Nelson, Justice.
The relator and taxpayer filed its income and franchise tax return for the year 1947 according to a separate accounting system previously used by it. The commissioner of taxation for the state of
Minnesota issued a deficiency letter on August 3,1951, and an order on August 31, 1951, in which he disallowed the taxpayer’s accounting method and assessed a deficiency for 1947. An appeal was taken from this order to the state board of tax appeals, which affirmed the commissioner’s decision to disallow the taxpayer’s separate accounting system and to assess the tax on the basis of the three-factor formula provided for in M. S. A. 1945, § 290.19, subd. 1(2) (b).
A writ of certiorari was issued to review the decision.
The taxpayer is a Missouri corporation engaged in the retail and wholesale merchandising of automotive parts and accessories and other merchandise on a nationwide basis. During the calendar year 1947, taxpayer operated 257 retail stores in 30 states and the District of Columbia. This included eight such stores in the state of Minnesota. It also served 1,904 dealer stores of which five to seven were located in Minnesota.
All general, executive, and management functions of the taxpayer are performed at its home office in Kansas City, Missouri, with the exception of certain management functions which are performed at 16 different division warehouse offices, none of which were located in Minnesota. The centralized purchasing is done by expertly trained personnel whose department is located in Kansas City, Missouri. All the merchandise sold by it is purchased from manufacturers and jobbers by the company’s central purchasing department and initially shipped from the manufacturers and jobbers to the warehouses of the taxpayer and from there to the company stores or sold and delivered to the dealer stores. No segregation of the merchandise between company stores and dealer stores is made either in the purchasing or the storing in the warehouses. It so operated in 1947. The Minnesota dealer stores and retail stores were serviced mainly by the Dubuque, Iowa, warehouse, but in ordering merchandise they would either send their orders to the home office in Kansas City, Missouri, or to the warehouse in Dubuque, Iowa.
From time to time schools are set up by the company for the purpose of instructing the personnel of the company’s retail stores and the dealer stores in the principles of merchandising and selling. Its
representatives are sent to contact and interview persons interested in establishing dealer stores. Such representatives survey the trade territory, and when a dealer store is set up they help the dealer by advising him how to get started. These stores are strategically located so as to serve the customer and capture the market. Thereafter a representative calls on the dealer from time to time to see if the merchandise sold to the dealer is satisfactory and he also makes his service available in their merchandising problems. He makes it his business to show them good merchandising procedures.
The merchandise furnished to both the company retail stores and the dealer stores is the same, and the stores on the whole are more or less uniform in each state. Any person who sees the sign on a dealer store knows that it is a store where the Western Auto Supply Company’s line of merchandise can be purchased. Prices as to articles sold are set and the policies determined by the home office. Certain articles carry the Western Auto brand name. For instance, the Wizard batteries are made by the manufacturer under the Western Auto label and according to Western Auto specifications, and they are sold at all retail and most dealer stores. This battery is guaranteed, and it may be taken into any Western Auto store anywhere for adjustment. The same is true as to Davis tires. Other items carrying brand names are Truetone radios and certain oils and greases. All directive activities point to the reasonable conclusion that the wholesale side of the business helped the retail side of the business and vice versa and that both the wholesale side and the retail side contributed to the existence and the operational success of the entire organization as a centralized business operation.
The taxpayer, in filing its 1947 Minnesota corporation income tax return, apportioned to Minnesota on the basis of separate accounting a net taxable income of $22,518.20, which amounted to approximately .812 percent of total taxable income.
The commissioner, upon auditing said return for 1917, apportioned taxpayer’s income upon the statutory three-factor formula of property, payroll, and sales. In arriving at the sales factor of the formula, the commissioner included in the denominator (the bottom figure of the fraction) all sales by the taxpayer to dealer stores and sales by its retail stores, but in the numerator (the top figure of the fraction) included only sales made by its Minnesota retail stores and did not include sales to Minnesota dealer stores.
The figure used, namely $2,056,919.91, over $121,395,031.32 resulted in a ratio of 1.69113 percent. The other factors in the formula used by the commissioner were Minnesota tangible property ($791,635.16) as compared to total tangible property ($16,199,120.22) and Minnesota payroll ($289,316.11) as compared to total payroll ($11,528,029.18). This resulted in a property factor of 1.72002 percent and a payroll factor of 1.99111 percent. The arithmetical average of the three factors produced a percentage of 1.80196 percent which, when applied to the total taxable net income of taxpayer ($2,772,700.72) resulted
in apportioning to Minnesota after certain allowances and credits a taxable net income of $17,836.18.
The board of tax appeals stated in part in its findings that the commissioner and the taxpayer each conceded that the sales formula alone does not result in a proper reflection of Minnesota taxable net income. The record indicates that the state and the taxpayer agree that the sales formula, § 290.19, subd. 1(2) (a), prescribed as the primary or specific method, does not properly reflect taxable income assignable to Minnesota and that it was therefore rejected by both sides. This agreement, even though seemingly harmonious, proceeds from different premises and somewhat opposite conclusions. The taxpayer contends that it was rejected because the sales formula alone allocated too much income to Minnesota. That contention of the taxpayer is not admitted by the state. The commissioner contends that the tax as assessed by the three-factor formula employed by the state is higher than that allocated by the use of the sales formula alone; that it is the practicable one here and the one which will properly and fairly reflect Minnesota income as required by the statute; and that the sales or single-factor formula does not here properly reflect taxable net income assignable to the state.
Other contentions of the taxpayer are that its business does not constitute a sufficiently homogeneous unit to be considered a unitary business and therefore the unit rule of taxation is not applicable; that there is no reasonable correspondence between the unit of measure and the subject to be measured; that the sales formula allocates to Minnesota income which, as a matter of economic demonstration, is earned elsewhere; that Minnesota income is properly identified by the taxpayer’s separate accounting; and that any attribution of additional income to Minnesota under the auspices of apportionment, whether by the single sales formula or the three-factor formula, if applied, fails to “properly and fairly reflect” Minnesota income within the meaning of § 290.19, subd. 1(2) (b).
Further, taxpayer contends that the commissioner’s constitutional duty is to avoid the taxation of extraterritorial income or illegal exactions upon interstate commerce; that it is his statutory duty (which is more exacting) to make a determination which “will
properly and fairly reflect” income assignable to Minnesota; that the use of the multiple formula (or any unit formula) is not practicable in the instant case within the meaning of § 290.19; and that its use in the premises is unreasonable because the subject matter is not suitable for unit-rule apportionment. Also, that the multiple formula (or the sales formula if used) reaches income not derived from business within Minnesota and therefore is inapplicable in either case and that its separate or segregated accounting methods correctly reflect Minnesota taxable income, whereas the application of the multiple formula does not. The taxpayer, by an analysis of what it claims are the pertinent facts, discusses the factors governing choice between the multiple factor and the separate or segregated accounting method and again urges that, under the multiple formula, there is no sufficient correspondence between subject and measure and hence that the multiple formula used by the commissioner has no tendency to effect a segregation of the net income from its wholesale division sales, no part of which, it claims, is taxable in Minnesota. It thus contends that the application of the three-factor formula has no tendency to reflect a proper segregation of net income within the retail division; and that the multiple formula siphons extraterritorial “buying profit” and “administrative profit” into Minnesota, and in so doing violates the rule of Hans Rees’ Sons v. North Carolina, 283 U. S. 123, 51 S. Ct. 385, 75 L. ed. 879, and ignores the rule laid down by the state board of tax appeals in some of its own decisions, citing Allied Building Credits, Inc., Docket No. 254, and Crane Company of Minnesota, Docket No. 320, stultifying consistent administrative practice of many years’ standing.
In the Hans Rees case the United States Supreme Court ruled that the application by North Carolina of a single-property-factor formula to the unitary net income of a company, whose manufacturing was done in North Carolina but most of whose selling activity was done outside the state, allocated to North Carolina income which was earned elsewhere and was therefore, in its application to that business, an unconstitutional method of apportionment. The defect pointed out by the court in that case was not in the use of the unit-formula method but in the use of the wrong formula. The record in
the instant case does not offer any proof in fact establishing that the three-factor formula used allocates to Minnesota any more than its proper share of the taxpayer’s net income.
Another case cited by the taxpayer as controlling in its favor is Montgomery Ward & Co. v. State Tax Comm. 151 Kan. 159, 98 P. (2d) 143. That case did not involve the validity of the application of a formula method of apportionment. It merely involved the validity of one of the • steps in Montgomery Ward & Company’s separate accounting method, the state of Kansas having taxed the net income of the Kansas stores of this company as it had been determined by the company’s accounting method. The court in that case did not say that the business was not a unitary business nor did it say that a formula method of apportionment would not be valid.. We are unable to find where the holding in that case sufficiently supports the taxpayer’s position. We are inclined to the view that the dictum in that opinion supports the commissioner and the board in the case at bar.
Cases representative of the action of the court on issues similar to those before us in the instant case are: S. S. Kresge Co. v. Bennett (S. D. N. Y.) 51 F. (2d) 353; Butler Bros. v. McColgan, 315 U. S. 501, 62 S. Ct. 701, 86 L. ed. 991; North American Cement Corp. v. Graves, 243 App. Div. 834, 278 N. Y. S. 920;
Id.
269 N. Y. 507, 199 N. E. 510;
Id.
299 U. S. 517, 57 S. Ct. 311, 81 L. ed. 381 (facts found at 17 Cal. [2d] 673, 111 P. [2d] 339); Edison California Stores v. McColgan, 30 Cal. (2d) 472, 183 P. (2d) 16.
Finally, the taxpayer sums up its position as follows: That the multiple or three-factor formula is not “practicable” within the meaning of § 290.19 upon the ground that it is unreasonable.in terms of constitutional doctrine, the commissioner having a constitutional duty to avoid taxation of interstate commerce in violation of the commerce clause, citing McLeod v. Dilworth Co. 322 U. S. 327, 64 S. Ct. 1023, 88 L. ed. 1304, and is not a proper and fair reflection of Minnesota income as required by § 290.19, since any action of the commissioner resulting in a tax on any part of the income of the wholesale division is also a violation of U. S. Const. Amend. XIV
because of extraterritoriality; that every circumstance which required the commissioner in the first instance to review the sales formula requires, a fortiori, the rejection of a multiple formula, again urging that the taxpayer’s separate accounting method reflects Minnesota taxable income and that the taxpayer’s system of separate accounting is to be preferred to the multiple formula; and that, although the commissioner’s determination has in the first instance prima facie validity, the aforesaid reasons advanced by the taxpayer establish that it is no longer entitled to recognition. The taxpayer’s contention when reduced to simple terms is that the use of the three-factor formula in this instance by the commissioner and affirmed by the board of tax appeals allocates to Minnesota income which was earned elsewhere, namely, the income from its wholesale business, none of which is conducted in Minnesota, and some income from its retail business not attributable to the Minnesota retail stores.
The taxpayer admits the validity of the use of the three-factor formula in the case of a homogeneous business unit but contends that its business as operated by it is not sufficiently homogeneous to justify applying the multiple formula. It bases its claim of the lack of homogeneity on the premise that no part of its wholesale business is conducted in Minnesota and on the further fact that there is no uniformity of operating experience in its retail stores; therefore, that § 290.19, which requires that the method of allocating income to Minnesota as adopted by the commissioner must properly reflect income earned in Minnesota, does not justify the use of the multiple formula. The taxpayer’s contention, however, that none of its wholesale business is conducted in Minnesota, even if this were the fact and it were also the fact that the operating experience of the retail store is not identical or the same everywhere, becomes irrelevant as far as the question of homogeneity or unitary character of the taxpayer’s business is concerned. A multistate business is a unitary business when the operations conducted in one state benefit and- are in turn benefited by the operations conducted in another state or states. The test to be applied in determining whether a busi
ness is a unitary one is based upon the following inquiry: Is the operation of the business within the state “dependent upon or contributory to the operation of the business outside the state”? See, Altman & Keesling, Allocation of Income in State Taxation, p. 101. It is not required, as the taxpayer contends, that the taxpayer have a portion or all parts of its business in the taxing state before the state may apply the formula method of apportionment. Various parts of its business may be carried on exclusively in different states without destroying its unitary nature if these parts are of mutual benefit to one another. Cases so holding are very numerous. The leading ones are: Underwood Typewriter Co. v. Chamberlain, 254 U. S. 113, 41 S. Ct. 45, 65 L. ed. 165; Bass, Ratcliff & Gretton, Ltd. v. State Tax Comm. 266 U. S. 271, 45 S. Ct. 82, 69 L. ed. 282; Butler Bros. v. McColgan, 315 U. S. 501, 62 S. Ct. 701, 86 L. ed. 991.
The commissioner, applying the multiple formula and acting in accordance with § 290.19, subd. 1(1), did not include the company’s wholesale sales to the Minnesota dealer-owned stores in the numerator of the sales factor of the formula, so the question of whether it would be constitutional to do so is not at issue in this case. The McLeod case does not hold that it would be unconstitutional to do so. Even if the actual sales did not take place in Minnesota so that this state could not constitutionally impose a sales tax on wholesale sales, Minnesota nevertheless would still be justified in including those sales in the numerator of the sales factor in the formula for apportioning the net income of the taxpayer, if conducting a unitary business, because the effect of so doing is not to tax a sale but simply to cause to be reflected in the formula applied certain selling activity of the taxpayer in Minnesota in connection with its dealer-owned stores which contributed to the earnings of the taxpayer’s total unitary net income. This is indicated by the activity set forth in the statement of facts. The sales factor is used in the formula applied in order to give weight to the activity of the taxpayer in obtaining customers, and, even though the activity of the taxpayer in Minnesota would not be sufficient ground for the imposition of a sales tax on these wholesale sales, that would not mean that the activity in laying the foundation for and promoting those
sales could not be found to justify their inclusion in the numerator of the sales factor in the formula used in apportioning the taxpayer’s unitary net income. Compare the statement in Maxwell v. Kent-Coffey Mfg. Co. 204 N. C. 365, 370, 168 S. E. 397, 400, 90 A. L. R. 476, 480, affirmed, 291 U. S. 642, 54 S. Ct. 437, 78 L. ed. 1040;
“* * * The bare fact of sale produces no income. It is merely the act by which the income is captured; the capital, the organization or efforts which produce the sale, are the things to be considered in ascertaining the amount of income to be credited to the sale.”
The test of whether a business is unitary is whether its various parts are interdependent and of mutual benefit so as to form one business unit rather than separate business entities and not whether the operating experience of the parts is the same in all places. If the taxpayer’s objection to the method of apportionment used, based on the fact that the operating experience of its retail units is not uniform throughout the nation, were valid, then no business could be found to be unitary and the formula method applied would always be invalid. Furthermore, the three-factor method of apportionment reflects these variables and compensates for them.
The constitutional questions raised by the taxpayer find no support in the record.
The taxpayer, in its analysis of the sources of its profit, treats its business as if it were made up of independent, separate business units. The fact remains that the buying profit and the administrative profit which its business reflects were not created solely by activities in the state of the taxpayer’s home office or in states where its warehouses are located. It was created by the operation of the entire business unit, through the co-ordinated and standardized activity of numerous stores throughout the country which made possible the central purchasing, the central management, .the warehousing as carried on, and the advertising methods adopted. The stores in Minnesota contributed in part to make all this possible, and the multiple-formula method simply allocates a fair share of the profit to Minnesota, a share corresponding to the contribution of the taxpayer’s Minnesota operation (measured by its Minnesota property, payroll, and sales activity) to the existence and operation of the taxpayer’s entire business unit.
Applying this test, any argument that the taxpayer’s business is not a unitary one lacks persuasion. A cursory review of the facts of the case indicates quite conclusively that the taxpayer’s retail sales and wholesale sales stand in the relationship of reflected dependency, mutual contribution, and mutual benefit.
Our statute leaves the use or rejection of the separate accounting method to the discretion of the commissioner initially, subject to the requirement that the method he adopts will properly and fairly reflect taxable net income assignable to Minnesota. It cannot be questioned that the proposition that multiple-formula method of apportionment adopted by the commissioner in the instant case does properly and fairly reflect the taxable net income earned in a given state by a unitary business finds support in the cases and among the leading authorities of accounting and allocation of income in state taxation. This is even conceded by the taxpayer in its argument. Taxpayer’s argument, however, is that its business is not unitary and hence not a proper subject for the application of the formula method of apportionment which the commissioner applied here. This contention of the taxpayer, however, is .not sufficiently borne out by its activities as shown by the record upon which the
•commissioner acted in rejecting the separate accounting method nor by the evidence upon which the board of tax appeals affirmed the commissioner’s action taken therein.
The authorities, we believe, preponderate in holding that the separate accounting method is inherently incapable of accurately apportioning the income of a unitary business. See analysis in Altman & Keesling, Allocation of Income in State Taxation, pp. 89 to 97. This, a recognized authority, after a discussion of the principles of allocation of business income, sums the situation up as follows on p. 97:
“For the above reasons it is an inescapable conclusion that the separate accounting method is wholly unsuited to the proper apportionment of the income of a unitary business. It is believed that the only proper method for making the apportionment is to employ a formula which gives weight to the various factors responsible for the earning of income, such as property, which reflects either the capital investment or the general size and importance of the business, payroll, which indicates the value to the taxpayer of the services' of its employees, and sales, which give weight to the activity of the taxpayer in obtaining customers without which, of course, the business would not function.”
While the taxpayer’s separate accounting method may be quite suitable for taxpayer’s business purposes, it nevertheless seems definitely inadequate for the purpose of attributing to Minnesota with a fair degree of accuracy that portion of the taxpayer’s income which was earned by its Minnesota operations for the reason that it rests in part on estimates and resorts to arbitrary allocation formulae.
The commissioner makes no claim of impropriety with respect to the taxpayer’s separate accounting system. It must be recognized that even the so-called separate accounting system cannot be kept with, exactness. In actuality no time records were kept, according to the evidence, and mental records were often relied upon. The taxpayer’s system of separate accounting divided its total business
into wholesale and retail divisions and, wherever possible, items were credited to the division to which they applied. The taxpayer’s separate accounting showed a $35,501,155.81 gross profit after deducting cost of goods sold from total corporate sales. It then showed $19,363,753.59 in direct charges against either the wholesale or the retail division as the particular item required. This resulted in what is referred to as a direct operating profit of $16,137,402.22. It showed an ultimate net income per federal return of $7,237,058.12. The difference between the direct operating profit and the net income per federal return is $8,900,344.10. This amount could not be charged directly to either the wholesale or the retail divisions and was therefore allocated by the taxpayer between the two on the basis of actual operating experience.
From mental records made and estimates applied, salaries were apportioned 60 percent to retail and 40 percent to wholesale. The chairman of the board, while presiding at meetings, had his salary allocated between retail and wholesale sales. While the assistant treasurer of the company was in Minnesota in connection with the taxpayer’s business, his salary was apportioned 60 percent to retail and 40 percent to wholesale operations. In 1947, 60 percent of administrative salaries was allocated to retail sales and 40 percent to wholesale sales. These allocations suggest a recognition on the part of taxpayer’s management that the business overall was a unitary business.
Approximately 57 percent of the total corporate sales for 1947 was made through the company’s 257 retail stores, of which eight were in Minnesota. These stores produced about 72 percent of the gross profit for 1947. The division referred to as wholesale sold to dealer stores as opposed to company-owned and operated units. Sales to the 1,904 (five to seven operating in Minnesota) dealer stores represented about 43 percent of the total corporate sales for 1947 and produced about 28 percent of the gross business.
Both total expenses itemized and direct expenses shown reveal a substantial amount of indirect expenses allocated. There appear the following allocations: over $1,000,000 in advertising, over $325,000 in travel expense, over $350,000 in rent, over $664,000 in profit-
sharing and retirement contributions, and over $337,000 in miscellaneous expenses. These substantial items are allocated by the use of some sort of apportionment formula. While the record does not disclose all of the various formulae used, in addition to salaries of officers, depreciation on tools and other items were allocated 60 percent to retail and 40 percent to wholesale; legal expense was assigned two-thirds to retail and one-third to wholesale; tabulating department expense was assigned fifty-fifty; general catalog expense was assigned one-third to retail and two-thirds to wholesale; while profit sharing, interest expense, stationery, and supplies for accounting were divided 75 percent to retail and 25 percent to wholesale. Other factors were used for warehouse operations, indicating that taxpayer treated seven wholesale outlets as the equivalent to one retail outlet. When the taxpayer comes to determining net income considered assignable to Minnesota stores, it is compelled to again resort to the use of allocation formula because of the unitary nature of its business. These matters are of course fully covered in the record, and no further listing of the same will be made here.
Suffice it to say that the taxpayer’s allocation formula must of necessity become somewhat arbitrary at points and relatively inflexible because so often based upon mental records and estimates, hardly to be avoided even in separate accounting when a business becomes as definitely unitary as this one has. While many of these may have been close estimates, nevertheless they remain mere estimates. Under all of these circumstances there would appear to be ample evidence to support the board’s finding that the wholesale part of the taxpayer’s business was part of one unitary operation and that the wholesale gross sales should be included with the retail sales in the total sum to be allocated by the múltiple formula as a part of its unitary income. Cases holding that the wholesale sales of a company conducting both wholesale and retail merchandising operations under conditions similar to those in the instant case are part of the company’s unitary income, though involving a different point of law, are: State v. Gamble-Skogmo, Inc. 63 Idaho 265, 120 P. (2d) 630; Bedford v. Gamble-Skogmo, Inc. 104 Colo. 424, 91 P. (2d) 475.
The law applicable on this appeal is M. S.
A.
1945, § 290.19, dealing with net income to be allocated. Subd. 1 deals with the question of apportionment where, as here, a business is conducted partly within and partly without the state. Subd. 1(1) deals with a business consisting “of the manufacture in Minnesota or within and without Minnesota.” It is not applicable here since the taxpayer is not a manufacturer.
In determining taxable income assignable to Minnesota of a non-manufacturing business carried on partly within and partly, without the state, § 290.19, subd. 1(2) (a, b), are the controlling provisions of the statute. The single-factor formula is prescribed by this section as the primary one, and it must be given first consideration and applied unless as a method it does not properly and fairly reflect taxable income assignable to Minnesota. Stronge & Lightner Co. v. Commr. of Taxation, 228 Minn. 182, 36 N. W. (2d) 800. Under § 290.20 it is provided that the commissioner is to prescribe the methods, subject to the right of the taxpayer to apply for the application of some other method under the statute by taking the appropriate procedure therein provided. It was held in the Stronge & Lightner Co. case that, if an appeal is taken by a corporate taxpayer carrying on a nonmanufacturing business partly within and partly without the state to the board of tax appeals from a decision of the commissioner of taxation that the single-factor formula adopted by a taxpayer did not properly reflect income assignable to Minnesota for the purpose of computing the tax and that the three-factor formula should have been applied, the taxpayer had the burden of proving that the formula adopted by it properly reflected income assignable to Minnesota but that the taxpayer should not on such appeal be required to .shoulder the additonal burden of showing that the alternative method' adopted by the commissioner did not fairly reflect income assignable to Minnesota.
In this connection it should be noted, although heretofore mentioned, that in the sales factor used by the state in the three-factor formula the numerator (the top figure) in the fraction dues not include sales to the Minnesota dealer stores (five to seven in 1947) . However, in the denominator (the flgure below the line) the total
sales-are included; namely, those made at retail as well as those made to the dealer stores. The obvious effect of this procedure is to reduce the percentage of income allocable to Minnesota. This is made necessary by the holding in The Maytag Co. v. Commr. of Taxation, 218 Minn. 460, 17 N. W. (2d) 37, since the sales made to dealer stores were made by, through, and from an office outside the state, and under our law these are not construed as Minnesota sales.
Under the application of the three-factor formula, Minnesota sales of $2,056,919.91 are 1.69113 percent of the total sales of $121,395,031.32. This is true despite the exclusion from Minnesota sales figure as stated above of all sales made to the five to seven Minnesota dealer stores. The state contends that this sales percentage of 1.69113 does not differ from the 1.80196 percent used by the state to such degree that it can be said to produce an unfair or improper allocation and that in the light of these figures there is nothing arbitrary or capricious in the use of the three-factor formula.
There must of necessity be some approximation and estimate, some margin of error, in any system used to allocate income among several states, including any system of separate accounting, even that used by the taxpayer herein.
Taxpayer’s income arrived at by separate accounting equals .812 percent of the total apportionable income of $2,772,700.72. Since the facts point to the .taxpayer’s business as a unitary one, it is fair to assume that the commissioner selected the three-factor formula as practicable and the applicable one for the reason that the figure arrived at by the taxpayer by separate accounting was disproportionately small in relation to the percentage of sales made in Minnesota as compared to total sales made everywhere, and since the taxpayer’s business is clearly a unitary one, under leading authorities, the multiple formula does, when applied, properly and fairly reflect taxable net income assignable to the state.- In this connection it must not be overlooked that the sales percentage is the primary statutory formula prescribed in this state for allocating income from a multistate, nonmanufacturing business, but it is not made manda
tory.
.While this significance is not completely nullified by the fact that here sales concededly do not properly reflect assignable income, neverthless, it is the commissioner’s prerogative and duty to select the applicable formula as it appears to him in the first instance. When this is done, the order of the tax commissioner comes to the board of tax appeals clothed with prima facie validity deriving from specific statutes.
The taxpayer takes the position that the commissioner’s obligation is to choose “the method which constitutes the better ‘reflection’ of Minnesota income.” The taxpayer also states in speaking of the three-factor formula that the question is “more simply, whether it is better suited to reach a fair and proper result than separate
accounting.” This, however, is not a correct statement of the law which governs this case.
The commissioner here has used one of the two methods which the law says he “shall use” if the sales factor will not properly reflect taxable income. Having done so, the commissioner’s order comes to the board of tax appeals with the benefit of prima facie validity, which simply means that the burden is on the taxpayer to show not, as contended, that its method of separate accounting is a “better reflection” but that the commissioner’s three-factor formula does not “properly reflect” an equitable allocation of income. The two are not the same.
The fact is that, while an allocation of income arrived at by the three-factor formula might not meet the “better reflection” criteria, which taxpayer erroneously assumes is the test, it might well be adequate under the “properly and fairly reflect” criteria established by statute and applied by the commissioner.
Mathematical certainty is simply not possible in this difficult field. Any method of apportionment, separate accounting as well as every possible formula, contains estimates, and, in varying degrees, there is applied the exercise of human judgment in the allocation of costs and charges.
It is obvious that absolute accuracy is not possible by any method which may be used in allocating to Minnesota its proper share of total corporate income derived from a multistate activity with integrated buying, selling, credit, advertising, and general management policies. Therefore, in such a case all that is required is that the end result assign “to Minnesota a fair, honest, and equitable part of total corporate income measured in terms of corporate activity within Minnesota.”
The commissioner of taxation must of course guard against being arbitrary or capricious. When the provisions of the statute are clear, he must follow the statute. Having followed and. complied with the statute, his determination must stand, under the prima facie validity rule, even though his method may not have produced the most equitable or the most accurate reflection of apportionable
income, unless the taxpayer has sustained the burden of showing either the incorrectness or the invalidity of the tax.
Reasonable approximation, not meticulous precision, if arrived at will.measure the fairness of any formula to allocate income from a multistate activity “within the limits of fair human judgment.”.
The case of Stronge & Lightner Co. v. Commr. of Taxation, 228 Minn. 182, 36 N. W. (2d,) 800, came before this court upon relation of the commissioner of taxation to review a decision of the board of tax appeals. The court on review emphasized and followed the general rule stated in State ex rel. Inter-State Iron Co. v. Armson, 166 Minn. 230, 232, 207 N. W. 727, 728, to the effect that it is well settled that, in reviewing an order or determination of an administrative board, this court will go no further than to determine: (1) Whether the board kept within its jurisdiction; (2) whether it proceeded on a correct theory of the law; (3) whether its action was arbitrary, oppressive, or unreasonable and represented its will and not its judgment; and (4) whether the evidence was such that it might reasonably make the order or determination in question.
Stronge & Lightner Co. was a Minnesota corporation which had its home office in St. Paul. It owned and operated a chain of 48 retail stores which sold women’s apparel. Twenty-two of the stores were located in Minnesota, the remainder in nine other states. The merchandise which it sold was largely purchased by it in the east and shipped to St. Paul, where, through the home office, it was processed and stored in a warehouse until shipped to the retail stores. From the beginning of the business in 1936 until 1943, the company had used the single-sales-factor formula in apportioning its total net income to Minnesota for tax purposes. In 1943, the commissioner determined that the single sales factor did not properly reflect the income earned by the home-office activities and applied the three-factor formula instead. The three-factor formula ascribed 46.588 percent of the company’s net income to Minnesota; the single-sales faetor, 37.3738 percent; and a separate accounting method maintained by the company, 25.054 percent. The application of the three-
factor formula by the commissioner added only an additional tax of $268.05.
On appeal to the state board of tax appeals and a consideration of the matter
de
novo, the commissioner was reversed. The board determined that the application of the single-sales-factor formula fairly reflected the income earned in Minnesota. In its memorandum accompanying its decision, the board stated that its duty was “to determine whether under the Sales formula as applied to this Taxpayer there results in assigning to Minnesota a fair, honest, and. equitable part of total corporate income measured in terms of corporate activity within Minnesota.” The board stated its duty and its conclusion, but it offered little, if anything, upon the facts to support its conclusion. Regarding the fact that the single-sales-factor-formula method resulted in assigning to Minnesota in that case approximately 50 percent more taxable income than the separate accounting method did, the board stated that “Due to the difficulty of evaluating the importance of the numerous elements in the income-producing chain and of determining the proper reflection of each in the various formulas, we are unable to state with any degree of accuracy that this increase is or is not due to a reflection of home office activity or some other element.” It said further that “The requirements of this decision, however, do not demand such, an evaluation. All that is required is that the result show a proper reflection of Minnesota taxable net income.”
On the appeal by the commissioner from the decision of the board, this court said that the question before it was in the main whether the decision of the board was supported by the evidence, and this court held that it was. In evaluating the weight of that decision as regards the scope to be given to the primary method of apportionment, it must not be overlooked that this court was under the impression that the board had determined that the difference in the result of the two accounting methods did reflect the activities of the taxpayer’s home office and warehouse. However, in the last sentence of the board’s memorandum it refers to the difference in result in the two accounting methods as if it were in fact basing its finding
on it. This court said, in approaching its conclusion in that decision (228 Minn. 189, 36 N. W. [2d] 804):
“All questions of law and fact are to be viewed by us in the light of the rule that a decision of the board of tax appeals will not be disturbed if it has any reasonable basis in law.”
This court then said (228 Minn. 196, 36 N. W. [2d] 808):
“When we come to a nonmanufacturing business, the legislature, by § 290.19, subd. 1(2) (a), has first prescribed the single-factor formula. If the statute had gone no further than this paragraph, it also would be mandatory. However, the legislature thereafter added paragraph (2) (b), which provides that if paragraph (2) (a) does not properly reflect taxable net income assignable to the state there shall be used, if practicable and if such use will properly and fairly reflect such income, the three-factor formula or the separate or segregated accounting method.”
Recognizing the rule that a decision of the board of tax appeals will not be disturbed if it has any reasonable basis in law, this court nevertheless made this observation (228 Minn. 200, 36 N. W. [2d] 809):
“Obviously, it is impossible to lay down any definite rule that will apply to all cases. The facts in each case must determine, within the limits of fair human judgment, which formula should be applied.”
Had this court been charged with finding the facts in that case in the first instance, there is no certainty that it would have found upon the facts as the board did.
If the specific formula set forth in § 290.19, subd. 1(2) (a), does not properly reflect the income assignable to Minnesota in a given case, then it seems clear and obvious that the legislature intended that the commissioner should then be vested with the discretion initially of using the multiple formula or the separate-accounting method under subd. 1 (2) (b).
The application made of the single-factor formula or the three-factor formula or the separate accounting method must be made, as the various provisions of § 290.19 provide, according to the facts
of each particular case, and its application is not necessarily limited to what might be termed exceptional cases, as we inadvertently said in the Stronge & Lightner Co. case, there being no such designation of cases in the provisions of the statute.
The problem of allocation, however, is always one of securing for the state a fair, honest, and equitable part of the total corporate income measured in terms of corporate activity within the state. If, as in this case, it becomes upon the facts a question as to whether the three-factor formula or the separate accounting method is to be used, there must first be determined when may that portion of a business operated within the state be considered separate and under what condition does it become a portion of a unitary business conducted within and without the state.
No one in the instant case has contended for the application of the single-factor formula. The dispute in this case arises over the application of the three-factor formula and the separate or segregated accounting method. The commissioner applied the three-factor formula. The state board of tax appeals, upon a
de novo
consideration of the facts submitted, affirmed the commissioner’s use of the three-factor formula. The sole question of serious consequence here is whether the order and decision of the board is justified by the evidence upon which it made its findings. We are dealing here with apportionment of income which we conclude is income derived by a unitary business operated both within and without the state. The record in this case as a whole justifies us in the belief that the formula adopted by the commissioner and approved
de novo
by the state board of tax appeals does in some reasonable measure and in fact give weight to the different factors responsible for the earnings of the income of the taxpayer and, therefore, that the multiple formula applied does properly and fairly reflect the taxpayer’s income from its entire business among the various states in which it is conducted, including Minnesota. The taxpayer has the burden of overcoming the presumption that the commissioner’s order, which the board affirmed, is based on a formula which is practicable and which properly and fairly reflects income assignable to Minnesota. The taxpayer’s burden, which he must assume in
order to show that the commissioner’s order and the board’s affirmance thereof was not justified by the evidence, is not satisfied by a mere showing that separate accounting allocated less income to Minnesota and therefore produced a lower tax. Of course, this position is but the converse of the principle of the Stronge & Lightner Co. case that (228 Minn. 197, 36 N. W. [2d] 808) • “The test is not which method will produce the greatest revenue for the state of Minnesota.”
Under the circumstances here, even though the primary formula was not applied or contended for by the taxpayer, the commissioner applying the three-factor formula, rejecting the taxpayer’s separate accounting method, we conclude that the board’s determination in affirming the commissioner in the application of the multiple formula as properly and fairly reflecting the net income of the taxpayer assignable to Minnesota is justified by the evidence and must be affirmed.
Affirmed.