) is the basis for the tax
in the present case. Relevant to the
instant case, the Department of Revenue is not imposing a tax on expenditures
of more than $16 million that Menasha Corporation made over the next seven
years to customize the R/3 System.
¶137 In
sum, the R/3 System was not written solely for Menasha Corporation or upon its
request. SAP had developed the R/3
System before meeting up with Menasha Corporation and has leased the same R/3
System to thousands of customers.
¶138 The
majority opinion agrees that "the R/3 System exists and can be sold to
everyone in the same form."[39]
¶139 Here's the legal question of statutory
interpretation presented: Is the R/3 System of 70 modules leased by SAP to
Menasha Corporation in 1995 a taxable "computer program" or a
tax-exempt "custom computer program" as these words are defined in
the administrative rule?
¶140 Here's a snapshot of the majority opinion's
answer to the legal question: The majority opinion hides behind the
decision of the Tax Appeals Commission.
It cannot hide. It bears full
responsibility for the result.
¶141 Interpretation
of a statute or a rule is a question of law determined by this court
independently of other courts or administrative agencies. In interpreting a statute or an agency rule,
a court may, but need not, accord an agency's interpretation of a statute
deference or weight. The court reserves
its authority to interpret the law.
Indeed, it is the court's responsibility to decide questions of law and
to determine whether deference is due and what level of deference is due to an agency
interpretation and application of a statute.[40] By granting deference or weight to an agency interpretation, the
court does not, and should not, abdicate its authority and responsibility to
decide questions of law.[41]
¶142 Furthermore,
deference or weight is due to an administrative agency only when the agency's
interpretation is consistent with the language, meaning, and purpose of the
statute or rule and is reasonable.[42] The court itself must always interpret the
statute to determine the reasonableness of the agency determination.[43]
¶143 In
the present case, the Tax Appeals Commission's interpretation of the statute
and rule governing computer software and custom computer software is
inconsistent with the language, the meaning, and the purpose of the statute and
administrative rule and is not reasonable.
The Tax Appeals Commission's determination therefore cannot be entitled
to any deference or weight as this court determines the question of law
presented.
¶144 Nevertheless,
the majority opinion hides behind the Tax Appeals Commission by spending many pages uselessly, confusingly,
and often erroneously discussing whether the Department of Revenue's
interpretation of its own rule or the Tax Appeals Commission's interpretation
of the Department's rule should be given some kind of weight as the court
determines the question of statutory interpretation presented.[44] All of this discussion in the majority
opinion is beside the point, given that the Commission's interpretation is, in
any case, unreasonable and should be given no deference at all.
¶145 The
majority opinion does not take seriously its duty to render its own
interpretation of the administrative rule and to scrutinize the reasonableness
of the Commission's interpretation.
Although conceding numerous imperfections in the Tax Appeals
Commission's Ruling and Order,[45]
the majority opinion does its best to paper over, or to ignore, the
Commission's obvious failures to remain faithful to the clear text of Wis.
Admin. Code § Tax 11.71(1). The
majority opinion abandons its responsibility as the ultimate authority to
decide issues of law, perfunctorily and unpersuasively concluding that the
decision of the Tax Appeals Commission is consistent with the administrative
rule and is reasonable.
¶146 Here's a snapshot of the Tax Appeals
Commission's erroneous reasoning (upon which the majority opinion erroneously
rests): The Tax Appeals Commission makes two critical errors in
interpreting and applying Wis. Admin. Code § Tax 11.71(1).
¶147 First,
and most importantly, the Tax Appeals Commission utterly disregards the
administrative rule's definition of nontaxable "custom programs" as
"utility and application software which accommodate the special processing
needs of the customer." Nowhere in
its opinion does the Tax Appeals Commission apply the rule's definition to the
facts of the present case. The
Commission's opinion effectively repeals the administrative rule's clear
language setting forth this definition.
¶148 The
R/3 System that SAP leased to Menasha clearly does not fit within the rule's
definition of "custom programs."
Everyone agrees that the R/3 System that Menasha Corporation acquired
from SAP did not "accommodate the special processing needs of"
Menasha. Everyone agrees that the R/3
System is leased in unmodified form to many customers and that the R/3 System
does not meet the special processing needs of any particular customer. The Tax Appeals Commission and the majority
opinion could not possibly have reached a contrary conclusion if they had
honored the text of the administrative rule defining "custom
programs."
¶149 Second,
the Tax Appeals Commission misconstrues the administrative rule's definition of
taxable "prewritten programs," adding words and ideas to the plain
language of the rule that cannot be found in the rule's text. Prewritten programs are defined in relevant
part as "programs . . . existing for general
use." The Commission erroneously
converts the words "existing for general use" in the administrative
rule to read "ready for use off the shelf without any modification."
¶150 The
essence of the Tax Appeals Commission's decision is that because Menasha
Corporation had to customize the R/3 System after the Corporation acquired it,
the R/3 System is a custom computer program, not a prewritten program existing
for general use. According to the Tax
Appeals Commission, the distinction between custom computer programs and
prewritten programs hinges on the amount of effort necessary to get the
software operational for a particular customer's needs. The Commission concludes that the R/3 System
does not "exist for general use," because it is, in the Commission's
view, "useful only after a significant investment of resources in
planning, testing, training, enhancement, and maintenance . . . ."[46]
¶151 The
plain language of the administrative rule does not support the Tax Appeals
Commission's interpretation. The
administrative rule does not define a prewritten program as one that requires
no additional investment from the customer after its acquisition. The rule requires only that a prewritten
program is "prepared, held or existing for general use," not that the
program is ready for use right off the shelf without any modification. The phrase "for general use" does
not include any mention of the amount of time necessary to get the program up
and running for the individual customer.
¶152 Nor
does the fact that the R/3 System is not ready for use off the shelf without
any modification mean that the program is not of use or useful in the
unmodified state in which SAP leases the program to its customers. Of course the unmodified R/3 System is used
and is useful. The R/3 System is used
and useful as, in the Tax Appeals Commission's own words, "a rudimentary
business and accounting computer software system[.]"[47] SAP's customers use this rudimentary business
and accounting computer software system to advance their abilities to build the
business and accounting programs that will meet their own particular needs. SAP's customers do not, as the Commission
implies, pay millions of dollars to acquire something that has no use.
¶153 When
the administrative rule's definition of "prewritten program" is
applied as written, not as rewritten by the Tax Appeals Commission and the
majority opinion, the R/3 System that SAP leased to Menasha clearly fits within
the definition of "prewritten program." The R/3 System is an existing program that
Menasha Corporation acquired in unmodified form, not a program that SAP created
or modified at Menasha Corporation's request.
The R/3 is suitable for general use; it is used by many customers other
than Menasha Corporation. The program
thus is a prewritten program "existing for general use normally for more
than one customer."
¶154 In
sum, the Tax Appeals Commission ignores the administrative rule's definition of
"custom programs" and rewrites the rule's definition of
"prewritten programs." The
majority opinion seriously errs in adopting the Commission's interpretation as
its own, when the Commission's interpretation is so plainly inconsistent with
the rule's language, meaning, and purpose.
¶155 This
court should, in my opinion, follow the excellent memorandum decision of Judge
Steven Ebert of the
¶156 I
shall first set forth the undisputed facts, then the applicable administrative
rule, and finally I shall apply the statute and rule to these undisputed facts
in greater detail than presented in the snapshots above.
I
¶157 SAP
leased its R/3 software to Menasha Corporation for a license fee of
approximately $5.3 million.[48] SAP's lease of its R/3 software to Menasha
Corporation was routine in nature and similar to the circumstances of its R/3
sales to its other customers.[49]
¶158 The
R/3 System is a software product that SAP leases in unmodified form to many
customers.[50] SAP's leases of R/3 System software are sales
of "off-the-shelf" standardized software; the program is written before
the sale and is sold to a wide variety of customers.[51] SAP does not modify the existing software
modules before shipment to customers.[52] SAP and Menasha Corporation's agreement made
no provision for customization of the R/3 System software by SAP.[53] The customer must implement the R/3 System on
its own or hire SAP consultants or SAP approved independent consultants.[54] SAP keeps the lease agreement separate from any agreement
relating to consulting and maintenance services.[55]
¶159 In
other words, the R/3 System is not designed specifically for any customer. SAP itself agreed in a Wisconsin Department
of Revenue audit that R/3 is subject to
¶160 The
R/3 System is an example of Enterprise Resource Planning (ERP) software, that
is, "generalized, integrated software that [can] be customized for
virtually any large business."[58] The R/3 System contains a basic business and
accounting system that did not meet Menasha Corporation's (or any entity's)
particular needs for a business and accounting software system.[59] SAP markets the R/3 System not as software
that is customized to the needs of SAP's customers but rather as
software that is customizable to the customer's needs.
¶161 Like
SAP's other customers, Menasha Corporation acquired the R/3 System from SAP to
customize it to meet Menasha Corporation's needs.[60] After licensing the R/3 System from SAP,
Menasha Corporation expended a large sum of money over seven years to customize
the software to fit its business needs.
All told, Menasha Corporation paid approximately $16,275,000 in
consulting fees for the purpose of customizing the R/3 software: $13 million to
ICS Deloitte (an SAP "logo partner" that Menasha Corporation elected
to hire); $2.5 million to SAP; and $775,000 to consultants associated with
neither ICS Deloitte nor SAP.[61] SAP was one of at least five different groups
of personnel involved in the implementation team that modified R/3 to meet
Menasha Corporation's needs.
¶162 The
Department of Revenue imposed $265,093 in sales or use taxes on the $5.3
million that Menasha Corporation paid to SAP for the lease of the R/3 software.[62] It is this tax that is at issue in the
present case.[63] Menasha Corporation's refund claim pertains
solely to the $265,093 in sales or use tax payments arising from Menasha
Corporation and SAP's $5.3 million transaction for the acquisition of SAP's R/3
software. The present case has nothing
to do with the nontaxable $16,275,000 in consulting fees that Menasha
Corporation paid in various transactions over the course of seven years for the
purpose of customizing the R/3 software.
¶163 Those
are the undisputed facts. I turn to the
law and its application to the undisputed facts.
II
¶164 The
Department of Revenue promulgated Wis. Admin. Code § Tax 11.71 under its
rule making authority.[64] Section 11.71, entitled "Computer
industry," restates (among other matters)
¶165 Section
11.71(1)(e), (k) and (m) of the
(e) "Custom programs" mean utility
and application software which accommodate the special processing needs of the
customer. The determination of whether a program is a custom program shall
be based upon all the facts and circumstances, including the following:
1. The
extent to which the vendor or independent consultant engages in significant
presale consultation and analysis of the user's requirements and system.
2.
Whether the program is loaded into the customer's computer by the vendor and
the extent to which the installed program must be tested against the program's
specifications.
3. The
extent to which the use of the software requires substantial training of the
customer's personnel and substantial written documentation.
4. The
extent to which the enhancement and maintenance support by the vendor is needed
for continued usefulness.
5.
There is a rebuttable presumption that any program with a cost of $10,000 or
less is not a custom program.
6. Custom programs do not include basic
operational programs or prewritten programs.
7. If
an existing program is selected for modification, there must be a significant
modification of that program by the vendor so that it may be used in the
customer's specific hardware and software environment.
. . . .
(k) "Prewritten programs", often
referred to as "canned programs", means programs prepared, held or
existing for general use normally for more than one customer, including
programs developed for in-house use or custom program use which are
subsequently held or offered for sale or lease.
. . . .
(m) . . . For purposes of this section a program is
either a prewritten or custom program.[66]
¶166 When
I apply the administrative rule to the undisputed facts, it is clear that the
R/3 System is excluded from being a custom computer program under the very
first sentence of the rule.
¶167 The
first sentence of Wis. Admin. Code § Tax 11.71(1)(e) defines the term
"custom programs" to "mean utility and application software
which accommodate the special processing needs of the customer." The Tax Appeals Commission recognizes that
the administrative rule's first sentence defines the term "custom
programs." Quoting directly from
the rule's first sentence, the Commission states that "[s]ection TAX
11.71(1)(e)(intro) defines
custom programs 'as utility and application software which accommodate the
special processing needs of the customer.'"[67] The Tax Appeals Commission is obviously
correct that the administrative rule's first sentence defines "custom
programs." The first sentence
begins, "'Custom programs' mean . . . ."
(emphasis added). The verb
"mean" denotes definition.[68]
¶168 Although
claiming to defer as a general matter to the Tax Appeals Commission's
interpretation of the administrative rule, the concurrence apparently does not
defer to the Commission's conclusion that the rule's first sentence defines
"custom programs."[69] The concurrence also does not explain what
the rule's first sentence means, or what function it serves, if it does not
define "custom programs."
¶169 Everyone
agrees that the R/3 System that Menasha Corporation acquired from SAP did not
"accommodate the special processing needs of" Menasha Corporation.[70] Software that is leased in unmodified form to
many customers and that does not provide adequate processing for any particular
customer obviously does not accommodate the customer's special processing
needs. Indeed Menasha Corporation had to
expend more than $16 million over seven years to fit the R/3 System to
Menasha's special processing needs.
¶170 The
Tax Appeals Commission completely disregards the administrative rule's first
sentence defining "custom programs."
In its opinion, the Commission neither acknowledges nor explains its
failure to apply the definition set forth in the rule. It appears that the Commission's error was
inadvertent, not the result of a conscious determination that the
administrative rule's definition of "custom programs" somehow is not
controlling in the instant case. The
Commission just seems to have forgotten that the definition is there.
¶171 The
majority, however, certainly is aware that the definition is there. The Department of Revenue argues to this court
that the rule's first sentence defining "custom programs" must be
applied in deciding the instant case.
¶172 The
Department of Revenue, for example, states in a brief to this court (using the
words of the rule's first sentence) that the outcome in the present case
depends upon "whether the software sold under the sales transaction with
the seller accommodates the special processing needs of the
customer . . . ."[71] In the same brief, the Department further
argues that "[a]s purchased by Menasha under its software sales contract
with SAP, R/3 was not 'custom' because it did not 'accommodate the special
processing needs of the customer' as required by subsection (e) of DOR's
rule."[72] Again the Department relies on the first
sentence.
¶173 During
oral argument to this court, the Department of Revenue again stated its
position that the court must apply the administrative rule's first sentence
defining "custom programs" in determining whether the R/3 System is a
"custom" program for purposes of the rule. Counsel for the Department of Revenue made
the following statement to this court:
[T]he rule says that "custom software" is software that accommodates the specific processing needs of the customer. The R/3 modules . . . as purchased by Menasha are not "custom software" within that definition because as purchased the modules did not accommodate the specific processing needs of Menasha. Menasha hired ICS so to customize the software so that it would accommodate Menasha's specific processing needs. . . .
When asked where in the rule counsel found this definition of "custom software," counsel referred the court to the first sentence of the rule.
¶174 Although
the majority is aware that Wis. Admin. Code § 11.71(1)(e) defines the term
"custom programs" in its first sentence, and although the majority is
also aware that the Department of Revenue relies upon the definition set forth
in the first sentence, the majority opinion does not apply the rule's
definition of "custom programs"; does not explain why it does not
apply the rule's definition of "custom programs"; and does not
acknowledge the Department's argument that the rule's definition of
"custom programs" should be applied.
¶175 Instead
of applying the definition as written, the majority opinion relies exclusively
upon the rule's second sentence, which states that the determination of whether
a program is "custom" shall be based on all the facts and
circumstances. The majority opinion
decides the present case simply by "weighing" the relevant facts and
circumstances, as if Wis. Admin. Code § Tax 11.71(1)(e) did not define
"custom programs" but instead merely set forth some sort of amorphous
balancing test with no bottom line. The
majority opinion concludes that the facts and circumstances "weigh in
favor" of holding that the R/3 System that SAP leased to Menasha
Corporation is a custom program. The
majority opinion nowhere determines whether the facts and circumstances show
that the R/3 System fits within the rule's definition stating that
"'[c]ustom programs' mean utility and application software which
accommodate the special processing needs of the customer."
¶176 The
Tax Appeals Commission and the majority opinion clearly err in disregarding the
administrative rule's first sentence defining "custom programs" and
in relying exclusively on the rule's second sentence stating that the
determination whether a program is custom shall be based on all the facts and
circumstances. The rule's second
sentence supports rather than displaces the definition set forth in the first
sentence, instructing the court what factors to consider when determining
whether a program accommodates the special processing needs of the customer.
¶177 For
example, when a vendor enters into a transaction to sell a standardized
software package but also agrees to modify the software for the customer (thus
giving the software elements of both a prewritten program and a custom
program), the software initially meets the administrative rule's definition of
a custom program. Then the factors set
forth in the rule's second sentence must be examined to determine whether,
under the totality of the circumstances, the program is a tax-free
"custom" program accommodating the special processing needs of the
customer. The second sentence, however,
does not come into play when the program obviously is excluded from the
definition of a custom program under the first sentence.
¶178 The
rule's first sentence defining "custom programs" clearly provides the
rule's bottom line. Put another way,
although the application of the rule need not end with the first sentence in
every case, it does begin with it. The
dissent examines and applies both the first and second sentences of the
rule. The majority opinion erroneously
attends only to the second sentence and entirely ignores the first in deciding
the instant case.
¶179 I
could end the dissent right here. The
plain language of the first sentence of Wis. Admin. Code § Tax 11.71(1)(e)
leaves no doubt that the present case must be resolved in favor of the
Department of Revenue. I shall, however,
proceed to the administrative rule's second sentence and to the list of seven
factors enumerated within the rule's text.
The rule's second sentence and enumerated factors reinforce the rule's
first sentence defining "custom programs."
¶180 The
sixth enumerated factor, Wis. Admin. Code § Tax 11.71(1)(e)6., shows that
the R/3 system cannot be a custom computer program. The sixth factor states that "[c]ustom
programs do not include basic operational programs or prewritten
programs." No one argues that the
R/3 System is a basic operational program.
Rather the dispute is whether it is a prewritten program. Prewritten programs are defined in Wis.
Admin. Code § Tax 11.71(1)(k) as "programs prepared, held or existing for general use normally for
more than one customer."[73] Furthermore, § Tax 11.71(1)(m) states
that a prewritten program for purposes of § 11.71 is either a prewritten
or custom program.
¶181 The
language of Wis. Admin. Code § 11.71(1)(e)6. and (m) make clear that if
the R/3 System is a "prewritten program," it cannot qualify as a
custom computer program. The Tax Appeals
Commission understood that the rule's sixth factor has this meaning, stating
that factor six "is as much a veto as a factor"[74]
and that "[t]his case hinges on whether the R/3 System is a prewritten
program."[75] The Commission's proper interpretation of the
plain language of § Tax 11.71(1)(e)6. as a veto is further supported by
Wis. Admin. Code Tax § 11.71(1)(m), which provides that for purposes of
§ Tax 11.71, "a program is either a prewritten program or a custom
program," not both.
¶182 Although
conceding that the R/3 System is "existing rather than created"[76]
and is leased in identical form to a large number of customers, the Tax Appeals
Commission and the majority opinion shockingly conclude, contrary to the very
words of the rule, that the R/3 System does not exist for general use by more
than one customer.
¶183 The
Tax Appeals Commission reaches this conclusion in disobedience of the plain
language of the administrative rule, explaining that the only "use"
of the R/3 System was its use after extensive post-sale modification had
customized the software to Menasha Corporation's particular business
needs. As the majority opinion states,
the Tax Appeals Commission reasons that the R/3 System does not exist for
general use because the software is "useless until modified."[77]
¶184 According
to the Tax Appeals Commission, "[t]he distinction between custom and
prewritten programs hinges on the amount of effort necessary to get the
software operational for a particular customer's needs."[78]
¶185 Put
another way, the Tax Appeals Commission concludes that SAP's sophisticated
business customers pay millions of dollars to acquire software that has no use
and is useless in the form purchased.
Nonsense! Menasha Corporation
paid $5 million for software that it knew had a use and that it knew would be
useful to it: Menasha used, and paid a lot of money to use, the R/3 System as a
rudimentary business and accounting software system that advanced Menasha
Corporation's ability to build its own business and accounting software system that
met Menasha's Corporation's particular needs.
¶186
The Tax Appeals Commission's decision rests on converting the words
"existing for general use" in the administrative rule's definition of
prewritten program to the words "ready for use off-the-shelf" without
any modification[79]
It is unreasonable to interpret the words "existing for general
use" as "ready for use off-the-shelf" without any modification. Reasonably construed, the words
"existing for general use" mean just what they say: that the program
exists (as opposed to a program that must be created at the customer's order)
and is suitable for general use (as opposed to the special use of one
customer). The text of the
administrative rule says nothing about whether a program is "ready for use
off-the-shelf" without any modification.
¶187 The
R/3 System that SAP leased to Menasha clearly falls within the administrative
rule's definition of prewritten program.
Because factor six (and (1)(m)) act as a "veto," rendering
prewritten programs and custom programs mutually exclusive, the R/3 System
cannot be a custom program under Wis. Admin. Code § Tax 11.71(1)(e).
¶188 The
Tax Appeals Commission concluded that although "SAP admitted that the R/3
software is prewritten and subject to the sales and use tax," and although
SAP "backed up" this admission by paying taxes on past sales of the
R/3 software and agreeing to collect sales and use taxes on future sales, SAP's
admission and actions have "no value" in the instant case and are not
"probative."[80] The Tax Appeals Commission's position that
SAP's admission and actions have no value and are not probative is not reasonable. How can the seller's agreement that its own
software is taxable as "prewritten" software be without any probative
value?
¶189 In
sum, the Tax Appeals Commission's interpretation and application of the
administrative rule's sixth factor is unconvincing for several reasons.
¶190 First,
nothing in the words of the administrative rule supports the Tax Appeals
Commission's interpretation of Wis. Admin. Code § Tax 11.71(1)(e)6.
changing the words "general use" to read "ready for use
off-the-shelf" without any modification.
The words "ready for use off-the-shelf" without any
modification are conspicuously absent from the text of Wis. Admin. Code
§ Tax 11.71(1)(k). The
administrative rule's text does not require that the "use" of a
prewritten program must be use of the program off the shelf and without
modification.
¶191 Second,
the Tax Appeals Commission's interpretation of "prewritten program"
in Wis. Admin. Code § Tax 11.71(1)(k) as ready for use off the shelf without
any modification is not only contrary to the words of the rule but also
contravenes the history of the administrative rule.[81] The history behind the rule shows that the
majority opinion rewrites the text of Wis. Admin. Code § Tax 11.71(1)(k)
to include the very language that the Department of Revenue expressly refused
to put in the rule at the time of the rule's promulgation.
¶192 The
computer industry apparently lobbied the Department during the promulgation of
Wis. Admin. Code § Tax 11.71 to define "prewritten programs" as
programs that are ready for use off the shelf.
When the Department was preparing the definition of "prewritten
programs" in Wis. Admin. Code § Tax 11.71(1)(k), an attorney
representing the computer industry (the very same attorney, as it happens, who
represented Menasha Corporation before the Tax Appeals Commission in the
present case) lobbied the Department to define "prewritten programs"
as programs intended for general use "and mass distribution as
prepackaged ready-to-use programs" (emphasis added).[82]
¶193 The
Department expressly denied the attorney's request, stating that it would stick
with "the prewritten program definition we originally used."
¶194 The
Department's refusal to define "prewritten programs" as
"prepackaged ready-to-use programs" stemmed from its longstanding
position that prewritten programs need not be ready for use off-the-shelf. In two "technical information
memoranda" dated April 2, 1976, and August 7, 1978, the Department of
Revenue defined the term "prewritten (canned) programs" as
"programs prepared, held or existing for general or repeated use,
including programs developed for in-house use and subsequently held or offered
for sale or lease."[83] The technical information memoranda
elaborated upon this definition by explaining that "prewritten (canned)
programs" need not be ready for use off-the-shelf. The memoranda stated that "[i]n some
cases [prewritten programs] are usable as written" but that "in most
cases it is necessary that the program be modified, adapted and tested to meet
the customer's particular needs."[84]
¶195 Third,
the Tax Appeals Commission's decision yields unreasonable and absurd
results. In its decision the Commission
considers the Menasha Corporation's seven-year history working with the R/3
System in determining that the R/3 System was not a prewritten program because
it was "useful only after a significant investment of resources in
planning, testing, training, enhancement, and maintenance . . . ."[85] Yet the seller and buyer need to determine
the taxability of the transaction in the year the agreement is completed. At the time of the agreement, the vendor and
customer may not know what future steps the customer may take to customize the
software to its own needs, or how much the customer may invest relative to the
purchase price.
¶196 Fourth,
the Tax Appeals Commission's interpretation of Wis. Admin. Code Tax
§ (1)(e)6. reduces the factor to mere surplusage. Thus the Tax Appeals Commission bases its
interpretation of factor six largely on the rule's first four factors, each of
which (according to the Commission) "hinges on the degree to which the
software is ready for use off-the-shelf."[86] Why conclude that factor six seeks to measure
the same attributes of the software program that the other factors
measure? The plain language of
enumerated factor six does not refer to the other five preceding factors. Each factor in the definition of "custom
program" is independent of the others, and each must be interpreted
individually based on the plain language of the rule.
¶197 Factor
six states simply that "custom program" does not include a prewritten
program. The other enumerated factors do
not influence the plain language of enumerated factor six. By interpreting factor six to mean nothing
more than the combined meaning the first four factors, the Tax Appeals
Commission reduced factor six to surplusage.
¶198 Again,
I could end the dissent here. Like the
first sentence of Wis. Admin. Code § Tax 11.71(1)(e) defining "custom
programs," the rule's definition of "prewritten programs" and
the veto provision in the rule's sixth factor clearly demonstrate that the R/3
System is not a custom program. Nevertheless,
I will address two other errors in the Tax Appeals Commission's decision that
render the decision in conflict with the plain text of the administrative rule.
¶199 The
Tax Appeals Commission errs in interpreting enumerated factor two, which has
two parts: (1) "[w]hether the program is loaded into the customer's
computer by the vendor" and (2) "the extent to which the installed
program must be tested against the program's specifications."[87]
¶200 With
regard to the first part, the Tax Appeals Commission found that "the fact
that a former SAP employee loaded the [R/3] software weighs in favor of a
finding that the software at issue is custom software."[88] The Commission's finding ignores the plain
language of the rule, which requires that the program is loaded "by the
vendor" (emphasis added). A
former employee of the vendor is not the vendor.
¶201 The
language of the rule is clear, and the Tax Appeals Commission's interpretation
and application of factor two is inconsistent with the language of the
administrative rule. The Commission did
not apply the "by the vendor" language literally. The Commission's interpretation is
unreasonable.
¶202 Finally,
the Tax Appeals Commission does not account for the seventh factor. The seventh factor states, "If an
existing program is selected for modification, there must be a significant
modification of that program by the vendor so that it may be used in the
customer's specific hardware and software environment."
¶203 The
Tax Appeals Commission concluded that the reference in factor seven to
"'existing program' means an 'existing program for general use' as that
phrase is used in the definition of 'prewritten programs . . . .'"[89] The Commission then determined that, because
the Commission had already concluded that the R/3 System is not a prewritten
program, "this factor does not come into play."[90] According to the Tax Appeals Commission, the seventh
factor comes into play only if the software is prewritten.
¶204 The
Tax Appeals Commission's reasoning is odd.
As counsel for Menasha Corporation has explained, the Commission's
decision does not apply factor seven literally.[91] Furthermore, the Commission's conclusion that
factor seven comes into play only when a program is prewritten implies that
factor seven is surplusage. As the
Commission itself seemed to recognize, the "veto" provision in factor
six states that any program which is prewritten cannot be a custom
program. If factor seven comes into play
only when a program is prewritten, it thus has no effect.
¶205 In
sum, I agree with the circuit court. I
would apply Wis. Admin. Code § Tax 11.71(1)(e), (k), and (m) according to
their clear text and would conclude that the R/3 System that SAP leased to
Menasha is not a "custom program" as that term is defined in the text
of the rule. It is undisputed that the
R/3 software that SAP leased to Menasha Corporation did not "accommodate
the special processing needs of the customer." Rather, the exact same package of 70 software
modules was sold to thousands of different customers regardless of each
customer's individual processing needs.
The R/3 System is a "prewritten program" already existing for
general use, not a program created or customized for Menasha's special
use. Because the R/3 System was not
customized, Menasha Corporation had to enter into separate service contracts,
subsequent to licensing the software, to customize the software to meet Menasha
Corporation's special processing needs.
¶206 To
reach a result favorable to Menasha Corporation, the Tax Appeals Commission and
the majority opinion do violence to the plain language of Wis. Admin. Code Tax
11.71(1). The Tax Appeals Commission and
majority opinion reach their holding in the only way they can: by disregarding
the definition of "custom programs" set forth in Wis. Admin. Code
§ Tax 11.71(1)(e) and by rewriting the definition of "prewritten
programs" set forth in Wis. Admin. Code § Tax 11.71(1)(k).
¶207 For
the reasons set forth, I dissent.
¶208 I
am authorized to state that Justices ANN WALSH BRADLEY and LOUIS B. BUTLER, JR.
join this dissent.
¶209 ANN WALSH BRADLEY, J. (dissenting). Here's the $300 million[92] question: did the R/3 computer program "accommodate the special processing needs" of Menasha? That's the question that the first sentence of Wis. Admin. Code § TAX 11.71(1)(e) requires that we ask. The Commission doesn't answer it, the majority doesn't answer it, and the concurrence doesn't answer it.
¶210 Not one of them answers the question in the affirmative. They can't because the 70 disks of the R/3 system did not accommodate the special processing needs of Menasha. Not one of them answers the question in the negative. They can't because doing so would undermine their conclusions. We therefore get three separate writings, totaling more than 100 pages of text, where no one addresses the very first sentence of a rule they purport to interpret.
¶211 I join the dissent of Chief Justice Abrahamson. I write separately, however, to address the issue of deference and to emphasize the failure of the Commission, the majority, and the concurrence to ask and answer this necessary question.
¶212 In a case where there are differing reasonable interpretations,
deference is often given. An interpretation is not reasonable, however, if it
ignores the language of the rule or if it is inconsistent with that language. Pfeiffer
v. Board of Regents of the
¶213 Instead of interpreting the language of Wis. Admin. Code § TAX 11.71(1)(e), the Commission altogether ignored the first sentence of the rule it purported to be interpreting. How can we defer to an interpretation of language when no such interpretation exists? The majority follows the Commission in failing to provide any analysis of the first sentence.
¶214 As a result of ignoring the first sentence of the rule, the agency
here has reached an interpretation that is inconsistent with that sentence.
This court gives no deference to an agency's interpretation of a rule that is
inconsistent with the language of that rule. Pfeiffer, 110
I
¶215 The issue here is the interpretation of Wis. Admin. Code § TAX 11.71(1)(e). The first part of the section states as follows:
"Custom programs" mean utility and application software which accommodate the special processing needs of the customer. The determination of whether a program is a custom program shall be based upon all the facts and circumstances, including the following . . . .
(Emphasis added.)
¶216 The very first sentence of the rule sets forth the definition of "custom programs," and the question in this case is whether the software here fits within the definition of a "custom program." However, the Commission never turns its attention to the definition.
¶217 We give agencies varying degrees of deference in interpreting their
own rules. See DaimlerChrysler v. Labor & Indus. Review Comm'n,
2007 WI 15, ¶15, 299
¶218 How can we give deference to a non-existent interpretation?
¶219 The answer, of course, is that we can't. We owe no deference whatsoever to an agency's failure to interpret a definition clearly set forth in the Wisconsin Administrative Code.
¶220 The concurrence attempts to circumvent this problem by claiming that the first sentence is merely part of the definition of custom programs. It accuses the dissents of taking "eight words of a sentence from the definition and saying it is the definition." Concurrence, ¶116 (emphasis in original). Instead, the concurrence argues that we should give controlling weight deference to the Commission's interpretation.
¶221 Yet the concurrence contradicts the Commission outright. The Commission explicitly determined that section 11.71(1)(e) defines "custom programs" in its first sentence. In its decision in this case, the Commission stated:
"Section TAX 11.71(1)(e)(intro) defines custom programs as 'utility and application software which accommodate the special processing needs of the customer.'" (Emphasis added.)
In other words, the Commission agrees with the dissents in this regard.[93]
¶222 Despite the concurrence's protests that it is giving deference to the Commission, its central argument overturns the Commission's explicit determination.
II
¶223 Not only does the Commission fail to analyze the most fundamental provision in the rule——the first sentence——it renders an interpretation that directly contradicts that unambiguous provision.
¶224 The Commission, as well as the majority and the concurrence, makes the mistake of interpreting the factors listed in section 11.71(1)(e) as the only things to consider when determining whether a computer program is custom. However, if one looks at the factors in conjunction with the definition of "custom programs," it is clear that the factors are based on the assumption that the program in question actually meets the needs of the customer. The factors must be examined in order to determine whether the needs met by the software are "special."
¶225 The factors operate within the premise set forth in the very first sentence of the rule: that software "accommodate[s] the special processing needs of the customer." They cannot operate to contradict it.[94] The concurrence's claim that the dissents render the rest of the rule "surplusage" misses the mark. The dissents are clear: the factors must be considered, but they cannot be deployed to contradict the definition set forth in the first sentence of the rule.
¶226 It is undeniable that the R/3 system purchased——the 70 disks containing
prewritten software——did not
"accommodate the special processing needs of the customer." The
processing needs of the customer, Menasha, were only accommodated after other
modifications were made.
¶227 Put
another way, the Commission interpreted section 11.71(1)(e) such that software
which does not accommodate the special needs of the customer is a
"custom program," which in turn is defined as software that does
accommodate the special needs of the customer. "Does not accommodate"
is the opposite of "does accommodate."
¶228 The
Commission's interpretation led it to a conclusion that software which does not
accommodate the special processing needs of the customer is "custom."
Such an interpretation is thus inconsistent with the plain words of the rule.
Because that's what the Commission's interpretation has done, we owe it no
deference.
¶229 Like the Commission and the concurrence, the majority utterly ignores the first sentence of rule. It does not provide a single word explaining how its conclusion comports with the requirement that custom programs "accommodate the special processing needs of the customer."
¶230 The closest the majority comes to such an analysis is its statement
that "the rule does not end with the first sentence." Majority op., ¶106. I agree——but it most assuredly
begins there.
¶231 Because
its analysis ignores the first sentence and renders an interpretation
inconsistent with the language of the rule, we owe no deference to the
Commission's interpretation. I therefore respectfully dissent.
¶232 I
am authorized to state that Justice LOUIS B. BUTLER, JR. joins this
dissent.
[1] DOR v. Menasha Corp., 2007
WI App 20, 299
[2] The
[3] All subsequent references to the Wisconsin Administrative Code are to the September 2006 version unless otherwise indicated.
[4] All subsequent references to
the Wisconsin Statutes are to the 2003-04 version unless otherwise indicated.
Wisconsin
Stat. § 77.51(20) provides:
"Tangible personal property" means
all tangible personal property of every kind and description and includes
electricity, natural gas, steam and water and also leased property affixed to
realty if the lessor has the right to remove the property upon breach or
termination of the lease agreement, unless the lessor of the property is also
the lessor of the realty to which the property is affixed. "Tangible personal property" also
includes coins and stamps of the
"Custom programs" mean utility and application software which accommodate the special processing needs of the customer. The determination of whether a program is a custom program shall be based upon all the facts and circumstances, including the following:
1. The extent to which the vendor or independent consultant engages in significant presale consultation and analysis of the user's requirements and system.
2. Whether the program is loaded into the customer's computer by the vendor and the extent to which the installed program must be tested against the program's specifications.
3. The extent to which the use of the software requires substantial training of the customer's personnel and substantial written documentation.
4. The extent to which the enhancement and maintenance support by the vendor is needed for continued usefulness.
5. There is a rebuttable presumption that any program with a cost of $10,000 or less is not a custom program.
6. Custom programs do not include basic operational programs or prewritten programs.
7. If an existing program is selected for modification, there must be a significant modification of that program by the vendor so that it may be used in the customer's specific hardware and software environment.
[5] At summary judgment before the Commission, Menasha
submitted its proposed Undisputed Material Facts. In response, the DOR agreed with the bulk of
Menasha's submissions and submitted additional proposed Undisputed Material Facts. On facts with which the parties agree, the
Commission treated the agreements as a stipulation of facts. Where disagreements arose, the Commission
addressed the differences and resolved any disagreements. Menasha Corp. v. DOR, Wis. Tax Rptr.
(CCH) ¶400-719, at 32,849-54 (WTAC 2003). At the court of appeals, the DOR argued that
there were genuine issues of material fact and thus summary judgment was
improperly granted. The DOR challenged
two undisputed facts at the court of appeals.
However, the court of appeals affirmed the Commission's conclusions and
stated, "we agree with the commission that the arguments the DOR raises in
relation to Undisputed Material Facts #29 and #47 are either unsupported by the
evidence or do not otherwise create a genuine issue of material fact." Menasha, 299
[6] While SAP agreed with the DOR——during the course of an audit——that the R/3 System was non-custom and thus subject to sales tax, the Commission found SAP's admission to be of "no value." The Commission concluded, "SAP is not a party to this proceeding, and, therefore, its admission is neither binding nor probative in this case. There may be any number of possible reasons why SAP elected to pay sales and use tax on its sale of the R/3 software, including reasons that may not reflect the merits of the issues raised in this case." We agree. SAP did not pay Menasha's taxes. SAP's agreement with the DOR has no precedential effect in this case. One's decision not to contest a matter is not binding on another who wishes to have its case heard and receive a full determination on the merits. Accordingly, we conclude that the Commission's conclusions were reasonable.
[7] This refund claim is based upon the tax that Menasha paid on the basic R/3 System that it purchased from SAP for $5.2 million.
[8] The Commission's Practice and Procedures is available at http://www.wisbar.org/AM/Template.cfm?Section=Legal_Research&Template=/CM/ContentDisplay.cfm&ContentID=32054 (Last visited 07/01/08).
[9] See
[10] While we review only the Commission's decision, we detail the circuit court's decision in the interest of completeness.
[11] The circuit court applied due weight deference to the Commission's interpretation of Wis. Admin. Code § Tax 11.71(1)(e). However, when giving deference to the Commission's interpretation of a rule, the proper deference is either controlling weight deference or a de novo review. Due weight deference is relevant only for an agency's interpretation of a statute. See ¶¶58-60 for the appropriate deference regarding rule interpretation.
[12] DOR v. Caterpillar, Inc.,
2001 WI App 35, 241
[13] While the court of appeals' decision stated that it applied the due weight deference standard to the rule interpretation, in reality it applied controlling weight deference. Due weight deference applies to statutory interpretation. Under rule interpretation, deference is either controlling weight deference or it is a de novo review. The court of appeals did apply the correct standard given that it concluded that the Commission's interpretation was reasonable and consistent with the plain language of the rule. See ¶¶58-60 for the appropriate deference regarding rule interpretation.
[14] In Racine Harley-Davidson, we stated:
By granting deference to agency interpretations, the court has not abdicated, and should not abdicate, its authority and responsibility to interpret statutes and decide questions of law. Some cases, however, mistakenly fail to state, before launching into a discussion of the levels of deference, that the interpretation and application of a statute is a question of law to be determined by a court. In any event, it is the court's responsibility to decide questions of law and determine whether deference is due and what level of deference is due to an agency interpretation and application of a statute. The court determines the appropriate level of deference by comparing the institutional qualifications and capabilities of the court and the agency by considering, for example, whether the legislature has charged the agency with administration of the statute, whether the agency has expertise, whether the agency interpretation is one of long standing, and whether the agency interpretation will provide uniformity and consistency.
Racine Harley-Davidson, Inc. v. Wis. Div. of Hearings & Appeals, 2006 WI 86, ¶14, 292 Wis. 2d 549, 717 N.W.2d 184.
[15] While Caterpillar
and River City Refuse Removal involve the interpretation of a statute
and not a rule, the Commission is still the final authority on all tax
questions in the state of
[16] Hillhaven Corp. v. DHFS,
2000 WI App 20, ¶12, 232
[17] State ex rel.
[18] See, e.g.,
Salvatore Massa, The Standards of Review for Agency Interpretations of
Statutes in
[19] Pfeiffer v. Bd. of Regents
of
[20] DOR v. River City Refuse
Removal, Inc., 2007 WI 27, ¶35,
299
[21] Wisconsin Stat. § 73.01(4)(e) provides in
relevant part:
The decision or order of the commission shall become final and shall be binding upon the petitioner and upon the department of revenue for that case unless an appeal is taken from the decision or order of the commission under s. 73.015. Except in respect to small claims decisions, if the commission construes a statute adversely to the contention of the department of revenue:
1. Except for hearings on ss. 341.405 and 341.45 and except as provided in subd. 2., the department of revenue shall be deemed to acquiesce in the construction so adopted unless the department of revenue seeks review of the order or decision of the commission so construing the statute. For purposes of this subdivision, the department of revenue has sought review of the order or decision if it seeks review and later settles the case or withdraws its petition for review or if the merits of the case are for other reasons not determined by judicial review. The construction so acquiesced in shall thereafter be followed by the department of revenue.
[22] Wisconsin Stat.
§ 73.015(2) provides:
Any adverse determination of the tax appeals commission is subject to review in the manner provided in ch. 227. If the circuit court construes a statute adversely to the contention of the department of revenue, the department shall be deemed to acquiesce in the construction so adopted unless an appeal to the court of appeals is taken, and the construction so acquiesced in shall thereafter be followed by the department.
[23] The DOR's arguments are in response to an order this court issued on March 21, 2008, asking the parties to discuss, "[w]hat impact, if any, do the acquiescence and non-acquiescence provisions of Wis. Stat. §§ 73.01(4)(e)1 and 73.015(2) (2003-04) have on the standard of review and degree of deference to be accorded the Department of Revenue and Tax Appeals Commission in the instant matter?"
[24] While factors one and two may hinge on the degree to which the software is ready to go off-the-shelf, that reasoning does not follow with factors three and four. In consideration of factor three, a very complicated non-custom program that is ready to go off-the-shelf may require substantial training even though it is ready for use. In consideration of factor four, whether programs need enhancement and maintenance has little to do with whether a program is ready for use directly off-the-shelf. While the Commission's language is somewhat perplexing, its reasoning is sound. In short, the Commission reasonably states that the more effort it takes to make a program usable, the more likely——even if not necessarily dispositive——that the program is custom, rather than canned or prewritten.
[25] When the DOR has
referenced the first sentence of the rule, it has been in tandem with the
remainder of the rule. See DOR briefs:
at 27 and 33 of its brief in chief, at 11-12 of its response to the non-party
brief, at 4 and 7 of its reply brief, DOR v. Menasha Corp., 2008 WI 88,
__ Wis. 2d __, __ N.W.2d __ (No.
2004AP3239) (asserting the first sentence of the rule but in discussion with
factors six or seven or subsection (k)).
The briefs can be viewed at the Wisconsin State Law Library,
[26]
(e) "Custom programs" mean utility and application software which accommodate the special processing needs of the customer. The determination of whether a program is a custom program shall be based upon all the facts and circumstances, including the following:
1.
The extent to which the vendor or independent consultant engages in significant
presale consultation and analysis of the user's requirements and system.
2.
Whether the program is loaded into the customer's computer by the vendor and
the extent to which the installed program must be tested against the program's
specifications.
3.
The extent to which the use of the software requires substantial training of
the customer's personnel and substantial written documentation.
4.
The extent to which the enhancement and maintenance support by the vendor is
needed for continued usefulness.
5.
There is a rebuttable presumption that any program with a cost of $10,000 or
less is not a custom program.
6.
Custom programs do not include basic operational programs or prewritten
programs.
7. If
an existing program is selected for modification, there must be a significant
modification of that program by the vendor so that it may be used in the
customer's specific hardware and software environment.
. . . .
(k)
"Prewritten programs," often referred to as "canned
programs," means programs prepared, held or existing for general use
normally for more than one customer, including programs developed for in-house
use or custom program use which are subsequently held or offered for sale or
lease.
[27] Wisconsin Stat. § 77.52(1)
imposes a tax on "the sale, lease or rental of tangible personal property .
. . ." Wisconsin Stat. § 77.51(20)
defines tangible property as follows:
"Tangible personal property" means all tangible
personal property of every kind and description and includes electricity,
natural gas, steam and water and also leased property affixed to realty if the
lessor has the right to remove the property upon breach or termination of the
lease agreement, unless the lessor of the property is also the lessor of the
realty to which the property is affixed. "Tangible personal property"
also includes coins and stamps of the
[28] An inappropriately narrow focus on eight words in a single sentence in the Commission's extensive analysis (see Chief Justice Abrahamson's dissent, ¶168) ignores the Commission's analysis in the same way that the inappropriately narrow focus on those words in the administrative rule ignores the rest of the definition. The record clearly shows that the Commission started its analysis with the administrative rule's first sentence, designating it parenthetically as the definition's "intro." "Intro" clearly meant "introduction," which in turn clearly meant that it was prefatory material, indicative that something else follows. The Commission then proceeded to construe and apply the remainder of the definition section——in other words, the seven factors. The Commission clearly indicated, by calling that portion the "intro," how it was proceeding. There is therefore no inconsistency in either the Commission's interpretation of the rule or my deference to it. (Justice Bradley cites to this paragraph and includes the Commission's parenthetical designation of the first sentence in Wis. Admin. Code § Tax 11.71(e) as the "intro," but she does so without acknowledging its significance. Justice Bradley's dissent, ¶221.)
[29] Majority op., ¶5.
[30] See letter from Robert Wm. Lang, Director, Legislative Fiscal Bureau, to Sen. Russell Decker & Rep. Kitty Rhoades, Chairs, Joint Committee on Finance (Jan. 30, 2007), available at http://www.legis.state.wi.us/lfb/Misc/2007_01_30_Revenue%20Estimates.pdf (last visited June 27, 2008).
[31]
[32] All references to the Wisconsin Statutes are to the 2003-04 version unless otherwise indicated.
[33]
[34]
[35]
[36]
[37] Martin Campbell-Kelly, From Airline Reservations to Sonic the Hedgehog: A History of the Software Industry 196 (MIT Press 2003).
[38] Campbell-Kelly, supra
note 9, at 197 (quoting Gail Edmondson et al., "Silicon
Valley on the
[39] Majority op., ¶88.
[40] Racine
Harley-Davidson, Inc. v.
[41]
In several tax cases the court has stated that when
the facts are undisputed, a court may substitute its judgment for that of the
Department of Revenue or the Tax Appeals Commission regarding the
interpretation and application of a statute to the undisputed facts. See,
e.g., DOR v.
Bailey-Bohrman Steel Corp., 93
[42] Majority op., ¶¶53-54, citing DaimlerChrysler v. LIRC, 2007 WI 15, ¶¶11, 13, 19, 299 Wis. 2d 1, 727 N.W.2d 311.
[43] Racine
Harley-Davidson, 292
[44] I do not reach the hypothetical issue whether deference would be due to the Tax Appeals Commission's interpretation of the administrative rule if the Commission's interpretation were reasonable. I do, however, disagree with the majority opinion's discussion of the issue of deference in the instant case. I note two obvious errors in the majority opinion.
First, it does not make sense for the majority opinion
to conclude that although this court owes only "due weight" deference
to the Tax Appeals Commission's interpretation of Wis. Stat. § 77.51, the court must
give the Commission "controlling weight" deference insofar as the
Commission interprets the Department of Revenue's administrative rule
interpreting Wis. Stat. § 77.51. See majority op., ¶53.
How can the majority opinion hold that the Commission's interpretation
of the Department's interpretation of Wis. Stat. § 77.51 is entitled to a
higher level of deference than the Commission's interpretation of the statute
itself?
Second,
the majority opinion is incorrect to conclude that whether the Department of
Revenue has acquiesced in the Commission's interpretation of a statute or
administrative rule has "no relationship" to this court's
determination whether deference is due (and how much deference may be due) to
the Commission's interpretation. See
majority op., ¶71. This court has previously recognized that
whether an adjudicative agency's interpretation of a statute has been embraced
by a line agency (the Department of Revenue in the instant case) may be
relevant to the question what level of deference is due to the adjudicative
agency's interpretation. See Racine
Harley-Davidson, 292 Wis. 2d 549, ¶53 (stating that when the Division of Hearings and Appeals adjudicates disputes
under certain statutes, the level of deference owed to the Division's decision
depends upon whether the line agency sharing concurrent jurisdiction with the
Division has adopted the Division's decision as its own).
For the
statutory provisions regarding the Department of Revenue's power to acquiesce
or not acquiesce in the Commission's interpretation of a statute, see Wis.
Stat. § 73.01(4)(e)1. and § 73.015(2) (providing that the Tax Appeals
Commission's interpretation of a statute is not binding upon the Department in
future cases when the Department seeks review of the order or decision of the
Commission construing the statute); Wis. Stat. § 73.01(4)(e)2. (providing
that even when the Department does not seek review of the Commission's decision
or order, the Department may issue a "notice of nonacquiescence," the
effect of which is that "although the decision or order is binding on the
parties for the instant case, the commission's conclusions of law, the
rationale and construction of statutes in the instant case are not binding upon
or required to be followed by the department of revenue in other cases.").
The Internal Revenue Service similarly may acquiesce or not acquiesce in decisions of the United States Tax Court. Susan A. Berson, Federal Tax Litigation § 1.01[7], at 1-13 (2008). In reviewing decisions of the United States Tax Court, the federal courts owe no deference to the Tax Court's interpretation of the Internal Revenue Code, or to the Tax Court's interpretations of the law generally. 3 Laurence F. Casey, Federal Tax Practice § 9.06, 9-13 to 9-14 (2007).
[45] See, e.g., majority op., ¶78 (acknowledging that a portion of the rule's text conflicts with the Commission's application of the rule); majority op., ¶85 n.24 ("While the Commission's language is somewhat perplexing . . ."); majority op., ¶96 (raising a question about the Tax Appeals Commission's interpretation of Wis. Admin. Code Tax § 11.71(1)(e)7.); majority op., ¶95 ("While the Commission could have used different language . . .").
[46] Menasha Corp. v. DOR,
[47] Menasha Corp. v. DOR,
[48] See majority op.,
¶20; Menasha Corp. v. DOR,
In 1995, Menasha Corporation agreed to lease the R/3 software from SAP for approximately $5.2 million dollars. In 1997, Menasha Corporation then made an additional payment of $100,000 in licensing fees to SAP, bringing the total sum that Menasha Corporation paid for the lease of SAP's R/3 software to approximately $5.3 million.
The record does not make clear what Menasha
Corporation received in consideration for the additional $100,000 in licensing
fees that it paid to SAP in 1997. The
parties do not seem to distinguish between Menasha Corporation's initial
payment of $5.2 million and its subsequent payment of $100,000. Thus, I assume for purposes of this opinion
that the additional payment of $100,000 was, like the initial payment of $5.2
million, for the lease of the R/3 software.
SAP's
transaction with Menasha Corporation is denominated a "license"
agreement, not a "lease" agreement.
However, the term "lease," as it is defined for purposes of
Wis. Stat. §§ 77.51-.66, "includes rental, hire and
license."
[49] See Menasha Corp. v. DOR,
[50] See Menasha Corp. v.
DOR,
Professor Campbell-Kelly explains that although SAP once operated "as a 50-person custom programming outfit rather than a software products firm," SAP's "switch to a software products firm came in 1978," when "SAP decided to rewrite its software as R/2 [a predecessor of R/3], with the medium-term aim of turning it into a product." Campbell-Kelly, supra note 9, at 193.
Professor Campbell-Kelly's history of the software
industry is accepted as an authoritative text.
The United States District Court for the Northern District of California
relied upon Professor Campbell-Kelly's book when deciding the much-publicized
case about Oracle Corporation's efforts to acquire the stock of PeopleSoft,
Inc. See
[51] See Menasha Corp. v. DOR,
[52] See Menasha Corp. v. DOR,
[53] Menasha Corp. v. DOR,
[54] Menasha Corp. v. DOR,
[55]
SAP's business practice is to distinguish between and to separate the sale of its R/3 software and any sale of consulting services to its customers. Thus SAP keeps its licensing agreement separate from any agreement for consulting services. The licensing agreement did not obligate SAP to modify the software to suit Menasha Corporation's particular work environment as part of the 1995 license transaction.
[56] Menasha Corp. v. DOR,
[57]
[58] Campbell-Kelly, supra note 9, at 172.
[59] Menasha Corp. v. DOR,
[60] Campbell-Kelly, supra note 9, at 172.
[61] See majority
op., ¶22; Menasha Corp. v. DOR,
[62] $265,093 is approximately 5%
of $5.3 million.
Both
Menasha Corporation and the Tax Appeals Commission are not consistent in
referring to the taxes at issue as "sales" taxes or "use"
taxes. The Tax Appeals Commission
variously characterizes Menasha Corporation's claim as one for the refund of
"use tax" payments, "sales tax" payments, and "sales
and use tax" payments. Menasha
Corp. v. DOR,
The
retail sales tax ordinarily is imposed on the retailer, in this case SAP.
[63] Menasha Corporation
initially sought a refund of $342,614.45.
Majority op., ¶24;
Menasha Corp. v. DOR,
The $77,521 in tax payments arising from Menasha
Corporation and SAP's transactions for maintenance of the R/3 software is no
longer at issue. The parties settled
their dispute about Menasha Corporation's liability for these taxes while
proceeding before the Tax Appeals Commission.
See Menasha
Corp. v. DOR,
[64]
[65]
[66]
[67] Menasha Corp. v. DOR,
The Commission refers to "Section TAX 11.71(1)(e)(intro)." In legislative parlance, an introduction is
"an unnumbered subunit of a section, subsection, paragraph, or subdivision
of the statutes with a colon at the end followed by a list of two or more items
in numbered subunits."
[68] See Wisconsin Bill Drafting Manual 2007-2008, § 2.01(1)(i) at 39 ("In a definition do not use 'means and includes.' 'Means' is complete and 'includes' is partial. Using 'includes' allows a court or administering agency to adopt additional meanings; using 'means' restricts them to reasonable constructions of your wording.") (emphasis added).
[69] See concurring op., ¶¶115-116.
[70] As the majority opinion acknowledges, the R/3 System "did not provide adequate processing for Menasha." Majority op., ¶16.
[71] Brief and Supplemental
Appendix of Petitioner Wisconsin Department of Revenue in Response to Non-Party
Brief of
[72]
[73]
[74] Menasha Corp. v. DOR,
Strangely, the majority opinion does not appear to defer to the Commission's conclusion that factor 6 functions as a veto. The majority opinion determines, in contradiction to the Commission, that factor 6 cannot alone be dispositive of the question whether a computer program is custom. See majority op., ¶92.
[75] Menasha Corp. v. DOR,
[76] Majority op., ¶88.
[77]
[78] Menasha Corp. v. DOR,
[79] Menasha Corp. v. DOR,
[80]
[81] The Department of
Revenue promulgated Wis. Admin. Code § Tax
11.71 in 1986. 361
The
legislature amended Wis. Stat. § 77.51(20)
in 1992 to provide that "tangible personal property" includes
"computer programs except custom computer programs." 1991
[82] The record contains a letter that the Department of Revenue sent to the attorney in response to his request.
[83] DOR Tech. Info. Mem. 38.4
(Aug. 7, 1978); DOR Tech. Info. Mem. 38.1 (Apr. 2, 1976).
This
definition of "prewritten (canned) programs" is nearly identical to
the definition of "prewritten programs" now set forth in the text of
Although the Department of Revenue's technical information memoranda are not binding upon the Department, the Department may be equitably estopped from collecting taxes in a manner inconsistent with the positions set forth in its technical information memoranda. See DOR v. Family Hosp., Inc., 105 Wis. 2d 250, 255-56, 313 N.W.2d 828 (1982) (holding that the Department was equitably estopped from collecting a certain tax from the defendant hospital when the hospital had reasonably relied to its detriment upon a DOR technical information memorandum stating that hospitals were exempt from the tax).
[84] DOR Tech. Info. Mem. 38.4 (Aug.
7, 1978); DOR Tech. Info. Mem. 38.1 (Apr. 2, 1976).
The Department of Revenue also remained consistent in
its position after promulgating Wis. Admin. Code § Tax 11.71. For
example, in a 1992 "tax release," the Department stated that under
Wis. Stat. § 77.51(20) (as amended by 1991
The Wisconsin Tax Bulletin, a quarterly newsletter that includes the Department of Revenue's published tax releases and private letter rulings, is available at the Department's Web site, http://www.revenue.wi.gov/ise/wtb/index.html (last visited June 30, 2008).
Under Wis. Stat. § 73.035, the DOR may issue and publish a private
letter ruling in response to a request for a ruling about facts relating to a
tax that the DOR administers. A private
letter ruling does not bind the requester and may not be appealed.
[85] Menasha Corp. v. DOR,
[86]
[87]
[88] Menasha Corp. v. DOR,
[89]
[90]
[91] See Foley and Lardner LLP, Legal News: Federal and State Controversy, Apr. 2004, at 6, available at http://www.foley.com/publications/newsletters_archive.aspx?year_2004 (last visited June 30, 2008).
[92] See Legislative Fiscal Bureau revenue and expenditure projections, January 30, 2007, 4. Available at http://www.legis.state.wi.us/lfb/Misc/2007_01_30_revenue%20estimates.pdf (last visited June 25, 2008).
[93] As an explanation for why the Commission failed to analyze the definition of "custom programs," the concurrence offers a strained interpretation. It surmises that the use of "intro" by the Commission clearly meant that the first sentence of the rule is merely "prefatory material," implying that it is not the definition of "custom programs." Concurrence, ¶115 n.3. The use of "intro" carries no such implication.
As noted in Chief Justice Abrahamson's dissent, the use of "intro" simply means that language is "preceding the rule's colon and numbered subunits." Chief Justice Abrahamson's dissent, ¶167 n. 39; see Wisconsin Bill Drafting Manual 2007-08, § 1.001(2) at 9. The use of "intro" certainly does not contradict the Commission's clear statement that the first sentence defines "custom programs."
[94] In other words, whether
there was extensive presale consultation and analysis is a factor in
determining whether the needs met by the software are special to the customer.
If the software does not meet the customer's needs (be they special or
general), the software is not custom. Likewise, whether the program is loaded
by the vendor, the extent of training required, the amount of maintenance
support, the cost of the system, whether the software is a basic operational or
prewritten program, and whether an existing system is significantly modified,
go to the question of whether any needs accommodated by the software are
special to the customer. See