Missouri Pacific Railroad Company v. the United States, 392 F.2d 592 (Ct. Cl. 1968)

This case was referred to Chief Trial Commissioner Marion T. Bennett with directions to make findings of fact and recommendation for conclusions of law under the order of reference and Rule 57 (a). The commissioner has done so in an opinion and report filed on June 23, 1967. Plaintiff and defendant except in part to the commissioner's opinion and findings^*  and the case has been submitted to the court on oral argument of counsel and the briefs of the parties. Since the court agrees with the commissioner's findings, opinion, and recommended conclusion of law, as hereinafter set forth, it hereby adopts the same as the basis for its judgment in this case. Therefore, plaintiff is entitled to recover and judgment is entered for plaintiff with the amount of recovery to be determined pursuant to Rule 47(c).

OPINION OF COMMISSIONER

BENNETT, Chief Commissioner:

This is a suit by the taxpayer for the recovery of excess profits taxes paid by taxpayer's predecessor for the year 1950. The sole issue remaining for determination is defendant's setoff of $67,309.02, arising by virtue of the defendant's treating taxes paid by taxpayer in 1950 to the Republic of Mexico as a deduction from gross income under section 23(c) of the 1939 Internal Revenue Code, ch. 2, § 23(c), 53 Stat. 12, rather than as a foreign tax credit against the United States tax liability under section 131(a), 131 (b), or 131(h) of the Code, as amended, ch. 619, § 158, 56 Stat. 856. This issue presents two separate questions: (1) Whether the tax imposed by the Republic of Mexico is an income tax or a tax in lieu of an income tax within the meaning of section 131 of the 1939 Code so as to entitle taxpayer to a foreign tax credit, and (2) if the Mexican tax is either an income tax or a tax imposed in lieu of an income tax, what is the amount which taxpayer may take as a foreign tax credit for 1950? For reasons hereinafter shown, it is concluded that taxpayer is entitled to the foreign tax credit. The amount of the credit is determined in this opinion.

Taxpayer is a common carrier by rail, and operates in interstate commerce subject to the jurisdiction of the Interstate Commerce Commission, hereinafter referred to as the I.C.C. One aspect of this commerce of particular relevance to this case is the freight car interchange system. Under this system, railroad cars owned by taxpayer were delivered, by interchange, to other railroads, for use by those other railroads on their lines.

Under the equipment interchange rules established by the Association of American Railroads, the using railroads paid a daily rental to the taxpayer known as "per diem," for the period during which the taxpayer's cars were located on their lines. Similarly, when cars of another railroad, referred to as "foreign cars," were located on the taxpayer's lines, taxpayer was required to pay these railroads the per diem.

The Mexican railroads also participated in this interchange system and subscribed to the interchange rules. Freight cars of the taxpayer were transported to the Mexican border, at which point the receiving Mexican railroad would hook on to them with one of its own locomotives and haul them away.

In 1950, the per diem to taxpayer for its freight cars in Mexico was $2.05. Taxpayer's reported income for the year from the rental of freight train cars to railroads in Mexico was $330,251, representing payment for 161,098 freight car days in Mexico at the stipulated rate of $2.05 per day.

In 1950, taxpayer, in conjunction with Mexican railways, operated a passenger train service into Mexico. Taxpayer received from Mexican railroads a rental income of 10 cents for each mile each car was operated on the lines of railroads in Mexico, resulting in a total income for the year from this source of $47,510. Taxpayer operated 781 passenger car days in Mexico in 1950.

Thus, the total income taxpayer received in 1950 from rental of freight and passenger cars to Mexico was $377,761. Taxpayer had no other commercial association with Mexico, since taxpayer had no operations in that country, owned no facilities there, and had no control over cars after an interchange with Mexican railroads until the cars were returned.

The above-mentioned rental income from Mexican sources was subject to tax imposed by the Republic of Mexico pursuant to the provisions of the Mexican law entitled "Ley del Impuesto Sobre la Renta," the translation of which is "Income Tax Law." Pursuant to the provisions of the Mexican income tax law, taxpayer paid to the Republic of Mexico in 1950 a tax in the amount of $116,050 on the rental income received. In filing its federal income and excess profits tax return for the taxable year 1950, on a calendar-year basis and under the accrual method of accounting, taxpayer claimed and was allowed the full amount of the tax paid to Mexico as a foreign tax credit.

Subsequently, the Commissioner of Internal Revenue, hereinafter referred to as the Commissioner, determined that taxpayer was liable for additional income and excess profits taxes in the amount of $203,239. The deficiencies were duly assessed and were paid by taxpayer on January 17, 1956. On November 16, 1956, taxpayer filed a timely claim for refund of $184,048 of excess profits taxes paid in 1950. That claim was amended on June 10, 1960, to claim the relief provided by section 94 of the Technical Amendments Act of 1958 (Pub. L. 85-866, 72 Stat. 1606, 1669 (1958)). In considering taxpayer's claim for refund, the Commissioner disallowed as a foreign tax credit the $116,050 tax paid to Mexico and instead treated the tax as a deduction from gross income under section 23(c) of the Internal Revenue Code of 1939 and treated the barred deficiency resulting from that determination ($67,309) as an offset against taxpayer's claim for relief under section 94 of the Technical Amendments Act. Following the institution of this suit, a taxpayer received a $116,739 refund for the balance of its claim. Plaintiff now seeks to defeat the offset.

Sections 131(a) and 131(h) of the 1939 Code¹  provide that United States taxpayers may credit against their United States tax liability amounts paid or accrued on account of foreign income taxes or taxes imposed in lieu of income taxes. The first prong of attack by the Commissioner's setoff is his contention that the tax which the taxpayer paid to the Republic of Mexico was neither an income tax nor a tax imposed in lieu of an income tax within the meaning of section 131.

Article 1 of Ley del Impuesto Sobre la Renta, hereinafter referred to as the Mexican income tax law, sets forth the general nature of the tax:

The Income Tax is payable on profits, gains, rentals, products, benefits, participations and in general, on all amounts received in cash, in securities, in kind or in credit, which by reason of any of the items set forth in this Law, modify the taxpayer's possessions.

The rest of the Mexican statute goes on to establish a schedular tax system. It imposes a graduated tax on income and gains under five different schedules, each schedule having its own determination of taxable income and its own specific tax rates. Almost all economic activity or sources of income will fall within and be taxed under one of the five schedules. Schedules I and II are the only schedules relevant to the instant case.²  Schedule I is the general schedule covering business or commercial income. It subjects to tax those who "execute acts of commerce, or operate an agricultural or industrial business." In general, all acts performed by a business entity in connection with its business are classified by the Mexican Commercial Code as acts of commerce, and therefore fall under schedule I unless specifically included elsewhere.

The taxpayer's car rental income was specifically included under schedule II, however, and this was the schedule under which the tax was imposed. Compared with schedule I, schedule II is a somewhat narrow and specific schedule, covering, inter alia, the following sources of income:

ARTICLE 15. * * *

XIII. — Rentals, premiums, royalties and returns of all kinds which the owners or possessors of movable property receive from persons to whom they have granted exploitation rights without transferring ownership.

The concerns which lease or in any other manner grant the use or enjoyment of rolling stock to concerns which operate in the country, even though the contracts are made abroad * * * are specially considered as included under this Section.

The tax imposed on taxpayer's car rentals under schedule II was payable on the gross amount of the revenue, since schedule II makes no provision for deductions. This is the basis of the Commissioner's contention that the tax was not an income tax.

It is well settled that the question of whether the foreign tax is an income tax within the meaning of section 131(a) must be decided under criteria established by our own revenue laws and court decisions. Commissioner of Internal Revenue v. American Metal Co., 221 F.2d 134 (2d Cir. 1955), cert. denied, 350 U.S. 829, 76 S. Ct. 61, 100 L. Ed. 740; Biddle v. Commissioner of Internal Revenue, 302 U.S. 573, 58 S. Ct. 379, 82 L. Ed. 431 (1938). In other words, to be creditable under section 131(a), the foreign tax must be the substantial equivalent of an income tax as the term is understood in the United States. Lanman & Kemp-Barclay & Co. of Colombia, 26 T.C. 582 (1956); New York & Honduras Rosario Min. Co. v. Commissioner of Internal Revenue, 168 F.2d 745, 12 A.L.R.2d 355 (2d Cir. 1948); Keasbey & Mattison Co. v. Rothensies, 133 F.2d 894 (3d Cir. 1943), cert. denied, 320 U.S. 739, 64 S. Ct. 39, 88 L. Ed. 438.

The Commissioner cites Keasbey & Mattison Co. v. Rothensies, supra, and other cases, for its assertion that in order to be considered an income tax, a tax must be measured by net income. St. Paul Fire & Marine Ins. Co. v. Reynolds, 44 F. Supp. 863 (D. Minn. 1942), also lends some support for this proposition.

For its part, the taxpayer cites several cases rejecting the theory that a tax levied on gross income cannot be credited as an income tax. Santa Eulalia Mining Co., 2 T.C. 241 (1943), petition to review dismissed, 142 F.2d 450 (9 Cir. 1944), involving a Mexican income tax statute which was a predecessor to the one in issue here; Seatrain Lines, Inc., 46 B.T.A. 1076 (1942). To the taxpayer, the allowance of deductions from gross income in computing the amount of tax due is a matter of legislative grace and should not affect the classification of a tax as an income tax. The significant point, plaintiff contends, is that both the United States and the Mexican statutes "are predicated upon the concept of deriving national revenues by imposing a direct tax upon income, gains or recurrent benefits (usually measured in money) which proceed from labor, business or property." This is in contrast to taxes imposed upon sales, property, production, or the privilege of doing business or operating as a corporation. In other words, plaintiff taxpayer says the Mexican income tax law is an income tax in the sense that (1) it imposes tax on items which fall within the definition of gross income under the Internal Revenue Code of the United States and (2) it is not an excise or other nonincome type of tax, but a tax the sole purpose of which is to reach income.

The taxpayer points specifically to sections 231(a) (1) and (c) of the 1939 Internal Revenue Code, now section 881 of the 1954 Code, as amended, as an example of a provision in the United States Internal Revenue Code comparable to the Mexican tax in question. Section 231 provides for a tax on the gross income from United States sources of nonresident foreign corporations and is nearly identical to a Philippines tax which the Internal Revenue Service ruled was creditable in Rev.Rul. 66-65, 1966-1 CUM. BULL. 175. The Commissioner disputes the comparison since the Mexican income tax law apparently applied to car rental income earned by residents as well as nonresidents.³  The taxpayer, of course, replies that the nature of the tax is not affected by the residence of the taxpayer against whom it is applied.

Finally, there are some early Revenue rulings which deal with the characterization of predecessor Mexican tax statutes. Two of these rulings held that a tax on gross receipts in the form of interest or rents comes within the United States concept of an income tax, while at the same time either reserving judgment on other provisions of the applicable statute which imposes a tax on gross receipts from other sources, or actually holding such other provisions to be not within the United States concept of an income tax and, therefore, not creditable. I.T. 2620, XI-1 CUM.BULL. 44 (1932); I.T. 3385, 1940-1 CUM. BULL. 103. Thus, these rulings seem to place in a special category income from interest and rents, as if this type of increment from capital represents a classical flow of "income" which requires no deductions.

A slightly later Revenue ruling, dealing with sources of income other than interest and rent, went even further, and in deference to the Santa Eulalia Mining Co. case, supra, revoked several old Revenue rulings holding that particular taxes imposed under the general Mexican income tax law of the period were not creditable because they wre imposed upon gross receipts. I.T. 3787, 1946-1 CUM. BULL. 232. Although over 20 years old, this Revenue ruling has not been revoked.

However, despite the able presentation of the parties, it is unnecessary to resolve the issue of whether or not the Mexican income tax was an income tax under section 131(a), since the case is so clearly controlled by section 131(h) of the 1939 Code, as amended.⁴  That section provides that a tax which is imposed by a foreign country in lieu of an income tax otherwise generally imposed may be taken as a credit.

The purpose of section 131(h), added to the Code by the Revenue Act of 1942, 56 Stat. 858, was to expand judicial interpretation in this area so as to allow a foreign tax credit for a foreign tax which, although not a tax on net income, was imposed as a substitute for a tax on net income. Motland v. United States, 192 F. Supp. 358 (N.D. Iowa 1961). Thus, the scope of the section was extended to cover the case where "a foreign country in imposing income taxation" authorized the payment of "a tax in lieu of such income tax but measured, for example, by gross income, gross sales * * *," etc. S.REP. No. 1631, 77th Cong., 2d Sess. 131 (1942-2 CUM. BULL. 504, 602).

Treas.Reg. 111, § 29.131-2 (1939) spells out in detail what is required for a tax to be considered a tax in lieu of income taxes generally imposed by a foreign country: (1) That the country have in force a general income tax law; (2) that the taxpayer claiming the credit would, in the absence of a specific provision applicable to such taxpayer, be subject to such general income tax; and (3) that the general income tax is not imposed upon the taxpayer thus subject to such substituted tax.

All three of the foregoing requirements are met in the instant case: (1) Schedule I of the Mexican income tax law is certainly a general income tax law of that country. I.T. 3945, 1949-1 CUM. BULL. 88. As mentioned above, this schedule covers generally all business transactions of commercial concerns, that is, of those engaged in commerce, industry, or agriculture. The tax is imposed on net income only, since the deductions permitted under this schedule are very similar to the deductions allowed under the Internal Revenue Code of the United States. Article 6 states that the tax under schedule I "shall be payable on the difference resulting between the taxpayer's revenue and the deductions authorized by the Regulations. These deductions shall comprise solely the ordinary and necessary expenses required for the purpose of the business."

Articles 39 and 40 of the Mexican regulations allow taxpayers under schedule I the following deductions, inter alia: Cost of goods sold, amortization, depreciation, rental expense, salary expense, interest, insurance premiums, local taxes, and charitable gifts.

(2) Except for the specific provision relating to rolling stock contained in article 15, section XIII, under schedule II, quoted supra, taxpayer would have been subject to the general income tax under schedule I.⁵ 

In 1937, when the predecessor of section XIII, article 15, covered rentals from movable property, but did not specifically include railroad rolling stock, the Supreme Court of Mexico held that the rental income on rolling stock was not taxable under schedule II for the reason that rolling stock was classified under the new civil code of Mexico as being real property. In a later case in the same year the Court held that such income should be taxed under schedule I of the Mexican income tax law, applicable to commercial concerns doing business in Mexico. In January 1939 the Mexican income tax law was amended to provide specifically, as it did in 1950 in article 15, section XIII, that rental on railroad rolling stock would be taxed under schedule II. This background makes it clear that the tax imposed on taxpayer's car rentals is intended by the Mexican Government as a substitute for the general income tax and is not merely an additional or unrelated tax.

Defendant argues that taxpayer would not, in fact, have been subject to the general income tax under schedule I because taxpayer had no operations or facilities in Mexico in 1950. However, nothing in the statute indicates this requirement for the applicability of schedule I. One can "execute acts of commerce" without the physical presence of facilities. Article 2, which lists those persons subject to the Mexican income tax statute, does not distinguish between the various schedules:

Article 2. — The following persons are subject to the Income Tax: * *

II. — Foreigners domiciled in or outside of the Republic, when modification of their possessions is derived from sources of wealth located in the National Territory or from business transactions carried out therein.⁶ 

(3) It is conceded that the general income tax under schedule I was not imposed upon the taxpayer.

A review of section 131(h) cases before this court reveals the court's substantial agreement with the design of the requirements set forth by the above Treas.Reg. 111, § 29.131-2 (1939). In Prudential Ins. Co. of America v. United States, 319 F.2d 161, 162 Ct. Cl. 55 (1963), the taxpayer insurance company paid a 2-percent tax imposed by the Dominion of Canada and two provinces on the insurance premiums collected during the year from Canadian policyholders. The court succinctly stated its reason for holding that the premium tax was a tax in lieu of an income tax: "* * * plaintiff was excused from paying the Ontario tax calculated on the net income of corporations because it was an insurance company paying a premiums tax under another section of the same statute." 319 F.2d at 164, 162 Ct. Cl. at 61.

The court later reaffirmed its reasoning in three more Canadian insurance cases with similar facts. In a second Prudential case, the court disposed of the case under the rule of stare decisis. Prudential Ins. Co. of America v. United States, 337 F.2d 651, 167 Ct. Cl. 598 (1964). In Equitable Life Assur. Soc'y of the United States v. United States, 366 F.2d 967, 177 Ct. Cl. 55 (1966), the court held that "both the Dominion and the provinces decided to impose or retain the premiums taxes because these insurance companies were not being asked to pay income taxes * * *. For the purposes of the Internal Revenue Code's foreign tax credit, that is enough to characterize the premiums taxes as imposed `in lieu of' income taxes." 366 F.2d at 974, 177 Ct. Cl. at 68-69.

In the most recent case of this kind, Metropolitan Life Ins. Co. v. United States, 375 F.2d 835, 179 Ct. Cl. 606 (1967), the Government contended that Congress intended the "in lieu" part of the foreign tax credit to apply only to so-called "empirical" income taxes — "those clearly seeking to levy solely on net income but using a simplified or convenient formula or method for calculating such profit in order to avoid the administrative difficulties of ascertaining a non-resident's net income within the country." 375 F.2d at 837, 179 Ct. Cl. at 609.

The court rejected this interpretation of statutory history. Reviewing the history, the court said the excerpts in the Senate report referring to the administrative difficulties in calculating a non-resident's net income⁷  were merely examples intended to illustrate the need for expanding the foreign tax credit. The primary consideration was the avoidance of double taxation of the earnings of American companies in foreign countries.

Thus, the court concluded that a foreign tax may qualify for the credit if it was imposed not because of administrative or computational difficulties, but because, as was true for the Canadian premiums taxes, it was considered bad policy or inconsistent with the country's legal theory to levy the normal income tax upon a particular class of company. The tax is creditable so long as it "was levied by the foreign country in place of or instead of or as a substitute for some existing income or profits tax," regardless of the "reason the other country might consider it proper to substitute the `in lieu' levy in the special case for the ordinary income tax generally imposed."⁸  375 F.2d at 838-839, 179 Ct. Cl. at 610, 612.

Similarly, this court held in Compania Embotelladora Coca-Cola, S.A. v. United States, 139 F. Supp. 953, 134 Ct. Cl. 723 (1956), that a Cuban production tax was a tax in lieu of an income tax, because a specific provision in the general income tax law exempted payers of the production tax from paying the income tax.⁹ 

Thus, in all these cases, the determining question is whether the tax in question was imposed as a substitute for, instead of, or in place of, and otherwise generally imposed income tax to which the taxpayer would otherwise be subject. For the reasons explained above, the tax imposed on taxpayer's car rental income clearly meets this test.

Once it has been determined that the Mexican tax imposed on taxpayer's Mexican car rentals is a tax in lieu of an income tax, it becomes necessary to compute the credit to which taxpayer is entitled. Section 131(b) (1)¹⁰  imposes a limitation upon the amount of the credit which can be taken for creditable foreign taxes. This is designed to prevent the tax credit from reducing or eliminating the United States tax on income from sources within the United States. Under this section the maximum amount of the credit is a figure which bears the same relation to the total United States tax against which the credit is taken as the normal-tax net income derived from sources within the foreign country bears to the taxpayer's entire normal-tax net income. Thus, the maximum amount of the credit may be determined by use of the following formula:

 Normal tax net income from Mexico U. S. tax before ______________________ × Total normal tax net adjustment in issue income

Since all the other factors in the formula are known, the only problem before the court is to determine taxpayer's taxable income from Mexico. Since the gross income the taxpayer earned from Mexican sources is known, what really remains to be determined is the amount of expenses which the taxpayer incurred in earning the Mexican income. Of course, we are concerned only with expenses which constitute deductions from the gross income, since we are trying to determine taxpayer's taxable income in Mexico, using standard United States tax accounting concepts. Therefore, it might be more accurate to define the problem as one of determining what part of taxpayer's total deductions is properly allocable to the Mexican gross income.

For guidance section 131(e) of the 1939 Code¹¹  provides that the amount of (taxable) income from sources without the United States shall be determined in accordance with the provisions of section 119.

Section 119(d) in turn provides as follows:

(d) NET INCOME FROM SOURCES WITHOUT UNITED STATES. — From the items of gross * * * [foreign income] there shall be deducted the expenses, losses, and other deductions properly apportioned or allocated thereto, and a ratable part of any expenses, losses, or other deductions which can not definitely be allocated to some item or class of gross income. The remainder, if any, shall be treated in full as net income from sources without the United States.

Restating the statutory mandate in terms of the instant case, if an expense or other deduction, or some part thereof, can be definitely assigned to gross income earned in either the United States or Mexico, then it should be so assigned and no proration is necessary. De Nederlandsche Bank, 35 B.T.A. 53 (1936) (decided under a statutory provision involving the other side of the coin — i. e., the determination of a foreign corporation's taxable income within the United States.) Those deductions which cannot reasonably be directly assigned to one country or the other should be allocated ratably between the two. International Standard Elec. Corp. v. Commissioner of Internal Revenue, 144 F.2d 487 (2d Cir. 1944), cert. denied, 323 U.S. 803, 65 S. Ct. 560, 89 L. Ed. 640 (ratably allocating to foreign countries part of the general or overhead expenses of a domestic holding company which received dividend income from foreign subsidiaries in those countries).

Since determining net taxable income in Mexico is our first objective for purposes of this case, we are really not concerned with those deductible expenses which are assignable directly to the United States. Rather, we are concerned only with the following two classes of deductible expenses: (1) Those which are directly assignable to Mexico because they are directly attributable to and incurred in earning the rental income in Mexico; and (2) the Mexican portion of expenses which are not directly assignable to the income from either country and must, therefore, be allocated ratably between the two countries.

Expert witnesses for both parties prepared studies to determine the amount of such expenses. Both studies adopted the same general two-step approach: (1) Listing for the year all the taxpayer's expenses which were related to the ownership and maintenance of its car fleet; and (2) analyzing those expenses to determine which of them were incurred as a result of cars being used to earn income in Mexico.

From the foregoing, one can observe that the effect of not including an expense in the category of ownership and maintenance expenses will be to assign that expense directly against United States income in the first place. This is theoretically sound since the taxpayer has no facilities and no operations of any kind in Mexico, apart from the rental of its cars. Thus, it is felt that any expense completely unrelated to the ownership and maintenance of railroad cars is, by necessity, unrelated to earning car rental income in Mexico. In other words, since such expenses must relate solely to United States operations, they are assigned there initially by not even including them in the expert studies for our consideration.

Once an expense passes the first test and is included in the category of ownership and maintenance expense, its final disposition still remains to be determined — i. e., should it be (1) directly assigned (in whole or in part) against United States income, or (2) directly assigned against Mexican income, or (3) allocated between the two on some pro rata basis?

Once it has been determined that an expense should properly be allocated on a pro rata basis, as opposed to directly assigned, the parties are agreed that the proper method for so allocating expenses between the two countries is by use of a "car-day divisor." The following section on passenger coach expenses illustrates the operation of this approach in determining the portion of taxpayer's expenses incurred in earning income from the rental of passenger coaches to Mexican railroads.

Taxpayer's total 1950 passenger coach ownership and maintenance costs were $1,863,929 and consisted of the following elements of cost:

 Item of Expense Amount 1. Passenger train car repairs applicable to coaches ......... $1,212,991 2. Coach repair portion of maintenance of equipment overhead ........................................................... 112,322 3. Coach portion of equipment repair facilities expenses ..... 107,319 4. Coach portion of general expenses ......................... 44,684 5. Coach portion of interest ................................. 23,982 6. Property taxes applicable to coaches ...................... 76,760 7. Payroll taxes applicable to coaches ....................... 56,855 8. Depreciation applicable to coaches ........................ 229.016 __________ Total coach ownership expense--1950 ................... 1,863,929

None of the above expenses should be directly applied against either United States or Mexican income, but rather the total amount should be allocated between the two countries on the basis of car days.

The number of total potential passenger train coach days for 1950 is 54,385, derived by multiplying the simple average number of coaches owned by taxpayer during 1950 (149) by 365 days. From this total, the unserviceable car days of coaches (3,046) is subtracted to arrive at 51,339, the total potential serviceable car days of coaches.

The total passenger coach ownership and maintenance cost of $1,863,929 is then divided by the above potential coach days of 51,339 to determine the cost per day of coach ownership and maintenance. Thus, for 1950, the cost of ownership and maintenance per coach day was $36.31.

The cost per coach day figure is then multiplied by the number of coach days in Mexico (781), resulting in the total figure of $28,358.11, which represents the expenses which the taxpayer incurred in earning income from the rental of passenger coaches to Mexican railroads.

With two exceptions, items 5 and 6, the parties are agreed that the above items of passenger coach ownership and maintenance expenses are correct and that all of them should be allocated between the United States and Mexico on a car-day basis. The two disputed items relate to interest and property taxes. The following sections explain the reasons for the conclusions reached on these items.

Property Taxes. State property taxes on railroad cars, like property taxes in general, are assessed on the basis of value. In order to solve the special problem created by the rolling characteristic of railroad cars, property taxes are based upon a proration of the value of the cars to the several states in which a railroad operates on the basis of miles of road a railroad has in each state.

The defendant's expert assigned a portion of state property taxes and other miscellaneous taxes to the cost of passenger coach ownership,¹²  and allocated such amount between the two countries on a car-day basis. The taxpayer contends that since such taxes are incurred only in the states in which the taxpayer owns and operates railroad lines, no such taxes are properly allocable to earning Mexican income because no tax is incurred in Mexico by reason of the presence of taxpayer's cars in Mexico. The taxpayer's contention must be rejected.

It is true that taxpayer incurs no additional property taxes because of its car rentals in Mexico. Its property taxes would have been the same even if none of its cars had ever gone into Mexico. But this is also true of interest, depreciation, and other types of car ownership expense. For example, taxpayer paid interest only to American banks and incurred no interest charges specifically because of the presence of coaches in Mexico. But this does not mean, and taxpayer does not contend, that the total amount of interest charges attributable to coaches should be allocated to United States operation. It simply means that such owenrship expenses as property taxes are in the statutory category of expenses which cannot be specifically identified with the income from either country, and therefore, must be allocated ratably (by car days).

The reverse side of the taxpayer's coin is just as true — the amount of property taxes is not reduced by virtue of the presence of taxpayer's cars in Mexico and their absence from the United States. Thus, the taxes are not assessed on the basis of United States operation alone, but rather are assessed on the basis of taxpayer's overall United States and Mexico operation. The taxpayer would be correct if the method for assessing state property taxes made allowance for the fact that sometimes cars are not in any state and therefore divided less than 100 percent of the value of all cars among the states where lines exist. But, the fact is that property taxes are assessed against all passenger coaches, and not just against coaches operating in the United States. Thus, such taxes are overall ownership expenses which must be prorated against both Mexican and United States income. Defendant's evidence which employs the above approach, has been adopted for the purpose of determining the amount of property taxes applicable to passenger coaches.

Interest. Railroad cars are purchased through a down payment of 20 to 25 percent and the issuance of equipment trust certificates and conditional sales agreements for the balance. The taxpayer fixed the amount of interest charges on passenger coaches by simply computing the actual interest charges paid on outstanding equipment trust certificates and conditional sales agreements covering such coaches.

Defendant disputes taxpayer's approach on two grounds. First, it contends that the 20- to 25-percent down payment must have come from the proceeds of general mortgage or other non-equipment bond issuances and that, therefore, an approach should be adopted which takes into account more than just the interest on the trust certificates and sales agreements. The taxpayer replies, however, that the down payment does not come from general mortgage funds, because such funds are invested entirely in roadway property and do not cover equipment transactions. Taxpayer claims that the source of the down payment was current or accumulated earnings. The record supports the taxayer on this point. The last bond issuance of taxpayer was a very minor one in 1933. The great majority of Missouri Pacific bonds were issued prior to 1930. Certainly the taxpayer has acquired new railroad cars, including coaches, since then.¹³  Thus, the cars acquired since the 1930's were acquired without the benefit of proceeds from bond issuance. For at least 20 years prior to the taxable year in question, taxpayer financed the down payment on equipment purchases from sources other than bond issuance.

The defendant's second objection to taxpayer's method of computing the interest costs relating to the ownership of coaches is more sweeping than its first. It contends that taxpayer is attempting to isolate a portion of its operation and fails to recognize that the debts incurred in acquiring railroad cars must bear some relationship to the overall debt structure. One must look at the entire debt structure, says the defendant, since the taxpayer is not in the car renting business, but is in the railroad business. The entire debt is thus claimed to be a general cost of doing business, and a portion of that debt is reasonably chargeable to passenger coaches. Therefore, the defendant determined the interest costs chargeable to the ownership of passenger coaches by computing the ratio of taxpayer's investment in passenger coaches to its total investment in all tangible assets, and multiplied that ratio by the total interest costs for the year.

The problem with the defendant's method is that it ignores the statutory mandate of section 119(d) of the 1939 Internal Revenue Code. The use of the convenient short-cut process of dealing with expenses in terms of the ownership and maintenance of railroad cars, as discussed supra, should not distract us from the method of allocation established by the statute. What the defendant is really saying is that none of the interest expenses are directly identifiable with or attributable to the income earned in either the United States or Mexico.

On the other hand, the taxpayer is saying that all interest costs, other than those paid on the equipment trust certificates which financed the passenger coaches, are directly assignable to the United States. The taxpayer is correct in this.

The general mortgage funds went to finance road equipment in the United States, which properly contributes to the earning of United States income only. The only interest expense which contributes to activity which earns Mexican income is interest which finances railroad cars, since some cars are rented to Mexico. It is impossible to comprehend how any of the interest cost on money borrowed to finance a line of track in Texas, for example, has any relationship to the earning of Mexican income.

The defendant's method has two distorting effects. By lumping all interest costs together, it applies an average rate on funded debt to railroad coach acquisitions. But, in fact, as defendant's expert witness admits, the interest rate on railroad cars is much lower than the interest rate on road property. Why should the railroad cars, and thus ultimately Mexican income, be charged with the higher interest rate required to finance United States operations? The direct, exact, and clearly identifiable cost of railroad car financing is known, and it should be used.

A second distortion in the defendant's method results from the fact that about one-third of the equipment obligations are paid off. Thus, the defendant is allocating a portion of general mortgage interest to cars which have paid their full interest costs in a tax sense.

In the world of finance and economics, the defendant's approach has merit. All interest is theoretically interchangeable and all assets bear a certain imputed interest cost, since the assets could be sold and the proceeds invested at interest. But, in computing taxable income (in the United States and in Mexico) established tax accounting techniques must be used. The tax law has never recognized imputed expenses in determining taxable income.¹⁴  The defendant's theory would lead to the logical conclusion that railroad cars bear a certain interest expense even if all debt had been paid off, including debt on roadway property. Likewise, the defendant's theory would lead to the same conclusion if, to take an unlikely hypothesis, a car manufacturer sold the taxpayer railroad cars under noninterest-bearing conditional sales agreements. But, the effect would be similar to the taxpayer's present circumstances. Approximately one-third of the cars were incurring no interest expense in 1950. The Code allows as a deduction only "interest paid or accrued within the taxable year on indebtedness,"¹⁵  and for some of these cars in 1950 no amount was paid or accrued on account on indebtedness.

A final analogy can be made in the area of depreciation, where the tax concepts also often depart from economic concepts. If a substantial portion of these cars was fully depreciated for tax purposes, though still running and earning income in the United States and Mexico, depreciation on the said cars would have to be ignored in computing taxable income, and one could not impute to those cars a portion of the depreciation on other property.

In summation, the taxpayer's method of determining the interest expense attributable to passenger coaches is adopted, i. e., inclusion of interest on equipment obligations alone, since the interest on car financing ceases to exist when these obligations are paid, and the interest on road property in the United States has no relation to Mexican income. The interest attributable to coaches, like the rest of the coach ownership and maintenance items, was allocated between the two countries on the basis of car days, as explained above.

There is much less agreement in the parties' expert studies of freight car ownership and maintenance expenses. The parties disagree both as to the inclusion of certain amounts in the category of ownership and maintenance, and as to the ultimate allocation or assignment of various items within the category. The main difference in approaches stems from the taxpayer's much more detailed study.

For example, instead of determining the total amount of freight car maintenance expenses and allocating that total between the two countries on the basis of freight car days (as was done for passenger coach maintenance expenses), the taxpayer broke down direct maintenance expenses, i. e., freight car repairs, into three separate categories of repairs. One category it allocated by car days, another was allocated on the basis of where the repair was incurred, and the third was assigned completely to the United States. The taxpayer followed this approach through in its treatment of indirect freight car maintenance expenses (various overhead or general expenses) allocable to freight car repairs. This opinion will use the taxpayer's approach as a point of departure in its analysis. On issues where the taxpayer is correct, the greater detail of its study adds a greater degree of precision. Where taxpayer is incorrect, the greater detail is still valuable in that it more clearly defines the issues.

Both parties use as a starting point the total amount of taxpayer's account 314, "Freight Train Car Repairs," in its 1950 annual report to the I. C. C. Thus, the issues are actually presented in terms of what are the proper adjustments to this figure, in order to determine ownership and maintenance expenses. The disputes over particular items are resolved in general terms in the findings which follow. Finding 54 contains a summary which gives in dollar amounts the results of the findings and conclusions made here.

Repairs. The records kept by taxpayer in 1950 indicate that it divided its freight car repairs into three classifications: (1) Classified repairs, (2) running repairs, and (3) user repairs.

Classified Repairs. Classified repairs are normally made in one of the three major repair shops of taxpayer and involve all body work over $50. A great percentage of classified repairs is programmed or scheduled and, on the average, a car receives a classified repair every 5 or 6 years. Under the interchange rules, classified repairs are made only by the owning railroads, and not by user railroads operating foreign cars on their lines. Since the operation of freight cars in Mexico, as well as in the United States, contributed to the classified repairs expenses, these expenses should be allocated between the two countries on the basis of car days. On this item there is no dispute between the parties. Classified repairs cannot be definitely assigned to the income of either country, since they pertain to the entire operation of taxpayer.

Running Repairs. Running repairs are repairs which are made either in trainyards or on tracks, commonly called rip tracks, where relatively minor adjustments and repairs are made. Running repairs consist primarily of work not involving the body of the car at all, such as repairs concerning the wheels, axles, springs, and braking system, although running repairs occasionally include work on the body which is under $50. Such repairs are often unscheduled. A railroad makes running repairs to foreign cars on its lines as well as to its own cars, but these repairs are considered the owner's responsibility and the using railroad bills the owning railroad for such repairs at rates established under the interchange rules.¹⁶ 

The defendant did not distinguish between running repairs and classified repairs, but included running repairs in total repair expenses to be ratably allocated.

The taxpayer, on the other hand, assigned running repairs on an actually incurred basis. If the repair took place in Mexico, the expense was assigned to Mexico, and if the repair was performed in the United States, the expense was assigned here. The defendant, in turn, contends that there is no justification for segregating car ownership costs on the basis of the fortuitous circumstance of the location at which such repairs were made, and that allocating the total cost on a car-day basis is more accurate.

The two approaches lead to considerably different results because, as will be discussed shortly, the taxpayer actually spends less on running repairs per car day in Mexico than it does in the United States. Thus, the car-day allocation method results in a greater allocation of expenses to Mexico, since it lumps all running repairs together and assigns the average expense per car day to Mexico.

In effect, the taxpayer is claiming that the expenses of running repairs in Mexico are expenses which, under the direction of section 119(d), are properly assignable directly to Mexico. Whether this is so depends upon whether those expenses accurately reflect the true cost of running repairs in Mexico. After all, what the taxpayer is referring to as the expenses of running repairs in Mexico is merely the sum of the bills paid to the Mexican railroads for repairs performed there. If the taxpayer had no cars in use in Mexico at all, but merely delivered a portion of its cars there for repairs, it is obvious that the bills for repairs performed there would not reflect the true cost of its Mexican car-use, which is zero. Looking at the total running repair expenses, the true cost of taxpayer's Mexican use is controlled by two factors: (1) The contribution the Mexican use made to the necessity for repairs, i. e., the wear and tear taxpayer's cars incurred in Mexico; and (2) differences in the price of repairs in each country. This is elaborated in the following analysis:

Looking just at the bills paid, that is, at the amount of money spent, the record reveals that the taxpayer spent approximately 64 cents per car day for running repairs on its own cars on its own lines. The Mexican railroads spent only 8.67 cents per car day on running repairs to taxpayer's cars, or at least that is the amount for which the taxpayer was billed.

If this disparity results from the fact that it truly costs less to run taxpayer's car for a day in Mexico than in the United States, then a proper matching of (Mexican) income and expenses requires that the repairs billed in Mexico be directly assigned to Mexican income. This is in accordance with the concept of an expense properly assignable to foreign gross income under the statute. Known differences in cost should not be submerged into averages. This is why interest on equipment obligations, which bore a lower interest rate, was kept separate from other bonded indebtedness covering United States property.

The taxpayer, of course, does claim an actual lower cost per day of running repairs in Mexico. Lower Mexican cost could only result from two factors: (1) Mexican railroads charge less for a given repair than the same repair would cost the taxpayer at home, or (2) less repairs are necessary in Mexico. The taxpayer does not claim the Mexican rates are lower than its own, since rates charged for running repairs are set by the interchange rules and represent the average cost of such repairs in the United States.

The taxpayer does claim, however, that the use to which the cars are put in Mexico and the particular type of service they render there, necessitate less repairs. In other words, service on its own lines is supposed to be "harder service" than service on Mexican lines.

However, except for the bare statement to this effect by taxpayer's expert witness, there is no evidence whatever of this fact in the record. The taxpayer does not give us one example of how service in the two countries differs, or why Mexican service should be any easier on cars. The taxpayer's own expert witness admitted that there is no particular reason why a mile in Mexico should be treated differently than a mile in Colorado. And there is nothing in the record to indicate that more miles are run in a car day in the United States than in Mexico.

The defendant's explanation for the difference in the amount spent on running repairs in each country is more logical and better supported by the credible evidence. The difference results from a conscious policy on the part of the railroad industry to spend more to repair its cars on-line so that the fleet would be in such a condition that there would be less necessity for repairs off-line. In other words, where possible the taxpayer sought to repair its cars at home rather than have them repaired in Mexico. Also, a railroad tends to be more diligent in searching for and repairing defects of a nondangerous nature in its own cars than in the cars of another railroad on its line, since it is more concerned with the condition of its own fleet. Thus, there are no lower costs connected with the operation of taxpayer's cars in Mexico. The taxpayer is just seeing to it that more of the repairs (including a portion of the repairs necessitated by wear and tear in Mexico) are performed in the United States than in Mexico.

The fact that the disparity in the amount spent for running repairs does not exist only between taxpayer's on-line repairs and Mexican repairs, tends to bear out this contention. At the same time that taxpayer was spending 64 cents per car day for running repairs on its own cars on-line, it was spending only about 27 cents per car day for running repairs to foreign cars on its lines. Also, in contrast to the 64 cents taxpayer spent for running repairs on its own cars, is the approximately 27.1 cents per car day that other United States railroads spent on running repairs to taxpayer's cars while on their lines.

As plaintiff admitted on page 27 of its reply brief to the trial commissioner:

Plaintiff's study showed that plaintiff spent a greater amount for running repairs to its cars on its own line than for such repairs to its cars on foreign lines and for repairs to foreign cars on line. This procedure is generally followed throughout the railroad industry as owner lines are primarily responsible for the operating condition of their cars. [Citations omitted.]

The interchange rules also tend to support the idea that repairs should be made at home whenever possible. Rule 1(B) of the 1948 edition of the Code of Interchange Rules of the Association of American Railroads reads as follows:

Repairs should be made by the car owner, insofar as may be practicable. In event a foreign car requires repairs on account of owner's defects, such repairs may be made subject to the following conditions:

1. Repairs to loaded cars must be confined to the minimum necessary for the safety of car, lading and trainmen.

The above analysis completely destroys the assumption that must be made and the condition that must exist if direct assignment of the running repair expenses is to reflect accurately the true cost involved. That assumption and condition is that the frequency of running repairs is such that they are made where the wear and tear or other defect was actually incurred.

As to any particular car, the above would be true only by chance. But, given a large enough fleet of cars and assuming, as seems reasonable to a degree, that the frequency of running repairs varies with the use given the car, the above would be true as long as (1) taxpayer's cars are inspected as frequently in Mexico for running repairs as they are on-line in the United States, and (2) the standards concerning when a repair is to be made are applied equally.

As pointed out above, these conditions are not met. Given a car in the same condition, a repair might be made in the United States that would not be made in Mexico.

But, the taxpayer contends that running repairs are performed on an emergency basis, out of operating necessity and safety, and thus may not be deferred. Yet the taxpayer contradicts itself on page 30 of its reply brief to the commissioner, where it admits "railroads permit minor repairs which do not affect the safety of the car, the lading, or trainmen to accumulate."

The defendant is not contending that defects are allowed to accumulate to the danger point in Mexico. Defendant is merely pointing out that the taxpayer might make a running repair to its own car long before safety considerations absolutely necessitated it, in order to preclude its being done by another railroad. A Mexican railroad, seeing the taxpayer's car in the same condition, would be inclined to let it go, since it was not yet a safety hazard. A railroad prefers to spend its time and facilities keeping its own cars in good condition.

The taxpayer's point is perhaps well taken that the car-day method of allocation assumes that all cars travel the same number of miles within a given period, regardless of where used, and that the various uses to which various cars are put do not result in any differences in wear and tear experienced. But, the taxpayer has not demonstrated any departures from this assumption that would explain the fact that taxpayer spent over seven times as much per car day for running repairs in the United States as in Mexico. Taxpayer's criticism of that assumption would also apply to some extent to classified repairs, but taxpayer agreed that they should be allocated by car days.

The fact remains that the assumption behind defendant's method is much more reasonable than the assumption one must make to use the taxpayer's method, and until a precise time/use and mileage breakdown of the cause of repairs is made, a choice is forced between the two assumptions and the methods to which they lead.

In summation, the evidence does not establish that the nature of a daily run in Mexico was easier on the cars and therefore resulted in less need for running repairs. Rather, from all that appears in the record, a Mexican car day contributed equally with a United States car day to the need for running repairs, but the taxpayer endeavored to see that the repair occurred on-line in the United States rather than in Mexico. Therefore, to reflect accurately the true cost incurred in each country, running repairs should be allocated between the two countries on the basis of car days rather than be assigned directly to the country where the repairs took place.

User Repairs. User repairs are repairs of damage occurring because of unfair usage or improper protection by the using railroad, that is, repairs which are not due to normal wear and tear. Under the interchange rules, user repairs to foreign cars cannot be billed to the owner but must be borne by the using road.

The parties are agreed that any user repairs to the taxpayer's cars should be assigned directly against United States income. If any such repair was done to one of taxpayer's cars in Mexico, it was, of course, not an expense of the taxpayer, since the Mexican railroad bore the expense. All such repairs done in the United States resulted from a particular use or misuse to which the car was put while earning United States income for the taxpayer.

The taxpayer also assigned directly to United States income, user repairs which the taxpayer made to foreign cars, while the defendant allocated such amount ratably by car days to both countries. The taxpayer's method is correct, since the reason for assigning such repairs to the United States applies identically to foreign cars. Just as the taxpayer, as an owner of cars, incurred this expense while earning exclusively United States income, so the taxpayer, as a renter or lessee of cars, incurred this expense while earning exclusively United States income. The fact that the cars repaired belonged to another railroad does not detract from their assignability to United States income. In fact, it only emphasizes that the expenses incurred were not ownership or maintenance expenses at all, but operating expenses relating solely to United States operations.

Inspection. The parties disagree as to whether the cost of freight car inspection at terminal and junction points should be included in the class of freight car expenses to be allocated between the two countries. There are two basic types of car inspection expenses. One is an inspection expense in connection with running and classified repairs to determine what the repair was and whether it was properly made. This expense is treated by both parties as part of the repair expenses.

The inspection expense with which we are concerned relates to the inspection of freight cars at all terminal and junction points, including charges on account of coupling air hose and testing air. Inspection at terminal and junction points serves at least two purposes. In connection with the interchange system, such inspections at the time of interchange determine who is responsible for the condition of the car. Secondly, and most important, before a car leaves a yard or terminal to begin a run, it is inspected to see whether it is in safe and proper condition for continued movement.

Inspection at terminal and junction points is performed on foreign and privately owned cars as well as on system cars. The Code of Interchange Rules provides that such inspection costs are not billable to the owning railroad, but must be absorbed by the using road.

The taxpayer did not include in its cost of car ownership and maintenance any of the inspection expenses at terminal and junction points. This, of course, had the effect of assigning the whole amount to United States income. The taxpayer reasoned as follows: Such inspection expenses are directly attributable to the use and operation of the car; they add nothing to the investment in or value of a car, are not expenses of car ownership or maintenance, but are operating expenses borne by the user; the taxpayer is not a user of cars in Mexico, no user repair costs are incurred by taxpayer in Mexico, and therefore, none are assignable to earning income in that country.

The defendant contends that some of such inspection expenses are ownership expenses since the only way to find and program repairs is to inspect the cars. Therefore, says defendant, such inspection costs are just as much ownership expenses as the repair expenses are, and must be ratably allocated. However, the defendant did eliminate 40 percent of these inspection expenses as "assignable to private car inspection and transportation cost element of railroad owned freight car inspection."

Even if defendant were right that some portion of terminal and junction inspection should be included as an ownership or maintenance expense, its computations must be rejected. The only expenses which can possibly be considered as expenses of car ownership or maintenance are expenses which relate to the taxpayer's own cars. The defendant recognized this by eliminating from terminal inspection a portion of such expenses attributable to the inspection of privately owned railroad cars. But, the defendant neglected to eliminate the expenses attributable to the inspection of foreign cars (cars owned by other railroads).

Foreign railroad cars and privately owned cars spent a total of 9,391,088 active car days on taxpayer's lines. Taxpayer's own cars spent only 4,567,204 active car days on-line. Thus taxpayer's own cars accounted for less than one-third of the total of 13,958,292 active car days on taxpayer's lines. From this it is reasonable to assume that the greater part of the terminal inspection expenses were incurred by taxpayer on foreign cars. In this regard it should be noted that the Association of American Railroads has a different standard for the inspection of foreign cars than it does for the repair of foreign cars. Rule 1 (A) of the 1948 edition of the Code of Interchange Rules reads as follows:

Each railroad is responsible for the condition of all cars on its line and must give to all equal care as to inspection and lubrication.¹⁷ 

Nevertheless, the defendant included 60 percent of the terminal inspection as a cost of maintenance. Furthermore, included in the 40 percent excluded, was an amount assignable to "the transportation cost element" which defendant described elsewhere as "the normal operations of the taxpayer." This is apparently an admission by the defendant that some of taxpayer's terminal inspection of its own cars is a nonmaintenance car inspection factor. In any event, looking just at the amount of terminal inspection repairs that must have been made to foreign cars, it is obvious the defendant's 40-percent deduction is too small. But, the defendant's treatment of terminal inspection expenses is objectionable for the more important reason that taxpayer is right — none of such expenses were incurred in Mexico or are assignable to earning income in that country.

Although the parties at times argued as to whether terminal and junction inspections are expenses of car ownership and maintenance on the one hand or operating expenses on the other, a more direct analysis is possible. The terms "maintenance expense" and "operating expense" are not necessarily mutually exclusive. These inspections, which do not increase the life or value of the cars, might be called either one, without affecting the important question — should such costs be assigned directly against United States income? On the whole, we are talking about frequent spot checks as a train moves from point to point, to assure that the train is in condition to travel to the next point. It is use alone which necessitates these inspections. Every inspection expense taxpayer incurs in the United States is directly attributable to the fact that the taxpayer used that car in the United States. Any terminal and junction inspections which the cars received in Mexico were likewise directly attributable to the use of the cars in Mexico. If taxpayer had paid for such inspections, they would be properly assignable directly to Mexico. Direct assignment would be proper under the statute, because the inspections in each country are so clearly attributable to the income-earning activity or use in each country.

The fact that the Mexican railroads absorbed the expense and did not bill the taxpayer for inspection to MOPAC cars there does not change the analysis. A direct assignment of expenses is still proper, but the inspection expenses which taxpayer incurred in Mexico, to be assigned against Mexican income, are zero. The taxpayer incurred no such expenses in Mexico because the Mexican railroads paid the bills.

In other words, the direct assignment of the amount taxpayer spent on inspections in Mexico — zero — accurately reflects the true cost to taxpayer of inspections in Mexico. A comparison of inspection expenses with some of the previously analyzed expenses will demonstrate why this is so.

Running repair expenses are similar to inspection expenses in that both are minor maintenance or operating expenses attributable mainly to use. Direct assignment of running repair expenses would have been proper, also, except that special circumstances prevented the amount spent in each country from representing the true cost of operations. A great percentage of the repairs necessitated by use in Mexico was not done there. Since cars must be kept in repair, what was not done in Mexico was done in the United States, thus increasing the amount spent in the United States even though the wear and tear was incurred in Mexico.

This special situation does not exist as to the inspection expense. First of all, as mentioned above, it is likely that taxpayer's cars received pretty much the same inspection in Mexico as they did in the United States. But, even if they didn't, a lack of inspection in Mexico did not add to the amount taxpayer spent on inspections in the United States. If some inspections were not done in Mexico, then inspection expenses were simply not incurred by anyone. But this did not necessitate proportionate extra inspections when the car was returned. The need for inspections does not accumulate as does the need for repairs. Inspections take place only as a car is used. Thus, the inspection expenses taxpayer incurred in the United States, i. e., the entire inspection expenses, were attributable solely to United States operations and should be assigned directly against United States income.

The above analysis also applies to user repairs, supra. Any user repair expense necessitated by use of the cars in Mexico was borne by the Mexican railroads. There is no evidence in the record that the Mexican railroads did not make the necessary user repairs.

As to classified repairs, none were conducted on taxpayer's cars in Mexico, so it is obvious the Mexican use resulted in the repairs being done and the money being spent in the United States. Thus, direct assignment of the expenses to the United States would have been improper, so they were ratably allocated.

Switching Expense — Bad Order Cars. The nature of this expense is discussed infra. For reasons explained there, the taxpayers computations are adopted.

Other Freight Car Repair Expenses Since both parties agreed to the elimination of certain other expenses from the category of the freight car maintenance, their computations are accepted without discussion.

Adoption of Taxpayer's Computations. The result of the above conclusions is that all of the taxpayer's computations as to freight car repairs are found to be correct, except for taxpayer's treatment of running repairs. Those computations are adopted, except that the taxpayer's figures for running repairs and classified repairs are added together and treated as one, for eventual pro rata allocation between the two countries.

In computing the various indirect freight car maintenance expenses to be discussed in this section, the defendant continued to use total figures, while the taxpayer continued to break down its figures into the various types of repairs, and terminal and junction inspection. Both parties used taxpayer's 1950 I. C. C. annual report and other books of the taxpayer as a starting point. Because of the added detail, taxpayer's approach is preferred. Just as the direct user defect repair expenses were properly removed from the total freight car repairs to be allocated between the two countries (i. e., were directly assigned against United States income), so any portion of indirect maintenance expenses attributable to user defect repairs is properly removed from the total expenses to be allocated.

Taxpayer's breakdown contains an estimate of the portion of various indirect car maintenance expenses attributable to taxpayer's running repairs to foreign cars. This is proper since such costs are in no way part of taxpayer's car ownership and maintenance costs.

Similarly, since the I. C. C. annual report figures for freight car repairs included an amount for junction and terminal inspections, it is reasonable to assume that certain indirect freight car maintenance figures did also. To whatever degree possible, such amounts should be estimated and eliminated, as they were from direct freight car repairs.

In general, the taxpayer made its breakdown of indirect maintenance expenses by applying to the total figures the same ratios that existed for direct freight car repairs.

In truth, the record does not contain the detail as to the nature of the various accounts which is necessary to determine with any precision how reasonable taxpayer's breakdown is. However, as a logical matter, taxpayer's method is not prima facie unreasonable. Furthermore, the defendant does not dispute taxpayer's breakdown, or offer a breakdown of its own. For these reasons, taxpayer's method of breaking down the indirect freight car maintenance expenses has been incorporated in this opinion. Of course, taxpayer's separate categories of running repairs and classified repairs will be added together for pro rata allocation, as they were for direct freight car repairs. This means that for convenience, taxpayer's computations will be used as a starting point, with corrections made where required.

Miscellaneous Maintenance of Equipment Overheads. These are various overheads in conjunction with repair costs. The defendant is in basic agreement with taxpayer's approach and computations, so they are adopted here without further discussion.

Maintenance and Depreciation of Equipment Repair Facilities. Equipment repair facilities refer to the facilities where car repairs are made. The parties are in basic agreement as to the total maintenance and depreciation expenses on all equipment repair facilities, but differ in the ratio they applied to derive the portion attributable to freight car repair. The taxpayer's ratio is preferred since it is based on estimates from 1950 books and accounts, while the defendant's is based on estimates contained in a 1946 questionnaire filed with the I. C. C. by taxpayer. Therefore, taxpayer's ratio is adopted in this opinion.

State Property Taxes Allocated to Equipment Repair Facilities. The parties agreed in principle on this calculation, but again taxpayer's calculations are preferred because it used 1950 figures to estimate the freight car portion, while defendant relied upon a 1946 questionnaire. Therefore, taxpayer's calculation is adopted in this opinion.

General and Administrative Overhead Expenses. These expenses consist of such items as the salaries and expenses of general officers, clerks and attendants; general office supplies and expenses, law expenses, stationery and printing; and pensions. Both parties attempted to estimate the portion of such expenses which were related to freight car repairs by applying to such expenses the following ratio:

Total freight car operating expenses less depreciation, amortization & general & administrative expenses ------------------------------------------------------- Total railway operating expenses less depreciation, amortization & general & administrative expenses

However, before applying the ratio, the taxpayer first removed from the total of general and administrative expenses the salaries and expenses of clerks and attendants (the largest single item on the list) except for the salaries of clerks in the Car Accounting Department. A function of the Car Accounting Department is to keep records of the movement of cars for collection of per diem payments. There is no justification in the record for the taxpayer's exclusion of the rest of the clerical salaries from the formula allocating general expenses. Though the salaries of the Car Accounting Department are clearly related to freight car repairs, there appears in the record no assurance that some portion of the salaries of clerks in other departments is not also related to freight car repairs. In fact, the record does not even contain a complete list of what the other departments are. It would seem that other clerical departments would be involved in activity relating to ownership and maintenance of freight cars, such as, an accounting department handling the bookkeeping with respect to repair expenses. For the sake of consistency, the rest of taxpayer's approach and calculations will be adopted, except that clerical salaries will be added.

Payroll Taxes. In computing the portion of payroll taxes apportionable to freight car maintenance, the taxpayer first divided the actual 1950 payroll taxes (railroad retirement and unemployment taxes) by the total compensation paid for 1950. The taxpayer then applied the resulting payroll tax per dollar ($0.89) of compensation to the total estimated labor portion of all the freight car maintenance expenses it had computed.

The defendant also estimated the labor portion of the freight car maintenance expenses it had computed. But it then applied to this figure a ratio of 0.575, which it says represents the actual payroll tax rate of 0.650 adjusted to reflect the taxable portion of total payroll, i. e., compensation not in excess of $300 per month.

Neither party offers any argument as to why its method is to be preferred, or why the other party's method might be wrong. Nothing in the record indicates that either method is prima facie unreasonable. Accordingly, the court is given no reason to prefer one method over another.

Keeping in mind that for this particular item the method of each party uses all the previous calculations of the party, it will be more convenient to adopt the taxpayer's method, since most of its previous calculations have been adopted. The proper adjustment will be made for the one instance — general and administrative overheads — where the calculations adopted in this opinion differ from those of the taxpayer.

Deadhead Haul Cost of Car Repair Material. This refers to the nonrevenue freight expense chargeable to the transportation of car repair material. The parties are in basic agreement as to the approach, but as was true of the calculation for payroll taxes above, this calculation incorporates parts of previous calculations by the parties. Therefore, the taxpayer's calculations will be adopted here also, again with the proper adjustments for the changes made in the general and administrative overheads.

Three categories of expenses are involved here: (1) Depreciation and amortization, (2) interest, and (3) state property taxes. These are expenses which the taxpayer incurred strictly by virtue of owning a fleet of freight cars and regardless of the type and amount of use to which they were put.

Depreciation and Amortization. The taxpayer used as freight car depreciation and amortization the same amount as it showed for that item on its 1950 tax return. The defendant came up with an amount that differed by about 1 percent, by going back to taxpayer's 1950 records and making the tax calculations all over again. The record does not indicate the reasons for the discrepancy.

In view of the lack of a strong reason for doing otherwise, the taxpayer's 1950 tax return figures should be used. After all, what we are trying to do is allocate taxpayer's 1950 tax deductions (i. e., deductible expenses) against its 1950 gross Mexican income. The defendant audited taxpayer's 1950 return twice, and accepted its calculations for depreciation and amortization. Lacking some stronger showing of cause, it seems a little late in the game to start all over and object to those figures now.

As to the allocation of the expenses, the parties are agreed that an allocation between the two countries on the basis of car days is correct.

Interest. Taypayer included in its interest expense attributable to freight cars only the interest paid on outstanding equipment trust certificates and conditional sales agreements covering freight cars.

The defendant, on the other hand, constructed a cost of capital rate for the taxpayer (an interest rate taking into account taxpayer's total capital stock and long-term debt) and applied this rate to the ledger value (less accrued depreciation) of taxpayer's freight cars. This had the effect of eliminating any difference in interest rate between equipment obligations and other long-term debt and of assigning a portion of the interest cost to freight cars which had already paid off the obligations covering them.

The arguments and considerations here are identical to those discussed in relation to interest on passenger coaches, supra, so the same conclusions are adopted here.

However, the taxpayer did omit one necessary element of interest cost which the defendant included. The defendant separately estimated the interest cost attributable to freight car repair facilities. The taxpayer does not dispute the correctness of such inclusion. The defendant's method seems reasonable, so it is adopted. Although they differed over the amount of interest expenses attributable to freight cars, the parties are agreed that such expenses are properly attributable between the countries on the basis of car days.

Property Taxes. Again, taxpayer assigned all of the property taxes on freight cars to United States income, while the defendant allocated them between the two countries on the basis of car days. The dispute is identical to the one concerning property taxes on passenger coaches, supra, so for reasons already discussed, the taxpayer's computations will be modified by allocating the freight car property tax on a car-day basis.

Following the method used for passenger coach expenses, the total freight car ownership and maintenance expenses to be allocated between the two countries must first be divided by the proper freight car-day divisor to obtain the costs per car day, and the result multiplied by the number of freight car days in Mexico to derive the portion of such expenses allocable to Mexican income.

The parties disagree as to what the proper car-day divisor is. They do agree upon the first step — multiplying the simple average number of freight train cars owned by taxpayer during 1950 by 365. This gives the total available system car days per year. The parties also agree that the number of bad order car days for 1950 should be subtracted from this amount. Bad order car days reflect days during which cars cannot be used because they are awaiting or undergoing repairs.

Among the records taxpayer filed with the I.C.C. every year are so-called CS-60 reports, which purport to list the number of bad order car days for the year. Some doubt was cast upon the accuracy of such reports by the results of a 1947 questionnaire. This questionnaire, developed by the Association of American Railroads and sent out to railroads in 1947, asked for statistics covering all facets of railroad operation, based upon the experience of a test week in 1946. The taxpayer's 1947 questionnaire indicated that taxpayer's CS-60 reports understated the amount of bad order car days by 18.62 percent. This understatement seems to be due to the fact that the CS-60 reports indicated only cars awaiting repairs and not cars also undergoing repairs. The taxpayer calculated its 1950 bad order car days by taking the bad order car days reported on its 1950 CS-60 reports and increasing that amount by 18.62 percent.

The defendant, complaining that the CS-60 reports are "notoriously inaccurate," used a different technique to calculate 1950 bad order car days. Using the 1947 questionnaire, the defendant divided taxpayer's bad order cars for 1946 by the average number of cars for the year. The defendant then applied this 1946 bad order car ratio to the average number of cars owned in 1950, and multiplied by 365 days, to come up with the number of 1950 bad order car days.

The taxpayer's method is the more reasonable. If the defendant has such confidence in the accuracy of the 1947 questionnaire, then it should accept plaintiff's finding that CS-60 reports are understated by 18.62 percent and be content to adjust the 1950 CS-60 reports accordingly.

Too many variables go into the ratio of bad order cars to total cars to assume that the ratio remains the same from year to year. A railroad may change its policy as to standard of upkeep or frequency of repair. Differences in intensity of traffic may lead to increased necessity for repair, or minimization of repair because of a backlog of business. The taxpayer contends that bad order car days in 1946 were much greater than usual because of the build-up of deferred maintenance during the World War II years when labor and material were scarce. It is not necessary to make a finding on this contention. It is merely indicative of possible changes in conditions which can affect the ratio. On the other hand, the degree to which a reported statistic will differ from an actual statistic would tend to be more stable, particularly when the difference is due to one continuing underlying reason. For the above reasons, the taxpayer's calculation of bad order car days will be accepted here.

So far, then, bad order car days have been subtracted from total available car days. The next item of dispute is whether so-called surplus car days should also be subtracted in arriving at a car-day divisor. Surplus car days reflect days during which cars are not in use because there is no available load within the next 24 hours. The taxpayer eliminates such days; the defendant includes them in its car-day divisor, for the reason that "the cost of owning such cars did not cease merely because of non-use."

The defendant is certainly correct that the expenses of owning such cars continue even while the cars are not being used,¹⁸  but that is beside the point. We are not viewing expenses in a vacuum; we are allocating expenses against gross income, as is always done in a determination of net income. Presumably, all expenses are incurred in the process of and for the purpose of earning income. In general, a railroad earns income through the use of cars. Therefore, expenses must be assigned proportionately to the use of cars. The active car-day divisor — that is, available car days minus bad order car days and surplus car days — is the only figure which does this.

A simple numerical example should make it clear why the active car-day divisor is the right figure to use. Assume the taxpayer had 1,000 available car days in the year, 200 of them surplus car days, 400 car days of use in Mexico, 400 car days of use in the United States and total expenses for the year of $80,000.

If one used an active car-day divisor of 800, as the taxpayer would advise, then the cost per car day would be $100 ($80,000 ÷ 800). Multiplying this by the number of car days in each country, the total expenses assigned to each country are as follows:

 United States Mexico 400 (car days) 400 (car days) × × $100 (cost per car day) $100 (cost per car day) _______ _______ $40,000 $40,000

If a car-day divisor of 1,000 is used, as defendant would advise, then the cost per day would be $80 ($80,000 ÷ 1,000). Multiplying the cost per car day by the number of car days in each country, the expenses assigned to each country are as follows:

 United States Mexico 400 (car days) 400 (car days) × × $80 (cost per car day) $80 (cost per car day) _______ _______ $32,000 $32,000

What about the expense of $16,000 for the 200 surplus car days? Should it be assigned against United States income or Mexican income? There is really no more reason to assign all such costs against United States income than there is to assign all of them against Mexican income. The cost of not doing business, of cars not being used for income-earning activity, is not directly attributable to taxpayer's activities or income sources in either country. The cars were not used in Mexico just as much as they were not used in the United States.

Since such expenses cannot be directly assigned against the income of either country, they should be allocated ratably. Fifty percent of the use has taken place in each country, so a pro rata allocation adds $8,000 of expense to each country, for a total of $40,000 expense in each. This is, of course, the same figure that was arrived at by using the active car-day divisor, and thus illustrates the theoretical correctness of that divisor.

In summation, the above considerations lead to the adoption of a car-day divisor of 12,708,194, representing the total number of car days of actual use for the entire freight car fleet of the taxpayer in 1950. This amount was arrived at by multiplying the simple average number of freight train cars owned by taxpayer during 1950 (35,923) by 365, resulting in a figure of 13,111,895 available system car days. From this amount was subtracted: (1) 268,167 bad order car days, i. e., the number shown on taxpayer's weekly CS-60 reports to the A.A.R. in 1950, plus the 18.62 adjustment described supra, and (2) 135,534 surplus car days — giving the final answer of 12,708,194 car days.

Total Switching Expense In Connection With Freight Car Repairs. While this item would have been more logically dealt with in the freight car repair portion of the opinion, an understanding of the preceding bad order car issue was necessary for its understanding. A bad order car, that is, a car undergoing or awaiting repairs, must be switched from trainyards to repair tracks or to shops. Both parties agree that the expense of switching bad order cars is properly includable in the freight car maintenance category. The defendant computed such expense by deriving from the 1947 questionnaire the ratio of the switching expense to the total account 314. It then applied this ratio to the 1950 account-314 total.

The taxpayer used the 1947 questionnaire only for the purpose of determining that it cost $2.208 to switch a bad order car ($16 labor cost per hour multiplied by the 0.138 hours required to switch a car). It then multiplied this cost per car by the number of bad order cars that plaintiff had computed for 1950.

For the sake of consistency, since the taxpayer's method of calculating bad order cars was accepted, so should its method of calculating switching expense in connection with bad order cars. Since defendant's computation relies upon the ratio of 1946 switching expense to 1946 account-314 expense, it is subject to the same objection its computation of bad order car days was — that it assumed that the ratio of bad order car days to total car days remained constant for the years 1946 and 1950. For, if the number of bad order car days in 1946 was unusually high in relation to total car days, this would be reflected in an unusually high switching expense in relation to total account-314 expense.

For the above reasons, taxpayer's calculations are adopted in this opinion.

The foregoing conclusions have been summarized in a chart in the findings. Using the Mexican freight car expenses of $194,929 shown on the chart in finding 54, and the Mexican passenger car expenses of $28,358 as computed in finding 21, it is now relatively simple to determine the taxpayer's normal tax net income from Mexico for 1950:

 MEXICAN INCOME AND EXPENSES Freight car expense ............. $194,929 Passenger car expense ........... 28,358 ________ Total expenses ............ 223,287 ======== Freight car revenue ............. 330,251 Passenger car revenue ........... 47,510 ________ Total revenue ................... 377,761 Less expenses ................... 223,287 ________ Normal tax net income from Mexico ....................... 154,474

Inserting taxpayer's normal tax net income from Mexico into the formula mentioned earlier in the opinion for determining the maximum amount of the tax credit to which taxpayer is entitled, the allowable credit is shown to be $64,848, as follows:

 COMPUTATION OF CREDIT U. S. tax before adjustment × Normal tax net income from Mexico ___________________________________ in issue Total normal tax net income $7,541,670 × $154,474 __________ = $64,838 tax credit 17,967,666 ==================

Since the Commissioner had determined that the taxpayer was entitled to treat the $116,050 tax paid to Mexico only as a deduction from gross income, the taxpayer has, accordingly, overpaid its income taxes by the difference between the allowance of $64,838 of the Mexican tax paid as a foreign tax credit and the allowance of the full amount of such tax as a deduction. Plaintiff is thus entitled to judgment for this difference with interest according to law.