OPINION.
Offer, Judge:Although the subject matter of the present proceeding is complex and technical, its general purpose is one which can be simply stated in terms of everyday general experience. The central problem is one dealing with inventories which, in the case of department stores, are kept by the use of a unit which is the dollar. When prices or costs are constant, any increase or decrease in the dollar amount will furnish a reasonably reliable measure of the increase or decrease in the physical inventory. But it is common and sometimes bitter experience that in a period of rising prices the same amount of dollars will represent a smaller physical volume of goods, and the same physical volume will be represented by a higher number of dollars. The fundamental proposition about which the present proceeding revolves is the possibility and desirability of eliminating from an inventory kept in dollars that part which represents not an increase in physical inventory, but merely an expanded dollar figure at which that inventory is being carried.
A review of the characteristics of the retail method of inventories is necessary to appreciate the genesis of the present problem. Because of the number and diversity of individual articles constituting the stock of the typical department store, a system has been devised, approved,1 and generally practiced of reducing such articles to their lowest common denominator — a dollar figure. The inventories are taken and carried over from year to year in that form, rather than by specific items or quantities. And the annual stock-taking process constituting the ascertainment of the physical inventory is likewise, because of the complications of recording and maintaining actual costs for a constantly fluctuating stock, permitted to be based on computation and averaging.
That process is, roughly, to take from the price tags of the merchandise in stock on the day of the closing inventory a total for each department which, of course, represents the retail price on that day of all the goods then on hand. From this there is deducted the department’s average mark-on for the year, and the resulting figure, representing a theoretical conversion back to the cost of the goods on the shelves, is the closing inventory of the department in question.2
The mark-on or difference between the cost to the store and its selling price is obtained by recourse to the original invoices. During the year, as goods are received the invoices are extended by a notation of the retail price at which such goods are to be sold. These invoices form the records from which the mark-on for the year is computed by comparing the total of the costs they show with the total retail prices extended on them.
The dollar amount of the inventory so determined is then used like any other in computing the cost of goods sold, gross merchandising profit, and. ultimate net income of the store. The immediately apparent consequences of the retail method are that the goods in a single department are treated for inventory purposes as being entirely fungible, even though in fact they may, and generally do, differ considerably as to type, quality, and price. This follows from the fact that inventories are taken, entered, and maintained not in terms of individual items of merchandise, but only in the form of dollars, which are necessarily completely interchangeable. And also for the same reason, and what is perhaps a different method of saying the same thing, the problem of pricing the inventory, or more accurately, of “costing” it, has been eliminated, since the dollar figure arrived at as the make-up of the inventory is also taken as its cost without further conversion. A taxpayer who keeps his inventories in terms of specific items or quantities must convert them into dollars before they can be entered on the books and used in arriving at income. The retail method of inventories eliminates that intermediate step.
Petitioner’s position is founded on the proposition that, in a period of consistently advancing price levels, the result is to show as a profit what is merely an increase in the retail price and hence in the apparent cost of its inventories. In order to eliminate the consequent unrealized income, it claims to be entitled to the benefits of section 22 (d), added by the Revenue Act of 1939, which provides as follows:
(d) (1) A taxpayer may use the following method (whether or not such method has been prescribed under subsection (c)) in inventorying goods specified in the application required under paragraph (2):
(A) Inventory them at cost;
(B) Treat those remaining on hand at the close of the taxable year as being: First, those included in the opening Inventory of the taxable year (in the order of acquisition) to the extent thereof, and second, those acquired in the taxable year; and
(C) Treat those included in the opening inventory of the taxable year in which such method is first used as having been acquired at the same time and determine their cost by the average cost method.
(2) The method described in paragraph (1) may be used—
(A) Only in inventorying goods (required under subsection (c) to be inventoried) specified in an application to use such method filed at such time and in such manner as the Commissioner may prescribe; and
(B) Only if the taxpayer establishes to the satisfaction of the Commissioner that the taxpayer has used no procedure other than that specified in subparagraphs (B) and (C) of paragraph (1) in inventorying (to ascertain income, profit, or loss, for credit purposes, or for the purpose of reports to shareholders, partners, or other proprietors, or to beneficiaries) such goods for any period beginning with or during the first taxable year for which the method described in paragraph (1) is to be used.
(3) The change to, and the use of, such method shall be in accordance with such regulations as the Commissioner, with the approval of the Secretary, may prescribe as necessary in order that the use of such method may clearly reflect income.
(4) In determining income for the taxable year preceding the taxable year for which such method is first used, the closing inventory of such preceding year of the goods specified in such application shall be at cost
[[Image here]]
Slight and presently immaterial amendments were made retroactively by section 118 of the Revenue Act of 1942.
The effect of this provision is to permit the use of a concept described as last in, first out in the determination of the contents of an inventory — familiarly known and sometimes here referred to as Lifo— in place of a first in, first out, or Fifo, theory, without regard to whether the actual physical content of the inventory conforms to the assumptions so made.3
A comparable provision appeared for the first time in the Revenue Act of 1938, but was there limited to special classes of taxpayers.4
In order to equip themselves to make the last in, first out method available, members of the National Retail Dry Goods Association, including petitioner, concluded that it was necessary to eliminate from the amount of their inventory taken at retail that portion which represented a rise in prices during the year. For the purpose of determining what this conversion factor would be, they employed the National Industrial Conference Board to make a nation-wide survey of relative prices in the member stores and to compute for each department a figure designed to represent the year’s increase in retail prices in the standard departments of a typical store. The selection of appropriate sample items, the ascertainment of basic figures by means of questionnaires, the weighting and averaging, and the computation of the ultimate index numbers used are set forth in more detail in the findings of fact and need not be repeated at this point. Employing the ultimate figures resulting, petitioner adjusted the dollar amounts of its inventories at retail accordingly, and thereupon computed cost from that reduced figure. It filed an application for, and claimed on its return for the taxable year to be entitled to, this use of the Lifo method.
The conversion back to opening inventory price levels, computed by using the N. I. C. B. index figure, was confined to the quantity of goods, stated in terms of dollars, which was in the opening inventory. If any department showed a larger closing inventory, the difference would not be computed at the opening inventory price level, but at its actual retail price on the date of the closing inventory. An example of this operation, taken from petitioner’s brief, is set forth in the margin.5
Respondent raises no question as to petitioner’s compliance with the technical requirements of the section such as the filing of the necessary application and use of the system in its financial reports. The controversies calling for decision can be reduced to whether the statutory grant of leave to use the last in, first out method is applicable to a taxpayer maintaining inventories by use of the retail system; and, if so, what method represents a permissible means of securing the benefits of the section. But in arriving at an answer to the latter question, consideration of a number of alternative possibilities is not required.
Refined to its simplest terms, the fundamental question is whether the identification of specific articles in an inventory is the prerequisite for application of the last in, first out method. Respondent’s contention that it is seems supported neither by logic nor by legislative intent. His principal reliance, the interpretative regulations, will not withstand examination. Those provisions most nearly dealing with the present question were inserted long after the tax year and, in fact, after the present proceeding had been brought to hearing. T. D. 5504, 1946-1 C. B. 52, 53. Their retroactive application seems unwarranted. Helvering v. Reynolds Tobacco Co., 306 U. S. 110. All of the regulations6 are as consistent with petitioner’s position as they are with that of respondent. See, e. g., Regulations 103, secs. 19.22 (c)-8 and 19.22 (d)-3. It is simpler and more rewarding to seek the meaning of the statute itself than of ambiguous and largely irrelevant administrative interpretations.
As to the language of the statute, it is urged that the term “goods” is used advisedly and “in the common and ordinary sense.” But the dictionary7 gives as synonyms “wares and merchandise.” Wares may have a slightly archaic .flavor, but merchandise is a term in the day-today vernacular, and it is difficult to conjure up one more accurately describing in the generic sense the stock in trade of a typical department store.
It is true that the legislative history indicates agitation for the amendment which became section 22 (d) as having its origin among certain groups of producers for whose special problem the Lifo solution would be particularly appropriate. It is also indicated that in these instances taxpayers followed generally an inventory system which dealt with specific articles, and that they were faced also with the problem of attributing a cost to the articles so ascertained.8 But it does not follow from this that, when the statute was amended to acquire its present all-embracing character, it was designed to cover only the industries which originally sought its enactment. The conclusion is rather to be drawn that, like all revenue legislation, particularly of a relief character, the purpose of the lawmakers was to have it apply in general terms to all those coming within its provisions.9
Similar results have been sought under what is known as the “minimum,” “cushion,” “base stock,” .or “constant average” method of inventories. It has been insisted that such a process does violence to the principle of annual accounting periods, since it has the effect of postponing the profit in a high price year and absorbing it in a subsequent period of low prices. For that reason it was held that the minimum inventory method does not clearly reflect income even in a base stock industry such as a steel fabricator and builder which kept its inventory records in terms of tons. Kansas City Structural Steel Co., 11 B. T. A. 877; and see the same case on review, Lucas v. Kansas City Structural Steel Co., 281 U. S. 264. It is, however, to precisely this type of taxpayer that apparently even respondent concedes the Lifo amendment embodied in section 22 (d) was directed. No such provision was in existence at the time the Kansas City Structural Steel Co. case was decided, and it must be accepted as the present legislative opinion that the broad doctrine for which that case stands is no longer authoritative. Unless we can discover some logical reason why the Lifo method, which is concededly available in these situations, is not equally applicable to inventories as to which ascertainment and valuation are a single process, the assumption seems to us required that petitioner may bring itself within the terms of section 22 (d) by conforming to its requirements.
We have already had occasion to note that the primary characteristic of inventories maintained under the retail method is the use of dollars as the measure of quantity. In a period of radically and rapidly changing price levels, the dollars so attributed to an inventory will be neither the current cost of replacement nor the presumptive level at which the inventory will be diminished by sales. But this is precisely the basic problem faced by the industries to which the legislation was concededly directed. The method of its application in those cases is simple, while with respect to such taxpayers as petitioner it is relatively complex. But, since the fundamental situation to which the provision is directed is identical, and since no portion of its operation or language is inherently inapplicable, it should, if possible, be given effect in all cases presenting the problem with which it deals. It remains to be considered whether the practical impediments to its application are so great as to eliminate the possibility of its use in such situations as that now before us.
The process envisaged by Lifo involves not so much the ascertainment of cost as the ascertainment of what it is of which we are to discover the cost. The last in, first out formula assumes that the merchandise remaining in the inventory is that which was first purchased. When that assumption is made, the process of ascertainment becomes difficult in the case of a retail merchant primarily because of the complications of the retail method itself.
If we were dealing with a fabricator or manufacturer, the first step would be to determine which merchandise it is.to which a cost is to be attributed, and the second, to determine that cost. The corresponding process engaged in by petitioner is to reduce the price level of the retail stock to that prevailing as of the opening inventory, and thereby to identify the merchandise remaining in inventory at the close of the year as that constituting inventory at the beginning of the year to the extent of the size of the opening inventory. That this is done by dealing with the merchandise stated in terms of dollars rather than of numbers or quantities is a requisite of the aspect of department store accounting which relies for its inventory volume on a statement in dollars alone.
But it will not do to say that this method does not truly reflect a taxpayer’s inventory or that the use of dollars instead of other measures of quantity has the effect of distorting the inventory content. The retail method has been used with respondent’s complete approval for too long a time for the assumption to be permissible that the contents of an inventory can not be satisfactorily represented for all purposes by its expression in dollars only; or for the result to be acceptable, unless entirely necessary, that the use of that method for the statement of an inventory forecloses the user from other benefits.
Having identified the closing inventory as containing the carryover merchandise from the opening inventory — that is, making the assumption of Lifo that the latest purchases were the first sold and the earlier purchases remained on hand — the process of putting a cost figure against the merchandise in a closing inventory as carried out on petitioner’s present theory is simple. It calls merely for the ascertainment of historic cost; in other words, it is the same cost which the same merchandise carried as of the opening inventory of the current year; and likewise, of course, the previous closing inventory. This is accomplished by resorting to the inventory itself, which is habitually carried at two figures, retail and cost. That proportion of the total cost is taken which the goods remaining out of the opening inventory bear to the total opening inventory; or, identically, by applying to the remainder of the opening inventory at retail the mark-on resulting from comparing the retail and cost figures of that inventory.
A summary description of the process might thus be stated as ascertaining the quantity of merchandise in the closing inventory, which is the equivalent at comparable prices of that in the opening inventory; applying to that portion of the closing inventory its historic cost; and applying to the remainder its current cost computed by processes which have become traditional under the retail method. Merchandise which is assumed to have been in both inventories is accordingly handled at the cost of the opening inventory. Any decrease is treated as having the same cost. Any increase is included at the cost of the current year’s purchases.
It is difficult to see that in either of these necessary steps — the selection of the merchandise constituting the inventory, or the assignment to that merchandise of a cost figure — such violence is done to recognized principles of retail inventory practice as to demonstrate either from principle or application that Congress intended the Lifo provisions to be inapplicable to the accounting procedure of retail merchants. The computations involved may be somewhat extensive, but they lack neither adequate clarity nor logic, and they can not be discarded as unduly burdensome in the light of the permissive character of the statutory option. We conclude that adaptation of the Lifo theory to inventories maintained in terms of dollars through the approach generally employed by the present petitioner is permissible and proper within the provisions of section 22 (d).
One aspect of the mechanics resorted to by petitioner and other members of the Retail Dry Goods Association, however, requires further mention. That is the selection of an index figure to be applied in reconverting to opening inventory prices. The record contains, in addition to the N. I. C. B. index, somewhat corresponding figures derived and published by a public body, the Bureau of Labor Statistics of the United States Department of Labor, not only for the Nation as a whole, but for individual cities, including Baltimore, in which this petitioner is located. It also contains material showing the weights attached by the Bureau of Labor Statistics to various items used to compute its index and which are found in the inventories actually carried by petitioner; and the prices as of the opening and closing inventories at which, among others, petitioner was carrying these or closely comparable items.
Petitioner has refrained from restricting its search for relief to the National Industrial Conference Board index figure or to the formula used by it in preparing its return. It states its willingness to make the adjustment by any of the possible alternative methods.
The adoption of appropriate details to accomplish the general pur* pose of applying the Lifo method to retail merchants could well be the subject of an administrative regulation. It might then be that any reasonable and practical method directed by the Commissioner would be mandatory to the exclusion of others not thought by him to be adequate or authoritative.
No such regulation has been issued. But respondent takes the position for purposes of this proceeding — not to be interpreted as generally applicable or effective beyond the confines of the present litigation— that there is one method of computing such an index figure which he finds less objectionable than others. That is, to take the opening and closing prices, department by department, which the record shows petitioner was actually charging for such of the items as it recorded, and to give such price changes equal weights — in other words, to compile an unweighted average — in computing the price rise for each individual department as a whole. Since petitioner has reiterated its willingness to adopt any one of the possible alternative methods for which the figures appear, and respondent has expressed his preference for that method, we need record no opinion as to the respective virtues of the various alternative possibilities. The procedure just described should be employed for computing price changes in petitioner’s store and, to afford an opportunity to compute the final inventory figures under the Lifo method.
Decision will be entered under Rule 50.
Reviewed by the Court.
T. D. 3058, 1920-3 C. B. 72; Regulations 103, sec. 19.22 (c)-8.
“The underlying principle of the retail Inventory method Is the determination of the inventory value by applying the complement of the percentage of mark-up to the retail value of the inventory.” Desmonda, Inc., 15 B. T. A. 738, 752.
See footnote 6, infra.
Revenue Act of 1938, section 22 (d) :
“(d) Inventories in Certain Industries.—
"(1) Producers and processors op certain non-vereous metals. — A taxpayer shall bo entiUed to elect the method of taking inventories provided in paragraph (2) if his principal business is—
“(A) Smelting non-ferrous ores or concentrates, or .refining non-ferrous metals, or both; or
“(B) Producing brass, copper products, or brass products, or any one or more of them, not further advanced than rods, sheets, tubes, bars, plates, or strips.
“(2) Inventories op Raw Materials. — A taxpayer entitled to elect, and who has so elected, shall, in taking his inventory as of the close of any taxable year beginning after December 81, 1938, of raw materials which are—
“(A) used in a business described in paragraph (1) ; and
“(B) not yet included in goods in process or finished goods ; and
“(C) so intermingled that they cannot be identified with specific invoices; treat such raw materials remaining on hand as being: First, those included in the inventory as of the beginning of the taxable year (in the order of acquisition) to the extent thereof, and second, those acquired in the taxable year, in the order of acquistion.
“(3) Tanners. — A taxpayer whose principal business is tanning hides nr skins, or both, shaU be entitled to elect (with respect to any taxable year beginning after December 31, 1938) the method provided in paragraph (2) as to the raw materials (including those Included in goods in process and in finished goods) in the business of tanning hides, or skins or both, if so intermingled that they cannot be identified with specific invoices.
[[Image here]]
“* * * The procedure was as follows: The Inventory of a department, at the close of the year, having been taken at the retail value, was stated in terms of an equivalent retail value at the beginning of the year by dividing the retail value of the inventory on hand at the end of the year by the index of retail price change determined by the National Industrial Conference Board for that department. By this means a comparison was made of the amount of goods on hand at the end and at the beginning of the year. For example, in the men’s shoe department, the retail Inventory at the end of the year was $35,663.38. The National Industrial Conference Board index showed a rise in prices of men’s shoes from 100 at January 31, 1941, to 110.52 at January 31, 1942. Dividing the retail value of the closing Inventory by the price change, it was determined that these shoes had a retaU value at January 31, 1941, of $32,268.71. On that date, however, the taxpayer had a retail inventory of only $25,837.92 of men’s shoes, thereby establishing that there had been an increase in the inventory, stated at January 31, 1941, retail prices, in the amount of $6,430.79. On a Lifo basis this increase would be stated at its current cost, while that part of the inventory which replaced the goods on hand at January 31, 1941, would be stated at their cost in the January 31, 1941, inventory. This cost of the January 31, 1941, inventory was $14,899.18. This is the cost basis used in the closing inventory for the taxable year ending January 31, 1941, and in the opening inventory for the year ending January 31, 1942. The increase in the inventory of $6,430.79 at January 31, 1941, retail basis, was restored to the January 31, 1942, retail basis by multiplying it by the NICB index of 110.52 to produce $7,107.31, being the January 31, 1942, retail value of the Increase in the Inventory. This increase was then reduced to cost in the customary manner by the use of the percentage of markup of that department which was 42.3. The cost of the increase of goods in the closing inventory of this department was computed thus at $4,100.92. This cost of the increased goods in the inventory was added to the cost of the inventory of goods on hand at the beginning of the year of $14,899.18 to produce a Lifo Inventory for the goods in this department of $19,000.10. In many of the departments there was a decrease in the amount of goods on hand, and the Lifo inventory was computed on a similar principle."
E. g., Regulations 103, sec. 19.22 (d), as amended by T. D. 5504, 1946-1 C. B. 52, 53:
"Sec. 19.22 (d)-l • • • This elective inventory method is not dependent upon the character of the business in which the taxpayer is engaged, or upon the identity or want of identity through commingling nt any of the goods on hand, and may be adopted by the taxpayer as of the close of any taxable year beginning after December 31, 1938. [Emphasis added.]
[[Image here]]
[[Image here]]
*‘SKC. 19.22 í (1}- 2. ♦ • *
"(4) Goods of the specified type on hand as of the close of the taxable year in excess of what were on hand as of the beginning of the taxable year shall be included in the dosing inventory, regardless of identification with specific invoices and regardless of specifto cost accounting records, at costs determined pursuant to the provisions of rule (A) or rule (B) hereof, dependent upon the character of the transactions in which the taxpayer is engaged. [Emphasis added.]
“Bule (X). In the case of a taxpayer engaged in the purchase and sale of merchandise, such as a retail grocer or druggist, or engaged in the initial production of merchandise and its sale without processing, such as a miner selling his ore output without smelting or refining, such costs shall be determined—
• ****••»»
Webster’s New International Dictionary, 2d Ed.
Conference Report on Revenue Bill of 1938 (H. Rept. No. 2330, 75th Cong., 3d sess.), pp. 34-35; Hearings before Ways and Means Committee on Revenue Revision 1938, pp. 888-890, 1181-1184; Hearings before Finance Committee on Revenue Act of 1938, pp. 143-167, 175-170, 429-430, 480, 483. 484-487; 83 Cong. Rec. 5043-5044.
"• • • Under the bill the option is extended to all taxpayers who use it, apply for it, and use it consistently, regardless of the business in which the taxpayer is engaged." Senate Finance Committee Report, S. Rept. No. 648, 76th Cong., 1st sess., p. 6.