Stonebridge Life Insurance v. Department of Revenue

18 Or. Tax 423

Opinion

HENRY C. BREITHAUPT, Judge.

I. INTRODUCTION

This matter comes before the court on a stipulated record and comprehensive cross-motions for summary judgment.

II. FACTS

During 2003, Stonebridge Life Insurance Company (taxpayer) engaged in the business of providing life, accident, and health insurance coverage. Taxpayer was licensed to do so in all 50 states and the District of Columbia. Taxpayer’s primary business locations were in Pennsylvania, Maryland, and Texas. Taxpayer had no physical operations, employees, telephone listings, or mail drops in Oregon. All of taxpayer’s Oregon insurance policies during 2003 originated through marketing by direct mail or telephone solicitation, and all business related to them occurred at one of taxpayer’s primary business locations outside of Oregon. In 2003, taxpayer received premiums of $5,978,898 from those Oregon policies for purposes of ORS 317.660(l)(a) and $661,443,717 in premiums from all its policies for purposes of ORS 317.660(l)(b) (together, the insurance sales factor). 1 That same year, taxpayer had no Oregon payroll for purposes of ORS 317.660(2)(a) but had a total payroll of $39,319,330 for purposes of ORS 317.660(2)(b) (together, the wage and commission factor). Also in 2003, taxpayer received gross income of $252,379 from Oregon real and tangible property for purposes of ORS 317.660(3)(a) and gross income of $1,565,829 from all its real and tangible property for purposes of ORS 317.660(3)(b) (together, the real estate income and interest factor). The income from Oregon real and tangible property was wholly from the interest received on two loans secured by Oregon property. Neither that income nor the loans was integral or necessary to taxpayer’s insurance business in Oregon or elsewhere.

In determining taxpayer’s insurance excise tax liability for 2003, the Department of Revenue (the department) applied the three-factor apportionment formula prescribed by ORS 317.660. Following that formula, the department determined that $12,787,485 of taxpayer’s total 2003 income was taxable by Oregon. The department derived that number from simple calculations. First, the department divided taxpayer’s gross income from Oregon real and tangible property ($252,379) by its gross income from all real and tangible property ($1,565,829) to determine the real estate income and interest factor (16.1179%). 2 Second, the department divided taxpayer’s income from Oregon premiums ($5,978,898) by its income from all premiums ($661,443,717) to determine the insurance sales factor (0.9039%). Third, the department divided taxpayer’s Oregon payroll ($0) by its total payroll ($39,319,330) to determine the wage and commission factor (0%). Fourth, the department averaged these three factors to determine the percentage of taxpayer’s total net income that should be apportioned to Oregon (5.6739%). Finally, the department multiplied that apportionment percentage by taxpayer’s total net income ($225,372,069) to determine taxpayer’s 2003 Oregon taxable income ($12,787,485).

Accordingly, the department asserted taxpayer’s net insurance excise tax liability to be $767,008. Taxpayer wrote the department requesting relief from the application of the three-factor apportionment under ORS 317.660. The department denied that request, reasoning that ORS 317.660 does not grant it the authority to use an alternate formula. Taxpayer appealed to this court and the matter was specially designated under Tax Court Rule (TCR) 1 C(2). Taxpayer argues that the department’s apportionment violates the Due Process Clause of the Fourteenth Amendment to the United States Constitution. 3

III. ISSUE

Did the department violate the Due Process Clause of the Fourteenth Amendment in apportioning $12,787,485 of taxpayer’s 2003 income to Oregon?

IV. ANALYSIS

A. Constitutionality of the Department’s Apportionment

In a line of cases stretching back more than 100 years, the United States Supreme Court has recognized that the federal constitution sets certain limits on the power of states to tax interstate enterprises. See Fargo v. Hart, 193 US 490, 24 S Ct 498, 48 L Ed 761 (1904); Trinova Corp. v. Michigan Treas. Dept., 498 US 358, 111 S Ct 818, 112 L Ed 2d 884 (1991). For instance, under the Commerce Clause 4 a state may tax the income of an interstate company only when the tax, in its practical effect, “is applied to an activity with a substantial nexus with the taxing State, is fairly apportioned, does not discriminate against interstate commerce, and is fairly related to the services provided by the State.” Complete Auto Transit, Inc. v. Brady, 430 US 274, 279, 97 S Ct 1076, 51 L Ed 2d 326 (1977). Under the Due Process Clause, a state may tax the income of an interstate enterprise only when there is “a 'minimal connection’ between the interstate activities and the taxing State.” Mobil Oil Corp. v. Comm’r of Taxes, 445 US 425, 436-37, 100 S Ct 1223, 63 L Ed 2d 510 (1980). While the “substantial nexus” requirement of the Commerce Clause is more stringent than the “minimal connection” requirement of the Due Process Clause, Quill Corp. v. North Dakota, 504 US 298, 313, 112 S Ct 1904, 119 L Ed 2d 91 (1992), the court need not probe that distinction in this case because taxpayer concedes that the Commerce Clause does not apply here; the Due Process Clause does. 5 Moreover, taxpayer does not dispute that the “minimal connection” requirement is met in this case.

Instead, this case turns on another requirement imposed by the Due Process Clause, “that the income attributed to the State for tax purposes must be rationally related to values connected with the taxing state.” Id. at 306 (internal quotation marks omitted). That requirement has been interpreted by the Supreme Court as similar or identical to the fair apportionment requirement of the Commerce Clause. See Container Corp. v. Franchise Tax Bd., 463 US 159, 165-66, 103 S Ct 2933, 77 L Ed 2d 545 (1983) (stating that the rational relationship requirement is one mandated by both the Due Process and Commerce clauses). Both clauses have been held to command that “a state may not, when imposing an income-based tax, tax value earned outside its borders.” Id. at 164 (internal quotation marks omitted). Accordingly, “the Court generally has drawn no distinction between the substantive requirements of the two clauses” but has “variously and interchangeably” attributed the same requirements to both. Walter Hellerstein, State Taxation of Corporate Income from Intangibles: Allied-Signal and Beyond, 48 Tax L Rev 739, 744 (1993); Trinova, 498 US at 373 (stating that “[t]he Complete Auto test, while responsive to Commerce Clause dictates, encompasses as well [the relevant due process requirements]”); see also Container Corp., 463 US at 164-66 (treating the analyses as one and the same); Norfolk & Western R. Co. v. State Tax Comm’n, 390 US 317, 323-26, 88 S Ct 995, 19 L Ed 2d 1201 (1968) (same). 6 At times, the Court has attempted to differentiate its analysis under the two clauses. See, e.g., Mobil, 445 US at 436-37, 442-46 (discussing the two inquiries as separate); Moorman Mfg. Co. v. Bair, 437 US 267, 98 S Ct 2340, 57 L Ed 2d 197 (1978) (same). 7 Most recently, in Quill, the Court admitted that its precedent had caused confusion and sought to clarify the distinction between the two clauses and the analyses they require. 504 US at 305-06. However, neither in Quill nor in any subsequent case has the Court clarified the specific distinction between the “rational relationship” requirement of the Due Process Clause and the “fair apportionment” requirement of the Commerce Clause. 8 To the degree that the analysis under each clause differs, this court would reach the same result in this case under any of the Court’s precedents.

In calculating a fair apportionment, a state may determine its fair share of an interstate enterprise’s income or value “by reference to the total system of which the intrastate assets are a part.” Norfolk & Western, 390 US at 324. Because the exact amount of a state’s fair share is often difficult to measure, the Supreme Court has held that “the Constitution imposes no single formula on the States.” Container Corp., 463 US at 164. Indeed, the Court has upheld the general use of a “number” of state formulas, Norfolk & Western, 390 US at 324, including a single-factor sales formula, Moorman, 437 US at 281 (upholding Iowa’s single-factor sales formula), and a three-factor formula like Oregon’s, Container Corp., 463 US at 170 (stating that the three-factor formula has become “something of a benchmark against which other apportionment formulas are judged”), even though those formulas “occasionally over-reflect or under-reflect income attributable to the taxing State.” Moorman, 437 US at 273.

The Supreme Court has held that, to be fair, an apportionment formula must be both internally and externally consistent. Container Corp., 463 US at 169. To be internally consistent, a formula must not subject to taxation more than all of a business’s income if applied by every jurisdiction. Id. Taxpayer does not claim that the three-factor formula of ORS 317.660 is internally inconsistent. 9 To be externally consistent, “the factor or factors used in the apportionment must actually reflect a reasonable sense of how income is generated.” Id. The three-factor formula, for instance, “has gained wide approval [as a tool for apportioning income] precisely because payroll, property, and sales appear in combination to reflect a very large share of the activities by which value is generated.” Id. at 183. 10 However, while the Court has not struck down any apportionment formula per se, it has in certain cases “found that the application of a [particular] formula to a particular taxpayer” was unconstitutional. Moorman, 437 US at 274 (emphasis in original). Taxpayer asserts that this is such a case.

To be sure, even the classic three-factor formula (based on property, payroll, and gross receipts), despite its many advantages, “is necessarily imperfect.” Container Corp., 463 US at 183. The Supreme Court has recognized several weaknesses, including the “essentially arbitrary” nature of the “one-third-each weight given to the three factors” and the potential existence of additional factors that generate income. Id. at 183 n 20.* 11 Indeed, the three-factor formula may sometimes be inappropriate for companies that are not engaged in either the mercantile or manufacturing business. See Jerome R. Hellerstein & Walter Hellerstein, I State Taxation ¶ 8.06 (3d ed 2005) (praising use of the three-factor formula for mercantile and manufacturing companies); id. ¶ 10 (describing various other methods used to apportion income for other industries, including service, transportation, publishing, broadcasting, finance, investment, professional sports, construction, and communications). 12 The formula’s weaknesses, however, are often outweighed by its strengths. “[I]n certain cases one factor may be unreasonably high and another unreasonably low but the application of the three factors together fairly represents business activity.” Twentieth Century-Fox Film v. Dept. of Rev., 299 Or 220, 233, 700 P2d 1035 (1985). Sales in one state may generate more net revenue than sales in another state, such that it would be fair to apportion to the first state a greater share of a company’s profits than would otherwise seem fair based on gross sales data alone. See Moorman, 437 US at 276 (“Obviously, all sales are not equally profitable. Sales in Iowa, although only 20% of gross sales, may have yielded a much higher percentage of appellant’s profits.”); id. at 272 (describing how that may have been so). In order to prove constitutional distortion, a company must have “convincingly demonstrated,” Trinova, 498 US at 384, that “a significant portion of the income attributed to [a state] in fact was generated by [operations in other states]Moorman, 437 US at 272.

Taxpayer bears a heavy burden. To succeed, it must prove “by ‘clear and cogent evidence’ that the income attributed to [Oregon] is in fact ‘out of all appropriate proportion to the business transacted * * * in [Oregon],’ or has ‘led to a grossly distorted result.’ ” Moorman, 437 US at 274 (citations omitted). The Supreme Court has considered dozens of cases in this area and has struck down a state’s apportionment as unfair only twice in the last 75 years. Norfolk & Western, 390 US at 329-30; Hans Rees’ Sons v. Maxwell, 283 US 123, 51 S Ct 385, 75 L Ed 879 (1931). 13 The parties cite no other case in which a court has struck down an apportionment as unfair in violation of the Due Process or Commerce clauses and this court is aware of only a few such cases. See In re British Land (Maryland), Inc. v. Tax Appeals Tribunal, 85 NY2d 139, 647 NE2d 1280 (1995); Tambrands, Inc. v. State Tax Assessor, 595 A2d 1039 (Me 1991). 14

Many taxpayers seeking to prove unconstitutional apportionment have foundered on the requirement that they present the court with “clear and cogent” evidence of constitutional distortion, a task that some commentators have deemed “daunting.” Hellerstein & Hellerstein, I State Taxation ¶ 8.15(1). Although “evidence may always be received which tends to show that a state has applied a method, which, albeit fair on its face, operates [unconstitutionally],” Hans Rees’ Sons, 283 US at 134, the Court has repeatedly expressed skepticism about the value of evidence derived from “formal geographical or transactional accounting,” also known as separate accounting. See, e.g., Container Corp., 463 US at 164-65, 181 (criticizing separate accounting as an appropriate baseline against which to judge the constitutionality of a state’s apportionment). But see Hans Rees’ Sons, 283 US at 135 (relying on separate accounting evidence to examine and ultimately strike down an apportionment); Moorman, 437 US at 272 (faulting a taxpayer for not introducing separate accounting evidence). In Moorman and Trinova, the Court also rejected attempts to show distortion through comparison to alternate apportionment formulas. Moorman, 437 US at 270; Trinova, 498 US at 384.

In Hans Rees’ Sons, the Court struck down the application of North Carolina’s single-factor property formula to a New York company that sold its goods worldwide but had its sole factory in North Carolina. 283 US at 126-28. To prove distortion, the company offered detailed separate accounting evidence regarding the geographic origin of its profits and the methods by which it derived those profits. Id. at 134-35. That evidence showed that while North Carolina apportioned 80% of the company’s profits to itself, no more than 22% of those profits actually originated in North Carolina, and, on average, only 17% did. Id. at 134. The distortion created by the application of North Carolina’s formula to the company was thus between 363% and 470%. But see Container Corp., 463 US at 184 (stating that the distortion was closer to 250%). The Court found that degree of distortion unconstitutional. Hans Rees’ Sons, 283 US at 135-36. 15

In Norfolk & Western, a property tax case, the Court held unconstitutional the application of Missouri’s single-factor formula, based on track-miles, to apportion the rolling stock of a Virginia railroad corporation with interstate operations and only incidental business in Missouri. 390 US at 319. The railroad had acquired by lease the fixed assets and rolling stock of the Wabash Railroad Company (Wabash), which had substantial operations in Missouri. Id. Prior to the lease, the railroad had owned no fixed property and only minimal rolling stock in Missouri, but did have substantial rolling stock in other states. Id. Under Missouri statutes, the assessment of rolling stock was determined by apportioning the value of all rolling stock owned or leased by the railroad according to the ratio of track miles owned or leased by the railroad in Missouri to the railroad’s total track miles. Id. at 320-21. The railroad showed that its business focused on coal operations, that it did not conduct those operations in Missouri, and that it had leased the Missouri rail stock only to “diversify its business, not to provide the opportunity for an integrated through movement of traffic.” Id. at 319.

In the year at issue, Missouri determined the value of the railroad’s rolling stock in Missouri to be nearly $20 million, even though the state had assessed Wabash for only $9.2 milbon of rolling stock the year before. Id. at 321-22. Indeed, the railroad’s “actual count of the rolling stock in Missouri” showed that its true value was closer to $7.6 million. Id. at 321-22, 326-27. The railroad also “demonstrated that neither the amount of rolling stock in Missouri nor the Missouri operations of [the railroad] and Wabash had materially increased in the intervening period.” Id. at 322. Finally, the railroad argued that if the rolling stock had been taxable to Wabash in the year in question, instead of to it, the assessed value would have been $10 million. Id. at 322 n 4. The Court described the 263% level of distortion between the state’s and the railroad’s apportionments 16 as a “chasm” and opined that “the record is totally barren of any evidence relating to enhancement or to going-concern or intangible value, or to any other factor which might offset the devastating effect of the demonstrated discrepancy.” Id. at 327-28. As the Court explained, “when a taxpayer comes forward with strong evidence tending to prove that [a] formula will yield a grossly distorted result in its particular case, the State is obliged to counter that evidence or to make the accommodations necessary to assure that its taxing power is confined to its constitutional limits.” Id. at 329. Because Missouri had not done so in the face of the railroad’s extensive evidence, the Court struck down the apportionment. Id. at 329-30. 17

In British Land, the New York Court of Appeals held unconstitutional the application of New York’s three-factor formula (only slightly different from Oregon’s) to a New York company that had acquired and sold a building in Maryland. 647 NE2d at 1282. The state had apportioned 64% of the company’s income from the sale to itself, primarily because the value of the company’s New York property dwarfed that of its Maryland property. Id. The company, however, offered extensive evidence, including separate accounting evidence, to show that its New York operations had little or nothing to do with its sale of the Maryland property. Id. at 1282, 1285. Among other things, that evidence showed that the company had decided to sell the Maryland property before it moved to New York, that the gain from the sale dwarfed the company’s other revenues, and that the value of the Maryland property had increased due principally to its Maryland location, management, and renovations. Id. at 1285. Accordingly, the court struck down New York’s apportionment, holding that the state had sought to tax 2,200% more of the company’s profits than it had a right to. Id. at 1285.

In contrast, the court in Unisys Corp. v. Commonwealth, 571 Pa 139, 812 A2d 448 (2002), rejected a taxpayer’s argument that the state’s apportionment was unfair because the taxpayer failed to present an alternate formula or evidence of actual business activity strong or coherent enough to counter the state’s formula. Similarly, in CSX Transp., Inc. v. Director, 2005 WL 1531329 (NJ Tax), the court rejected an apportionment challenge because the taxpayers’ evidence was insufficient. In CSX, the railroads’ evidence consisted of an alternate track-miles formula, a three-factor formula, and various ratios related to operating revenues, train hours, train miles, and freight car miles. Id. at *3. The court found, however, that none of that evidence bore a relationship to income arising out of New Jersey activities. Id. at *10.

In the present case, as an initial matter, taxpayer invites this court to engage in a thought experiment. 18 Had taxpayer received no income from Oregon real or tangible property in 2003 (i.e., had no loans secured by Oregon property), but still earned income from its other real and tangible property, the apportionment formula of ORS 317.660 would have resulted in the following calculations. Because both the real estate income and interest factor and the wage and commission factor would have been 0%, and the insurance sales factor would have remained 0.9039%, ORS 317.660 would have apportioned 0.3013% of taxpayer’s total net income ($225,372,069) to Oregon, resulting in 2003 Oregon taxable income of $679,058. Instead, because taxpayer derived an additional $252,379 in gross income from real and tangible property, which happened to be located in Oregon and not some other place, application of ORS 317.660 resulted in 2003 Oregon taxable income of $12,787,485. In other words, by collecting $252,379 in gross income from two loans secured by Oregon real and tangible property, taxpayer increased its Oregon taxable income by $12,108,427 (a difference of 1,883%). 19 The unconstitutionality of that result, taxpayer claims, is patent because the 1,883% level of distortion is far greater than other levels of distortion that the Supreme Court has found unconstitutional. See Hans Rees’ Sons, 283 US at 134 (between 368% and 470%); Norfolk & Western, 390 US at 321-22, 326-27 (263%).

Additionally, taxpayer emphasizes that the department has stipulated that taxpayer’s income from Oregon real and tangible property “was neither integral nor necessary to the insurance business carried on by Plaintiff within or without the State of Oregon.” To underscore that fact, taxpayer stresses that its gross income from all real and tangible property ($1,565,829), including its income from Oregon property, comprises less than 0.3% of its combined gross income from all real and tangible property and all insurance sales ($663,009,546). Taxpayer argues that it is fundamentally unfair to attribute one-third of its revenue generation to real and tangible property, as the three-factor formula of ORS 317.660 does, when that property accounts for less than 0.3% of its overall income from insurance sales and real and tangible property combined. That unfairness is compounded, taxpayer contends, when one considers the effect of attributing one-third of taxpayer’s revenue generation to real and tangible property in this case: the 1,883% increase in taxpayer’s Oregon taxable income attributable to its earning what amounts to 0.0381% of its gross income from insurance sales and real and tangible property. 20 Taxpayer claims that it is unconstitutional in this case to base such a large increase in its Oregon income tax liability on what is ultimately a minuscule part of its business.

The department defends the application of ORS 317.660 to this case by arguing that the three-factor formula accurately represents taxpayer’s Oregon income generating activities. For instance, the department notes that taxpayer “reported no wages or commissions in Oregon for 2003, regardless of the fact that it transacted $5,978,898 of insurance business in that year.” According to the department, the inclusion in the formula of a 0% wage and commission factor, and the low insurance sales factor, balances taxpayer’s relatively high real estate income and interest factor, such that the three factors together achieve a constitutionally acceptable apportionment. 21 The department also asserts that taxpayer’s income from Oregon real and tangible property, small though it may be in comparison to taxpayer’s total income, is significant to taxpayer beyond its dollar value for reasons peculiar to the insurance industry. Specifically, the department refers to the common practice among insurance companies of investing in securities and real estate to generate additional revenue. The department also refers to minimum capital and surplus requirements to which insurance companies are subject in Oregon and most other states. 22

The department’s contentions are unconvincing. The sheer numbers at issue in this case make a compelling argument that the high real estate income and interest factor neither properly nor permissibly balanced the low wage and commission and insurance sales factors. Even more tellingly, the department’s claims are contradicted by the key stipulation that taxpayer’s income from Oregon real and tangible property “was neither integral nor necessary to the insurance business carried on by Plaintiff within or without the State of Oregon.” Indeed, none of taxpayer’s income from real and tangible property appears to have been either integral or necessary to its business, constituting as it did less than 0.3% of taxpayer’s combined gross income from all insurance sales and real and tangible property. To the degree that any of taxpayer’s real and tangible property income was integral or necessary, it is nonetheless difficult to reconcile the one-third weight given to taxpayer’s income from Oregon real and tangible property with the admitted insignificance of that income. Although the value generated by taxpayer’s income from Oregon real and tangible property was not negligible, it could not have represented one-third of taxpayer’s total value in light of the stipulation. Yet the department, in applying ORS 317.660, assumed the latter and therefore drastically increased the amount of taxpayer’s income apportionable to Oregon. 23

Even absent the stipulation regarding the unimportant and unnecessary nature of taxpayer’s Oregon property income, other stipulated facts show that taxpayer’s Oregon apportionment factors did not generate much, if any, value beyond their numerical worth. As the department admits, taxpayer’s factors of production are easily sourced, at least with regard to Oregon. Taxpayer lacks any Oregon wages or commissions, or factory or office space, or any other means of production in Oregon that generate or help to generate income in other states. If such were not the case, then profits derived in other states could, in part, be attributable to Oregon, and thus apportionable to it. See Container Corp., 463 US at 182-84. However, where, as here, the real and tangible property taxpayer did have in Oregon, the security interests, did not play a larger role in contributing to taxpayer’s bottom line or otherwise generate value for taxpayer, but rather stood alone as an investment, it is easy to see that its inclusion in the three-factor formula might unconstitutionally allocate other income, derived elsewhere, to Oregon 24

On the side of income, as opposed to factors of production, the income from taxpayer’s loans secured by Oregon property is easily and obviously sourced to Oregon. Taxpayer’s insurance sales income is also easily sourced: the parties have stipulated to the exact amount of gross income taxpayer derived from Oregon premiums. It would appear that no other premiums can be attributed to Oregon, especially given the high levels of state specific regulation of insurance sales. Taxpayer contends that the department has, in effect, attributed to Oregon an amount of taxable income ($12,787,485) that is more than double taxpayer’s combined gross income from Oregon insurance sales and real and tangible property ($6,231,277). To illustrate its argument, taxpayer adopts an extreme and improbable hypothetical: taxpayer assumes that no costs were associated with its income from Oregon premiums. That assumption, beyond being improbable and in the department’s favor, would render taxpayer’s net income from Oregon premiums equal to its gross income from those premiums. One could even give the department the benefit of another assumption, that taxpayer’s gross/net income from Oregon premiums was worth more than their absolute dollar value. That hypothetically could be the case if one took a broader view of value, for instance, if taxpayer’s Oregon operations were particularly important to its brand image or if its Oregon sales were particularly reliable. 25 Taxpayer’s hypothetical, as augmented by the court’s additional assumption, ends in the argument that it is unreasonable to conclude that the true value of taxpayer’s Oregon insurance sales, the amount apportionable to Oregon, was twice their numerical worth, that taxpayer really “earned” and should be assessed on more than $2 for every $1 of actual income. Taxpayer holds that the same is true even if the value of its Oregon insurance sales were combined with the amount of its Oregon property investment income.

Ultimately, taxpayer’s argument can be reduced to the claim that the sheer accident that 16% of its minuscule gross income from real and tangible property came from Oregon cannot justify Oregon’s claim to almost 6% of taxpayer’s total income, given that taxpayer had no Oregon wages or commissions and that less than 1% of taxpayer’s insurance sales came from Oregon. The department has conceded the insignificance of the distorting factor in this case, yet it urges the court to allow that factor to attribute to Oregon a share of taxpayer’s income that was generated completely outside Oregon. As the Court explained in Norfolk & Western, “when a taxpayer comes forward with strong evidence tending to prove that [a] formula will yield a grossly distorted result in its particular case, the State is obliged to counter that evidence or to make the accommodations necessary to assure that its taxing power is confined to its constitutional limits.” 390 US at 329. Here, taxpayer has made the predicate showing described in Norfolk & Western and the burden shifts to the state to offer some explanation or evidence “relating to enhancement or to going-concern or intangible value, or to any other factor which might offset the devastating effect of the demonstrated discrepancy.” Id. at 327-28. The department must fall victim to the actual features of the apportionment formula of ORS 317.660, as applied here, and its own stipulation of facts. 26

Given the procedural posture in this case, in which the parties have not sought trial, but have been content to file comprehensive cross-motions for summary judgment based on a stipulated record, no more evidence may be received. Therefore, the court concludes that application of ORS 317.660 in this case allocated to Oregon a share of taxpayer’s 2003 income that was “out of all appropriate proportion to the business transacted” by taxpayer in Oregon. Hans Rees’ Sons, 283 US at 135. Simply put, taxpayer’s low insurance sales and wage and commission factors did not balance out taxpayer’s circumstantially high real estate income and interest factor. The three factors do not appropriately or permissibly reflect taxpayer’s 2003 Oregon business activity; instead, the high real estate income and interest factor “grossly distorted” the value generated by taxpayer’s Oregon operations. Norfolk & Western, 390 US at 326. 27

B. Remedy

Having concluded that the department’s apportionment to Oregon of $12,787,485 of taxpayer’s total net income is unconstitutional, the question remains what remedy to grant taxpayer. Taxpayer urges this court to adjust taxpayer’s tax liability by adopting, at least for purposes of this case, an alternate formula, which taxpayer asserts is reasonable. Taxpayer would first have the court average the insurance sales and wage and commission factors only, ignoring the real estate income and interest factor, to determine an “Oregon apportionment percentage” of 0.4520%. The court would then determine taxpayer’s “Total net income subject to apportionment” by subtracting from its total net income ($225,372,069) taxpayer’s total gross income from all real and tangible property ($1,565,829), arriving at $223,806,240. Next, the comb would multiply taxpayer’s “Oregon apportionment percentage” by its “Total net income subject to apportionment” to discern an “Income apportioned to Oregon” of $1,011,511. Finally, the court would add to that figure taxpayer’s gross income from Oregon real and tangible property ($252,379) to determine taxpayer’s 2003 Oregon taxable income: $1,263,890. Accordingly, taxpayer asserts that its 2003 Oregon tax liability should be $83,373. Because the department assessed taxpayer’s 2003 tax liability to be $767,008, taxpayer requests relief in the form of a refund of $683,635 plus statutory interest.

5. Taxpayer assumes that this court has the authority to reapportion and adjust taxpayer’s income and tax liability in this case. In support of that proposition, taxpayer cites ORS 314.670 and ORS 314.280, which allow the department to use methods of apportioning income alternative to those otherwise provided for by statute. However, neither ORS 314.670 nor ORS 314.280 apply to this case. First, the court notes that neither ORS 314.670 nor ORS 314.280 apply to insurance companies. See ORS 314.610(4) (defining insurance companies as financial organizations); ORS 314.615 (requiring financial organizations to report income under ORS 314.280, thereby exempting them from the provisions of ORS 314.670); ORS 317.660 (stating in direct and unambiguous terms that insurance companies are to use the three-factor formula of ORS 317.660 “[i]n lieu of the provisions of ORS 314.280” if they conduct business both within and without Oregon, as taxpayer does). 28 Second, neither ORS 314.670 nor ORS 314.280 grants power to a court to reapportion a taxpayer’s income; they grant that power only to the department. See Twentieth Century-Fox Film, 299 Or at 233-34 (ordering the department to reapportion a taxpayer’s income); Crocker Equipment Leasing, Inc. v. Dept. of Rev., 314 Or 122, 131, 838 P2d 552 (1992) (same). Third, ORS 317.660 allows for only one exception to the standard three-factor formula, and that relates to the calculation of the insurance sales factor, ORS 317.660(1), not the use of a different formula altogether. That the statute’s only exception retains the three-factor formula undercuts any argument that some other formula might be used to apportion taxpayer’s income. ORS 174.010; Fisher Broadcasting, Inc. v. Dept. of Rev., 321 Or 341, 353, 898 P2d 1333 (1995) (applying “the rule of statutory construction that the inclusion of one is the exclusion of the other (inclusio unius est exclusio alterius)”). Thus, the text and context of ORS 317.660 leave no doubt that neither the department nor this court possesses the authority to reapportion taxpayer’s income using a formula different from that found in ORS 317.660. See PGE v. Bureau of Labor and Industries, 317 Or 606, 610-11, 859 P2d 1143 (1993) (setting out statutory interpretation framework).

Some cases seem to indicate that courts possess an inherent power to adjust an unconstitutional apportionment. In Hans Rees’ Sons, the Court remanded the case for further proceedings after finding the apportionment unconstitutional. 283 US at 136. In Norfolk & Western, the Court remanded the case to the Missouri Supreme Court, leaving it “open” to that court to “remand the case to an appropriate tribunal to reopen the record for additional evidence to support the [state’s] assessment.” 390 US at 330. 29 In the present matter, the court sees no basis, and the parties have provided none, for concluding that the court possesses an inherent reapportionment power. Moreover, the procedural posture of this case, in which the parties have filed comprehensive cross-motions for summary judgment based on a stipulated record, does not lend itself well to a reopening of the record for additional evidence to support the department’s assessment. Accordingly, the court concludes that it lacks the authority to reapportion taxpayer’s 2003 income or to order the department to do so. Without that authority, the court cannot give taxpayer the exact relief it seeks. Instead, based on the conclusion that the taxes paid by taxpayer were paid under an unconstitutional application of the statutory regime, the court awards taxpayer a refund of $767,008, the full amount of insurance excise tax it paid for its 2003 Oregon business activity.

V. CONCLUSION

Based on the foregoing reasons, the court concludes that the department violated the Due Process Clause of the Fourteenth Amendment in apportioning $12,787,485 of taxpayer’s 2003 income to Oregon. The court also concludes that it lacks the authority to reapportion taxpayer’s 2003 income. Now, therefore,

IT IS ORDERED that Plaintiffs Motion for Summary Judgment is granted; and

IT IS FURTHER ORDERED that Defendant’s Cross-Motion for Summary Judgment is denied; and

IT IS FURTHER ORDERED that Defendant shall refund to Plaintiff $767,008, plus statutory interest.

1 All references to the Oregon Revised Statutes (ORS) are to the 2003 edition.

2 Percentage figures are rounded here to four decimal places; dollar figures, however, reflect calculations using the exact percentages. All figures related to this case are taken from the parties’ stipulated facts and other submissions unless otherwise noted.

3 The Due Process Clause of the Fourteenth Amendment states: “[N]or shall any State deprive any person of life, liberty, or property, without due process of law.” US Const, Amend XIV, § 1. Neither party has argued, and the court does not perceive, that any statutory issues or issues arising under the Oregon Constitution would be determinative in this case.

4 The Commerce Clause states: “The Congress shall have Power * * * To regulate Commerce with foreign Nations, and among the several States, and with the Indian Tribes.” US Const, Art I, § 8, cl 3.

5 Taxpayer recognizes that its claim rests on the Due Process Clause alone because “Congress removed all Commerce Clause limitations on the authority of the States to regulate and tax the business of insurance when it passed the McCarran-Ferguson Act, 59 Stat. 33, 15 U.S.C. § 1011 et seq.” W. & S. Life Ins. Co. v. State Bd. of Equalization, 451 US 648, 653, 101 S Ct 2070, 68 L Ed 2d 514 (1981).

6 In Norfolk & Western, the Court explicitly noted the question whether the analysis under the two clauses differed, stating:

“The problem under the Commerce Clause is to determine what portion of an interstate organism may appropriately be attributed to each of the various states in which it functions. So far as due process is concerned the only question is whether the tax in practical operation has relation to opportunities, benefits, or protection conferred or afforded by the taxing State. Those requirements are satisfied if the tax is fairly apportioned to the commerce carried on within the State. Neither appellants nor appellees contend that these two analyses bear different implications insofar as our present case is concerned.”

390 US at 325 n 5 (citations and internal quotation marks omitted).

7 In those cases, the Court discussed the Due Process Clause as being primarily concerned with preventing the unreasonable result of state taxation of profits earned outside the taxing state’s borders. Mobil Oil, 445 US at 442; Moorman, 437 US at 273-75. The Court treated the Commerce Clause, on the other hand, as being primarily concerned with the harms of duplicative taxation and undue burdens on interstate commerce. Mobil Oil, 445 US at 443-46; Moorman, 437 US at 276-81.

8 See 18 OTR 423, 429 n 10 (2006) (discussing the lack of clarity).

9 Indeed, such a claim would be difficult to make given that the Court in Container Corp. upheld as constitutional a three-factor formula in several respects similar to ORS 317.660.463 US at 184.

10 The Court has also stated that the three-factor formula “can be justified as a rough, practical approximation of the distribution of either a corporation’s sources of income or the social costs which it generates.” General Motors Corp. v. District of Columbia, 380 US 553, 561, 85 S Ct 1156, 14 L Ed 2d 68 (1965) (emphasis added); see also Jerome R. Hellerstein & Walter Hellerstein, I State Taxation ¶ 8.15(2) (3d ed 2005) (discussing the latter justification). The latter justification may be grounded in the Due Process Clause’s concern with the “opportunities, benefits, or protection conferred or afforded by the taxing State.”Norfolk & Western, 390 US at 325 n 5. It also appears to be embodied in the fourth prong of the Complete Auto test, rooted in the Commerce Clause, which requires that a tax be “fairly related to the services provided by the State.” 430 US at 279.

Accordingly, it might appear that the fourth prong of the Complete Auto test (fair relationship) would require an analysis similar to that required by the second prong of the test (fair apportionment) or the Due Process Clause (rational relationship between apportionment and value). However, the Court in Quill, 504 US at 313, did not see the fair relationship and fair apportionment concepts as similar or requiring the same analysis; instead, it found the fair relationship requirement similar to the substantial nexus requirement. See also Commonwealth Edison Co. v. Montana, 453 US 609, 625-26, 101 S Ct 2946, 69 L Ed 2d 884 (1981) (same). In Commonwealth Edison, the Court held that the fair relationship test does not relate to amounts or values, but rather requires that “the measure of the tax must be reasonably related to the extent” of a company’s contact with the taxing state.” 453 US at 625-26 (emphasis in original). Despite the similarities between that test and the fair apportionment and rational relationship tests, it is apparently different. See id. at 645 (Blaekmun, J., dissenting) (arguing that the Court had “emasculatefd]” the fair relationship prong of the Complete Auto test).

11 The property factor in the traditional three-factor formula is typically measured by the cost of property and only includes real and tangible personal property. It therefore could be said not to represent all income producing property where a taxpayer has income from intangible personal property such as stocks and bonds. Notwithstanding the general exclusion of intangible property from inclusion in the property factor, income from intangible property is generally included in a company’s net income, which is the tax base subject to apportionment through use of the formula. See generally, Hellerstein, State Taxation of Corporate Income from Intangibles: Allied-Signal and Beyond, 48 Tax L Rev 739. That is also the case under ORS 317.660. However, the real estate income and interest factor of ORS 317.660(3) includes only gross income from real and tangible property, omitting intangible property; and, it is measured by the income from, rather than the cost of, property.

12 Indeed, recent trends indicate that many states, including Oregon, are moving away from the three-factor formula to a single-factor formula, based on gross receipts, for most industries. Stanley R. Arnold & Frank D. Katz, When States Shift to Single-Factor Sales Apportionment: Is it a Win for Business, or “Be Careful What You Wish For’’?, 12 Multistate Tax R 575, 576 (2005). However, such changes may be driven by reasons unrelated to theoretical or actual deficiencies in the three-factor formula.

13 The court refers to apportionment, in this respect, in terms of both facial challenges to an apportionment formula and “as applied” challenges to a particular apportionment. “Indeed, the U.S. Supreme Court has never invalidated a multi factor formula on its face or as applied to an interstate manufacturing or mercantile enterprise on the grounds that it failed to satisfy the ‘fair apportionment’ requirement.” Hellerstein & Hellerstein, I State Taxation ¶ 8.14. Prior to 1931, the Court had struck down several apportionments based on single factor formulas (usually focused on property). See id. ¶¶ 8.06, 8.14; Norfolk & Western, 390 US at 324-26 (citing cases). Each of those cases involved railroads, telegraph companies, or other businesses of a similar nature, not manufacturers, or mercantile or financial companies. Hellerstein & Hellerstein, I State Taxation ¶ 8.12-.13.

14 In AT&T v. Dept. of Rev., 143 Wis 2d 533, 422 NW2d 629 (Wis Ct App 1988), the court held unconstitutional the application of that state’s three-factor formula because the state did not include in the property factor the value of the taxpayer’s investments in its subsidiaries. See also Kellogg Co. v. Herrington, 216 Neb 138, 343 NW2d 326 (1984) (same). Those decisions, however, run contrary to the “weight of authority from other jurisdictions.” Unisys Corp. v. Commonwealth, 571 Pa 139, 812 A2d 448, 459 (2002) (citing cases).

15 Although the company pleaded violations of both the Commerce and Due Process clauses, the Court did not specify on which clause it rested its holding, if not on both. 283 US at 125-26. In Moorman, the Court described its holding in Hans Rees’ Sons as based at least in part on the Due Process Clause. 437 US at 274.

16 The court in General Dynamics Corp. v. Sharp, 919 SW2d 861, 869 (Tx Ct App 1996), calculated the distortion found in Norfolk & Western as being between 162% and 205%. In Unisys, the court calculated the distortion found in Norfolk & Western as being between 266% and 300%. 812 A2d at 456-57. This court believes that the proper way to calculate distortion in a fair apportionment case is to divide the state’s assessment as measured by the apportionment formula ($19,981,757 in Norfolk & Western, 390 US at 326), by the company’s assessment as measured by its desired method ($7,600,000, id. at 322).

17 The Court explicitly based its holding on both the Due Process and Commerce clauses. 390 US at 329-30.

18 That tactic is similar to one adopted by the taxpayer in Norfolk & Western, where the taxpayer compared Missouri’s apportionment with that which would have occurred had the taxpayer never leased fixed assets or rolling stock from Wabash. 390 US at 322 n 4.

19 The 1,883% figure was derived by the court by dividing $12,787,485 (taxpayer’s Oregon taxable income under ORS 317.660 given its actual income from Oregon real and tangible property) by $679,058 (taxpayer’s Oregon taxable income under ORS 317.660 given no income from Oregon real and tangible property).

20 The 0.0381% figure was derived by the court by multiplying 0.2362% (the actual proportion of taxpayer’s combined gross income from all real and tangible property and sales that is composed of its gross income from all real and tangible property) by 16.1179% (the proportion of taxpayer’s gross income from all real and tangible property that is composed of its gross income from Oregon real and tangible property).

21 “Distortion in one factor * * * does not necessarily result in unfair reflection of the business activity in the state; the other two factors may well mitigate the distortive effect of the third, so that, ultimately,” the three factors in combination fairly represent the taxpayer’s business activity in the state. In re Appeal of Merrill, Lynch, Pierce, Fenner & Smith, Inc., 1989 Cal Tax LEXIS 18, at * 8 (Cal. State Bd. of Equaliz., No. 89-SBE-017, June 2, 1989). See also Paris Mfg. Co. v. Commonwealth, 505 Pa 15, 476 A2d 890, 893 (1984) (“We reject the Commonwealth’s assertion that the mere disparity between the magnitudes of the property, payroll, and sales fractions is, in itself, indicative of a failure of the apportionment formula to fairly reflect the loci of business activities.”); Trinova, 498 US at 381 (upholding an apportionment where the payroll factor was 0.2328%, the property factor was 0.0930%, and the sales factor was 26.5892%). In GATX Corp. v. Limbach, 21 Ohio App 3d 59, 486 NE2d 840, 843 (1984), the court required the taxing authority to eliminate the payroll factor when it was much greater than both the gross receipts and property factors because the disparity “put[ ] the entire apportionment formula out of focus.” Although the court adopted that approach under the equitable allocation provision of the Uniform Distribution of Income for Tax Purposes Act, and not the constitution, the decision reflects the reality that distortion in one factor, when combined with other evidence, can indicate an unfair apportionment.

22 Although the department does not make the argument, taxpayer’s Oregon property could theoretically generate value outside Oregon in that the property comprises part of the minimum capital required by other states for taxpayer to do business there. However, the department has already stipulated that taxpayer’s income from Oregon real and tangible property was neither integral nor necessary to taxpayer’s business. Moreover, even had the department not made that stipulation, the apparent gross distortion resulting from application of ORS 317.660, as shown by taxpayer, shifts the burden to the department to bring forth some explanation or evidence “which might offset the devastating effect of the demonstrated discrepancy.” Norfolk & Western, 390 US at 327-28. The department has not done so. Finally, the property factor of ORS 317.660(3) does not propose to measure a capital value for taxpayer’s Oregon property; it is based, not on any value of the property, but only on the amount of income generated from the property.

23 Here, the real estate income and interest factor, although weighted heavily toward Oregon, has a numerator that was admittedly inconsequential to taxpayer’s business and a denominator that was similarly insubstantial. Nonetheless, the three-factor apportionment formula of ORS 317.660 allowed that factor to overwhelm the insurance sales and wage and commission factors, both of which were weighted heavily toward other states and far more significant to taxpayer’s bottom line. The result was that taxpayer’s income apportioned to Oregon was determined largely with reference to the inconsequential real estate income and interest factor, grossly distorting the true sources of taxpayer’s profits. This is an instance of a very small tail wagging a very big dog. A similar thing happened in Norfolk & Western, where the same tail (Wabash’s track mileage in Missouri) gained sudden gravitational force once it was applied to a different, and much larger dog (the taxpayer’s much larger fleet of rolling stock). That also produced an unconstitutional result.

24 It is important to note here that the real estate income and interest factor of ORS 317.660(3) is not a measurement (through cost, value, or other amounts) of income producing property. Rather, it is a measurement of income from property. In this case, where there are no wages or commissions in Oregon, the statutory formula results in an apportionment of net income by way of a combination of two factors both of which are based on income from Oregon sources.

25 Although the Court in Moorman contemplated that some sales are more profitable than others, because some involve less cost, the Court did not imply that any sale might be more profitable than the gross sale price. 437 US at 276. That understanding makes sense because profits equal gross revenue less costs. Nonetheless, the Court has not foreclosed the possibility that some sales may create ripple effects, imbuing them with greater overall value than the net profit derived from their singular occurrence.

26 In that respect, the department is like the taxing authority in Allied-Signal, Inc. v. Director, 504 US 768, 112 S Ct 2251, 119 L Ed 2d 533 (1992). In that case, the parties stipulated that the taxpayer and another company were unrelated and had nothing to do with each other. Id. at 773. That stipulation ultimately proved fatal to the taxing authority’s claim that the taxpayer and the other company were a unitary business. Id. at 788-89.

27 Although decided under OES 314.280, instead of constitutional grounds, the court finds Crocker Equipment Leasing, Inc. v. Dept. of Rev., 314 Or 122, 131, 838 P2d 552 (1992), instructive. That case, like the present one, involved an assertion that use of a three-factor formula ultimately led to an impermissible apportionment to Oregon of income from intangible property administered elsewhere. In Crocker, application of an apportionment was found to yield a “grossly distorted” result where 98% of the taxpayer’s income-producing assets were intangible properties but the property factor of the apportionment formula did not include them. Id. at 131-32. The department argued in that case that the exclusion of intangibles was permissible because income from intangibles was included in the gross revenues factor of the apportionment formula. Id. at 132. The Supreme Court rejected that argument, concluding that such treatment was not sufficient to “correct the extremely disproportionate representation” of taxpayer’s business activity in Oregon. Id. In the present case, the department cannot even make the argument rejected in Crocker because neither the insurance sales factor, nor any other factor, includes income from intangible property; rather, the insurance sales factor is restricted to gross premium income.

28 In 1969 the legislature passed HB 1021, which divided all insurers doing business both within and without Oregon into two groups: domestic insurers, which were subjected to a new taxation regime, and foreign and alien insurers, to which the gross premium tax of ORS 731.816 applied. Or Laws 1969, eh 600, § 7. The new regime was codified at ORS 317.199 and subsequently moved to its present location at ORS 317.660. It remained unchanged in all material respects until 1995, when the legislature eliminated the distinction between foreign and alien insurers on the one hand, and domestic insurers doing business both within and without Oregon on the other, subjecting them all to the regime of ORS 317.660. Or Laws 1995, ch 786, §§ 12,17.

29 Similarly, in NCR Corp. v. Tax Comm’n, 304 SC 1, 402 SE2d 666, 674 (1991), the South Carolina Supreme Court ordered the trial court to determine a new tax liability for the taxpayer based on a corrected formula, and, if the trial court found that apportionment unconstitutional, to “order the new tax amount to be assessed and refund granted.” In that case, however, the Supreme Court did not order the trial court to apply a new formula, but rather to apply the proper formula as mandated by statute. At any rate, the trial court never ordered a new assessment or refund because it found the new apportionment constitutional. See NCR Corp. v. Tax Comm’n, 312 SC 52, 439 SE2d 254 (1993) (affirming the trial court).