Cummins Engine Company, Inc., Plaintiff-appellant, v. the United States, Defendant-appellee, 923 F.2d 826 (Fed. Cir. 1991)

Cummins Engine Company, Inc. (Cummins) appeals from the summary judgment of the United States Claims Court, 17 Cl.Ct. 854 (1989), holding that Cummins failed to make a proper election under 26 U.S.C. § 4223(b) to compute its federal excise tax liability on the purchase price of component parts in certain repair kits and denying a refund to Cummins of the claimed overpayment of $4,531,390, plus interest. We affirm.

* Cummins is an Indiana corporation which manufactures and sells diesel engines for heavy-duty trucks. Cummins provides replacement parts for these engines from two sources: it either manufactures them or it purchases them for resale. Sometimes Cummins sells these replacement parts in "kits"¹  which may contain both manufactured and purchased parts. Once packaged, Cummins recorded the cost of an entire kit as a single unit on its books, rather than recording the cost of each component part as a separate inventory item. The amount recorded as the cost of a kit was the sum of the cost of the parts plus the cost of the direct labor used to collect and package the parts into a kit.

Until 1983²  Cummins was subject to the manufacturers excise tax on its sales of truck engines and parts, including the parts in the kits. For the quarterly tax periods at issue in this case (January 1, 1977 through December 31, 1981),³  the Internal Revenue Code (Code) provided that this tax was to be based on (1) the selling price of manufactured parts, and (2) either the selling price or the purchase price of purchased parts (not used in the manufacture of other articles but simply resold for use "as is").⁴ 

When the election first became available in 1959, Cummins' Supervisor of General Accounting advised management that "further manufacture" as used in section 4223 of the Code did not encompass collecting parts and packaging them into kits and, therefore, the purchased parts in the kits would be eligible for the section 4223(b) (2) election. This meant that Cummins could have reduced its excise tax liability by making the purchase price election for the purchased kit parts, which were generally purchased for less than their subsequent selling prices. Thus, the lowest tax base for a kit containing a mix of purchased and manufactured parts would have been:

(1) the purchase price for the purchased parts; and

(2) the selling price for the manufactured parts.

Nevertheless, based on its accounting treatment of the manufactured parts, purchased parts, and parts included in kits as shown on its books and records, Cummins reported the parts for tax purposes in the following manner:

(1) if direct labor was expended and recorded to the part or kit, Cummins reported it on the returns under the selling price method;

(2) if no direct labor was expended and recorded, Cummins reported it under the elective purchase price method.

Since direct labor was expended assembling both the purchased and the manufactured parts into kits, the Cummins' formula had the effect of computing the tax on an entire kit as a unit--rather than on the component parts--under the selling price method. Accordingly, Cummins reported a greater excise tax liability on the kits than it would have reported for the parts had they not been in kit form. Cummins used this method of reporting despite its knowledge in 1959 that the purchased parts in kits were eligible for the purchase price election.

In March 1982, Cummins' management held a meeting in which the taxing of the kits was discussed. Outside counsel advised Cummins that basing the excise tax on the purchase price, instead of the selling price utilized by Cummins for the previous twenty-three years, would afford Cummins a tax savings. When Cummins realized the amount of money involved, it filed claims for a refund of the overpayment of excise tax for the twenty quarters, beginning January 1, 1977, for which the statute of limitations had not yet run.

The Commissioner of Internal Revenue determined that Cummins had never made an election on its books and records or on its tax returns to compute the excise tax on the kits under the purchase price method. The refund claims were denied for failure to make a proper election.

Cummins then filed a refund suit against the United States in the Claims Court. The United States subsequently moved for summary judgment and Cummins cross-moved. The Claims Court found that Cummins knew or should have known of the election and failed to act on it. The court rejected Cummins' argument that it had made a blanket election, finding no evidence of a formal or blanket election to tax the kits under the purchase price method. The court also determined that Cummins' claims for refund were not effective as new elections for the periods after January 1, 1977 because the statute expressly requires that the election be made in the return reporting the tax. The United States' motion for summary judgment was therefore granted.

II

On appeal, Cummins maintains that it has made one basic factual assertion and one corresponding legal contention from the inception of this case. The factual claim is that Cummins made a blanket purchase price election for all purchased parts in 1959, both for those in kits and for those not in kits. Cummins contends that the kits were misclassified on its records as a single manufactured unit which resulted in the kits being mistakenly taxed under the selling price method. The legal contention is that since Cummins made this mistake, it did not make a conscious choice to compute the tax on the selling price method; therefore, the Supreme Court's decision in Pacific Nat'l Co. v. Welch, 304 U.S. 191, 58 S. Ct. 857, 82 L. Ed. 1282 (1938), should not apply.

Summary judgment is appropriate when there are no genuine issues of material fact and the moving party is entitled to judgment as a matter of law. Rule 56(c), Rules of the United States Claims Court.

Cummins has not shown that there was any genuine dispute as to the material facts. Cummins' factual assertion is merely an argument that the facts presented to the court were legally sufficient to support a finding that Cummins had made a blanket election in 1959 to use the purchase price method. The Claims Court, however, determined otherwise. The court reviewed the undisputed facts and determined that there was no evidence of a purchase price election for the purchased parts in the kits. It noted that no extrinsic evidence to support Cummins' claim that it made a tax election for the kits was presented.⁵  Cummins was unable to produce any return evidencing a formal or blanket election to be taxed at the purchase price. Section 4223(b) (2) of the Code specifically states that the purchase price election is to be made "in the return reporting the tax."

Cummins' books and records accounted for its kits as inventory items. It did not separately inventory or account for the component parts contained in the kits. Based on this inventory method, Cummins computed its excise tax using the selling price method, which was the statutorily-prescribed method absent an affirmative election made in the return to use the purchase price method. At the time of the original election, there was nothing to prevent Cummins from electing the more favorable elective tax treatment based on the purchase price of the individual purchased components in the kits. The Claims Court found that Cummins knew or should have known at the time what election the corporation had in fact made. Moreover, the evidence clearly indicates that Cummins' management knew the election could be made for the purchased kit parts. Nevertheless, for the next twenty-three years Cummins used the selling price method, which was also a valid option.

Since Cummins chose, perhaps inadvertently, and continued for twenty-three years to compute the tax on the purchased kit parts under the selling price method rather than under the purchase price method, we are convinced that the Claims Court properly relied on Pacific Nat'l Co. There the Supreme Court held that when a taxpayer initially elects a valid method for computing its tax liability, it will be denied the opportunity to later recalculate this tax liability according to an alternative method even though the original computation method did not minimize the taxes. Pacific Nat'l Co., 304 U.S. at 194-195, 58 S. Ct. at 858-59. See also United States v. Kaplan, 304 U.S. 195, 58 S. Ct. 859, 82 L. Ed. 1285 (1938). In Pacific Nat'l Co., the taxpayer was not permitted to recalculate its tax liability on the gain from sales of property according to the installment method, when it had elected in its original return to use the deferred payment method. The Supreme Court stated:

Change from one method to the other, as petitioner seeks, would require recomputation and readjustment of tax liability for subsequent years and impose burdensome uncertainties upon the administration of the revenue laws. It would operate to enlarge the statutory period for filing returns ... to include the period allowed for recovering overpayments.... There is nothing to suggest that Congress intended to permit a taxpayer, after expiration of the time within which return is to be made, to have his tax liability computed and settled according to the other method. By reporting income from the sales in question according to the deferred payment method, petitioner made an election that is binding upon it and the commissioner.

304 U.S. at 194-95, 58 S. Ct. at 858-59.

Cummins relies on a number of authorities to distinguish the general rule in Pacific Nat'l Co. and to support its position that a taxpayer should later be permitted to exercise its full choice of available tax alternatives after having mischaracterized a transaction in a previously filed return. However, all of the cases cited by Cummins involve situations where the taxpayers either made initially improper elections which were set aside by the Commissioner of Internal Revenue or where the taxpayers incorrectly assumed that no taxes were due and thus failed to make elections at all.

The statute makes clear that a section 4223(b) (2) election must be made in the return reporting the tax. Cummins argues that the provision was satisfied by the filing of claims for refund. We do not agree that a claim for refund can be treated as the equivalent of a return. A claim for refund of excise tax seeks to adjust the tax that has already been reported on the return and is not itself a return reporting the tax. Because Cummins failed to make its election in the manner prescribed by the statute, the Claims Court properly concluded that the taxpayer was barred from changing to that election retroactively.

For the foregoing reasons, the judgment of the Claims Court is affirmed.

AFFIRMED.