Knight-Ridder Newspapers, Inc. v. United States

743 F.2d 781, 54 A.F.T.R.2d (RIA) 6120, 1984 U.S. App. LEXIS 18021

• Court: Court of Appeals for the Eleventh Circuit
• Decided: October 3, 1984
• Docket: 82-6157
• Status: Published

Opinion

GOLDBERG, Circuit Judge:

The United States appeals from a decision of the District Court, holding that Knight-Ridder Newspapers (“Taxpayer” or “Knight-Ridder”) is entitled to recover certain federal income tax payments with respect to its 1972, 1973, and 1974 taxable years. The payments represent various deficiencies assessed by the Internal Revenue Service and involve three separate legal issues. Taking them in order, we hold first that the Commissioner of Internal Revenue (“Commissioner”) did not abuse his discretion under I.R.C. § 446 when he determined that the cash method of accounting did not clearly reflect the income of two of Taxpayer’s subsidiaries. The Commissioner acted well within his authority in requiring those subsidiaries to adopt the accrual method.

Second, we hold that Taxpayer did not properly elect to use the “Guideline Class Life System” for depreciating the press equipment of several of its subsidiaries. Taxpayer failed to substantially comply with the regulatory requirements because its tax returns contained nothing to alert the Commissioner that an election had been made.

Finally, we hold that the use of an advertising rebate reserve is a “method of accounting,” such that the abandonment of the method triggers compensatory adjustments pursuant to I.R.C. § 481.

FACTS AND PROCEDURAL HISTORY

Taxpayer runs a large newspaper chain. It was formed on November 30, 1974, as a result of the merger of Knight Newspapers, Inc. (“Knight Newspapers”) and Rid-der Publications, Inc. Knight Newspapers in turn was the sole owner of a number of subsidiaries whose main business involved the publication of newspapers. For the taxable years 1972 and 1973, Knight Newspapers filed consolidated tax returns, reporting the income of the following subsidiaries: Beacon Journal Publishing Company, Tallahassee Democrat, Inc., Macon Telegraph Publishing Company, Boca Ra-ton News, Inc., Philadelphia Newspapers, Inc., Detroit Free Press, Inc. (“Detroit Free Press”) and Knight Publishing Company. The Lexington Herald Leader Company was a party to the consolidated return for 1973.

In addition, Knight Newspapers acquired the R.W. Page Corporation (“Page”) on October 1, 1973. Page had been actively engaged since 1927 in the business of publishing newspapers in Columbus, Ga. Page had also operated as a separate business the Bradenton Herald, Inc. (“Bradenton”), which published a newspaper in Bradenton, Florida. On November 1, 1973, Knight Newspapers organized Bradenton as a subsidiary of Page. Thus, Page and Braden-ton were parties to the 1973 consolidated return of Knight Newspapers. Page was a party to the return for the period from October 1 through December 31, 1973; Bradenton was a party for the period November 1 through December 31.

Finally, Knight-Ridder filed a consolidated return in 1974, covering among other things, the operations of the same subsidiaries whose operations had been covered by the 1973 return filed by Knight Newspapers,

ACCOUNTING METHODS OF PAGE AND BRADENTON

„ , . ... One of the disputes m this case concerns whether Page and Bradenton should use the cash and disbursements method of accounting or the accrual method. To simplify the distinction: under the cash method, expenditures are deducted when made and payments included in income when received. Under the accrual method, by contrast, income and deductions are generally recognized when an obligation becomes fixed, rather than when cash payments are actually made. Thus, accounts payable and accounts receivable are taken into account when the debts are fixed, even though no money has changed hands. 1

Throughout their history, Page and Bradentón have reported their income for tax purposes using the cash method. They have used the accrual method in preparing financial statements. During an audit of Page’s tax returns for the years 1964 through 1966, Page proposed to the I.R.S. examining agent that the company change its method of accounting to the accrual method. The agent considered the proposal but did not require 2 a change of method f°r those years. The Service later audited Page’s returns for 1969 and 1970 and again did not object to the use of the cash method. During that period, the cash and accrual methods reached substantially the same results in measuring the before-tax income of Page and Bradenton. 3

After Knight Newspapers bought Page and Bradenton in late 1973, however, a substantial discrepancy appeared between the two methods. In the tax year ending December 31, 1974, the cash method netted approximately $1,400,000 in taxable income for Page and Bradenton, while the accrual method would have yielded approximately $1,900,000, a difference of $500,000 (according to the calculations of Mr. William Pruitt, an accounting expert who testified for Taxpayer). See Plaintiff’s Exhibit 13, at 4; Plaintiff’s Exhibit 20, at 4; Plaintiff’s Chart # l. 3

This distortion was the result of Knight-Ridder’s investment of capital in the business of Page and Bradenton, building up accounts receivable and inventories of raw materials (newsprint and ink). The inventories in particular increased by about $240,000 in 1974, from $210,000 to $450,-000. The cash and accrual methods did not show such gross discrepancies again during the years 1975 to 1979.

In an audit of the 1973 and 1974 tax returns of Knight-Ridder and its subsidiaries, the Commissioner determined that the cash method did not clearly reflect income for Page and Bradenton in those years. He found that a timing distortion resulted because of the accounts receivable and inventory fluctuations. Therefore, he required Page and Bradenton to switch to the accrual method for the short tax period ending December 31, 1973 (i.e., the months after the take-over by Knight-Ridder), and the full tax year ending December 31,1974.

In line with that decision, the Commissioner required Taxpayer to include in its income for those years the adjustments prescribed by section 481 of the Internal Revenue Code, 26 U.S.C. § 481. Section 481 adjustments are designed to compensate for the duplication or omission of items of income or expense that occurs when a taxpayer changes its method of accounting. Taxpayer paid the tax resulting from these adjustments and, after satisfying the requisite administrative claim procedure, filed this suit for a refund.

The case was tried to the bench in the Southern District of Florida. The trial court held that the Commissioner had consented to Page’s and Bradenton’s use of the cash method and that, in any event, the method clearly reflected their income. Therefore, the Commissioner had abused his discretion in requiring Page and Bra-denton to use the accrual method for 1973 and 1974.

PRESS EQUIPMENT DEPRECIATION

The Internal Revenue Code has always permitted depreciation to be taken on the “facts and circumstances” method. That method requires a taxpayer to assign a separate useful life to each capital item based on a judgment as to how long it is expected to last. In recent years the Code and regulations have permitted taxpayers to elect various “class life” systems under which whole classes of assets may be assigned the same specified, and artificially short, useful life. The “Guideline Class Life System” applies to assets put into service before 1971. Since at least 1970, Taxpayer’s policy has been to take the most rapid depreciation allowed on the press equipment of its subsidiaries. It advised its tax accountant of the policy; and its subsidiaries accurately computed depreciation on their own books in accordance with the Guideline Class Life System.

A number of the subsidiaries, however, did not indicate on their tax forms for 1972 and 1973 that they were electing to use the system. Knight-Ridder’s consolidated returns were prepared by a national independent accounting firm. A local office of the firm prepared the returns for the local subsidiary of Knight-Ridder. Then, the local office sent the returns to the Miami office where all of the forms were collected and attached to Knight-Ridder’s consolidated return.

A number of the forms as prepared did not comply with Treasury Regulations governing the election of the Guideline Class Life System, Treas.Reg. § 1.167(a)-12, 26 C.F.R. § 1.167(a)-12 (1983). The Regulations require the taxpayer to “check” a box provided in Schedule G of its Corporate Tax Return and file a completed Form 5006 with the return. For their 1972 and 1973 taxable years, two of the subsidiaries — the Detroit Free Press and the Knight Publishing Company — checked the appropriate box and filed Form 5006. However, for the same years, neither Knight Newspapers nor any of its other seven subsidiaries 4 checked that box or filed Form 5006. Nor did the information required by the Treasury appear anywhere else in their returns. Most of them did, however, check a similar box electing to use the “Class Life ADR System,” which applies only to assets put into service after 1971. 5

The Commissioner concluded that Knight Newspapers and the seven subsidiaries had not elected the Guideline Class Life System for pre-1971 assets. He recomputed their depreciation according to the “facts and circumstances” test, determining that the proper useful life for the press equipment was 18 years rather than the 11 years prescribed by the Guideline Class Life System. 6 The Commissioner assessed deficiencies which the Taxpayer then paid.

The District Court, however, noted that Taxpayer and its subsidiaries had intended to use the Guideline Class Life System, had computed their depreciation in accordance with the system, and had properly maintained their own books and records. Therefore, the court held that Taxpayer had substantially complied with the election requirements and legislative purpose of the class life statute 7 and regulations. 8 Taxpayer was entitled to depreciate its press equipment pursuant to the system.

ADVERTISING REBATE RESERVE

Taxpayer’s subsidiary, the Detroit Free Press, maintained its books and records on the accrual basis. In its advertising contracts, the Free Press provided that an advertiser would qualify for a lower per line rate if its annual volume of advertising exceeded a specified level. The advertiser would be entitled to a rebate of the excess amounts previously paid at the higher rate. Because the measuring year did not always coincide with the Free Press’s tax year, rebates were often paid in a different tax year from that in which the advertising revenues were earned.

The Free Press established a reserve on its books for anticipated advertising rebates and each year added an amount based on an estimate of what the rebate liability would be for that year. Deductions were taken when the amounts were added to the reserve. Later, when rebates were paid, they were charged against, and served to reduce, the level of the reserve. As of January 1, 1972, the balance of the reserve stood at $120,000. In each of the years 1972, 1973, and 1974, the Free Press’ estimated deductions, actual rebate payments, and ending reserve balance were as follows:

Ending Taxable Balances Increases Year Estimated Actual Cash In Reserve for Taxa-Ended Rebates Disbursements Accounts ble Year

12/31/71 $120,000.00

12/31/72 $369,000.00 $339,000.00 150,000.00 $ 30,000.00

12/31/73 524,671.00 464,671.00 210,000.00 60,000.00

12/31/74 737,866.00 622,866.00 325,000.00 115,000.00

Upon audit of Taxpayer’s 1972 return, the Commissioner determined that the Free Press was not entitled to maintain the reserve for 1972 or thereafter. He eliminated the deductions for 1972 through 1974 and included in Taxpayer’s 1972 income the opening reserve balance of $120,000. Taxpayer objected that the statute of limitations barred the Commissioner from challenging any improper deductions prior to 1972. The Commissioner, however, concluded that he was entitled to correct the $120,000 balance pursuant to I.R.C. § 481. He felt that the reserve was an accounting method and, therefore, that Taxpayer had changed its accounting method when it ceased to use the reserve. Since it had changed the method, it was required to make adjustments under section 481 to prevent a double deduction. Otherwise, the change of methods would have provided a deduction when the $120,000 was first put into the reserve and a second deduction when rebates were actually paid after 1972.

Taxpayer paid the deficiencies and challenged only the $120,000 adjustment when it brought this action in the District Court. It pressed the statute of limitations argument, contending that section 481 does not apply to the rebate reserve which is not a method of accounting. The trial court agreed that the reserve is not an accounting method and held that the Commissioner is barred from recovering the $120,000 in deductions.

ISSUES

To summarize the issues on this appeal, the trial court held that

(1) the Commissioner abused his discretion in requiring Page and Bradenton to change to the accrual method;

(2) Taxpayer substantially complied with the requirements for electing the Guideline Class Life Depreciation System; and

(3) an advertising rebate reserve is not a method of accounting. 9 The Commissioner challenges each of these holdings on the law. We agree with the Commissioner on each issue and reverse.

ACCRUAL VERSUS CASH ACCOUNTING

A. The Cash Method, the Accrual Method, and Section 446.

The initial issue in this case concerns the authority of the Commissioner to force a taxpayer to change accounting methods, in particular to change from the cash to accrual method. These are the two most common accounting methods and could be said to emblemize the polar nature of the human spirit. The cash method — simple, plodding, elemental — stands firmly in the physical realm. It responds only through the physical senses, recognizing only the tangible flow of currency. Money is income when this raw beast actually feels the coins in its primal paw; expenditures are made only when the beast can see that it has given the coins away.

The accrual method, however, moves in a more ethereal, mystical realm. The visionary prophet, it recognizes the impact of the future on the present, and with grave foreboding or ecstatic anticipation, announces the world to be. When it becomes sure enough of its prophecies, it actually conducts life as if the new age has already come to pass. Transactions producing income or deductions spring to life in the eyes of the seer though nary a dollar has moved.

The Internal Revenue Code, the ultimate arbiter, stands to the side, shifting its eyes uneasily from the one being to the other. The Code is possessed of great wisdom and tolerance. It knows that man must generally choose his own way. Therefore, it leaves to the Taxpayer the original choice of which accounting method to use. Section 446(c) specifically authorizes both the cash and accrual methods. 26 U.S.C. § 446(c).

Yet the Code also understands that either extreme possesses inherent weaknesses and can become blinded to reality. Thus the Code and subsequent Treasury Regulations empower the Secretary of the Treasury and the Commissioner of Internal Revenue to cure the blindness. Section 446(b) of the Code provides that if

the method used does not clearly reflect income, the computation of taxable income shall be made under such method as, in the opinion of the Secretary, does clearly reflect income.

26 U.S.C. § 446(b). Treasury Regulations Section 1.446-l(a)(2) adds that “no method of accounting is acceptable unless, in the opinion of the Commissioner, it clearly reflects income.” 26 C.F.R. § 1.446-l(a)(2).

The courts have consistently interpreted this mandate to give the Commissioner “broad discretion to set aside the taxpayer’s method if, ‘in [his] opinion,’ it does not reflect income clearly.” Thor Power Tool v. Commissioner, 439 U.S. 522, 540, 99 S.Ct. 773, 785, 58 L.Ed.2d 785 (1979). His decision “should not be interfered with unless clearly unlawful.” Id. 99 S.Ct. at 781; Lucas v. American Code Co., 280 U.S. 445, 449, 50 S.Ct. 202, 203, 74 L.Ed. 538 (1930). A court may not “overturn his determination unless the evidence clearly shows that he has abused his discretion.” Drazen v. Commissioner, 34 T.C. 1070, 1076 (1960); accord Loftin and Woodard v. United States, 577 F.2d 1206, 1229 (5th Cir.1978).

Of course, in deciding whether the Commissioner has abused his discretion, we immediately face an age-old philosopher’s dilemma: how can we mere mortals know who sees the truth most vividly? How can we know whether the primal cash method or the mystical accrual method sees income more clearly without knowing what income really is? If it is really cash on hand, then the cash method is more accurate. If it is really fixed obligations, then the accrual method is more accurate. By embracing both conceptions, the Code provides no general baseline against which to assess the accuracy of an accounting method. In effect, we risk being led in circular fashion to arbitrarily choose one method as accurately reflecting income. When another method differs from it, that other must not clearly reflect income.

B. Inventories and the Accrual Method

Fortunately, we need not be so arbitrary in this case. The Code and regulations do provide guidance in the case of businesses that sell merchandise. Such businesses are generally required to use inventories, and where the taxpayer uses inventories, it must use the accrual method. Section 471 of the Code provides:

Whenever in the opinion of the Secretary the use of inventories is necessary in order clearly to determine the income of any taxpayer, inventories shall be taken by such taxpayer on such basis as the Secretary may prescribe as conforming as nearly as may be to the best accounting practice in the trade or business and as most clearly reflecting the income.

26 U.S.C. § 471. The Regulations implementing Section 471 provide:

In order to reflect taxable income correctly, inventories at the beginning and end of each taxable year are necessary in every case in which the production, purchase, or sale of merchandise is an income-producing factor. The inventory should include all finished or partly finished goods and, in the ease of raw materials and supplies, only those which have been acquired for sale or which will physically become a part of merchandise intended for sale * * *.

26 C.F.R. § 1.471-1. Once a taxpayer uses such inventories, it must adopt the accrual method unless the Commissioner consents to another:

In any case in which it is necessary to use an inventory the accrual method of accounting must be used with regard to purchases and sales unless otherwise authorized under subdivision (ii) of this sub-paragraph. 10

26 C.F.R. § 1.446-l(c)(2)(i).

The reasoning behind this regulatory scheme is straightforward. According to accounting wisdom, the income realized from the sale of merchandise is most clearly measured by matching the cost of that merchandise with the revenue derived from its sale. See Chirelstein, supra note 1, ¶ 12.03, at 221 (quoting the American Institute of Certified Public Accountants); Bittker, supra note 1, ¶ 105.41, at 77, 82. In order to achieve such a matching of revenue and cost, it is necessary to keep an inventory account reflecting the costs of merchandise, raw materials, and manufacturing expenses. These costs are not deducted immediately when paid but are deferred until the year when the resulting merchandise is sold. 11

To make the matching complete, the taxpayer must report income on the accrual method. That method helps to ensure that income from the sale (like the inventory costs) is reflected in the year of the sale. For example, if the sale is made on credit, the accrual method nevertheless treats the income as accrued and reflects it when the sale occurs. See Caldwell v. Commissioner, 202 F.2d 112, 114 (2d Cir.1953). The prophetic skills of the accrual shaman permits it to recognize both income and deductions in the same year.

By contrast, the primal cash method is unable to achieve such a mystical joinder of inventory deductions and credit sale income. To be sure, the cash method could theoretically operate in tandem with inventories. The beast could conceivably close its eyes to deductions until the year of the sale. It could never learn, however, to prophesy future cash payments. If there were a credit sale, the beast could not grasp income and deductions simultaneously in its rugged paw. See id. The goal of matching costs and revenues would fail. 12

Thus, “the ‘inventory’ system is generally recognized as synonymous with the accrual system of accounting.” 2 Mertens, supra note 1, § 16.03, at 5. If the taxpayer must use inventories, the Commissioner may also require it to adopt the accrual method. 26 C.F.R. § 1.446-l(c)(2)(i). 13

C. The Accounts of Page and Braden-ton: When Must a Newspaper Keep Inventories?

1. The Newspaper as Merchandise

Returning to our case, the initial question arises whether Page and Bradenton were required to keep inventories. Taxpayer argues that Page and Bradenton are not the type of merchandisers envisioned by the inventory regulations. These companies sell an extremely perishable commodity; a two day-old newspaper is stale. Thus, if they reported on the inventory method, they would have virtually no inventories of finished goods. In addition, most of the companies’ revenues are not derived from sales to readers. Almost 80% of total revenues come from advertisers. Taxpayer contends that the newspaper business is most accurately described as a service business, providing information to its readership and running advertisements for its clients.

We hold, however, that these companies do come within the requirements of Treas. Reg. § 1.471-1. The sale of merchandise is an “income-producing factor” for Page and Bradenton. Twenty percent of their revenues come directly from the purchasers of newspapers. Moreover, the sheer cost of producing the physical pages (let alone the intellectual content) is significant in comparison to total revenues. The cost * of newsprint and ink alone in 1973 and 1974 was $2,456,428, or 17.6% of the total cash receipts. 14

In Wilkinson-Beane, Inc. v. Commissioner, 420 F.2d 352 (1st Cir.1970), the First Circuit faced a similar “mixed service and merchandising business” and held that the sale of merchandise was an income-producing factor within section 1.471-1 when the cost of the physical merchandise averaged 15.1% of total cash receipts. The taxpayer ran a funeral home. It provided caskets as part of its funeral service, but did not charge for them separately. Instead, the cost of a casket was included in the total price for the service. The taxpayer contended that it ran a service business and that the sale of caskets was not an income-producing factor.

The court, unable to separate out the relative prices of caskets and service, compared the cost of the caskets to total receipts. The court held that the cost of the caskets (15.1% of revenues) 15 was high enough to make them a substantial income-producing factor, even assuming that the taxpayer sold them at cost. Id. at 355. 16 Similarly, in our case, where the cost of raw materials for the newspapers was 17.6% of total revenues and the actual sales price accounted for 20% of revenues, we hold that the sale of newspapers was a material income-producing factor.

We do not believe the newspaper is any less merchandise because it provides information. Books are classic examples of merchandise requiring the use of inventories, yet their primary value likewise inheres in the message they communicate. 17

2. The Requirement that Inventories or Inventory Fluctuations Be Substantial

Taxpayer contends that inventories (and the accrual method) are unnecessary in the newspaper business because the overall inventory account would be insubstantial. The number of finished goods held in inventory would be virtually nil, and the value of raw materials reflected in the account would be insignificant relative to total revenues. The Tax Court has held that inventories are not necessary when they would be “so small as to be of no consequence.” Ezo Products v. Commissioner, 37 T.C. 385, 393 (1961). 18

We agree that in deciding whether a taxpayer must adopt inventories, the size of the account and fluctuations therein are relevant. Admittedly, section 1.471-1 does not explicitly direct us to consider whether inventories are insignificant. The regulation requires inventories in “every case in which the ... sale of merchandise is an income producing factor.” 26 C.F.R. § 1.471-1. Nevertheless, given that the ultimate goal of the regulation is “to reflect taxable income correctly,” id., we hold that that purpose is not served where inventories and inventory fluctuations would be de minimis and have virtually no effect on the reflection of income. See Bittker, supra note 1, ¶ 105.4.1, at 88. On the other hand, if either the absolute level of the inventory account or its fluctuation during the year would be substantial, then the taxpayer must use inventories if it meets the other requirements of section 1.471-1.

In the case at bar, we need not decide whether the absolute level of Taxpayer’s account would have been significant enough. The fluctuation in the inventory of newsprint and ink clearly would have been substantial in 1974. Such a fluctuation must be an increase and must have a significant effect on taxable income. Taxpayer suggests that we compare the fluctuation to total revenues. See Reply Brief at 22, note *. We believe, however, that the proper standard for comparison is taxable income. The Internal Revenue Code authorizes the Commissioner to require inventories “in order clearly to determine the income of [the] taxpayer.” I.R.C. § 471, 26 U.S.C. § 471. This language apparently refers to taxable income because inventories have an impact only at that level; total revenues are the same whether or not the taxpayer keeps an inventory account. The implementing regulations are even more specific. They provide that inventories are necessary “[i]n order to reflect taxable income correctly.” 26 C.F.R. § 1.471-1.

Since the goal of the inventory provisions is to reflect taxable income clearly, we hold that the Commissioner may require inventory accounting when fluctuations in the inventory would be substantial relative to taxable income. Cf. Ezo Products v. Commissioner, supra, 37 T.C. at 387, 388, 392 (inventory method required where fluctuation in inventories substantial relative to before-tax income). 19 Such was the case with Page’s and Bradenton’s purchase of raw materials in 1974. Had the companies reported their income using inventories, the raw materials inventory would have increased by $240,000 in that year (from an opening balance of $210,000 to a closing balance of $450,000). 20 Put differently, in the absence of inventories, Page and Bradenton took $240,000 in deductions that they would not have had under the inventory method. Clearly, the $240,000 fluctuation was not insignificant relative to the $1,400,000 taxable income that the cash method yielded for that year. 21 We hold that a fluctuation of 17.1% 22 relative to taxable income is substantial enough to trigger section 1.471-1. 23

Taxpayer argues, however, that we should overlook the distortion in 1974 because there was not a great divergence in other years: throughout most of the history of Page and Bradenton, the cash and accrual methods produced substantially the same level of taxable income. Thus, according to Taxpayer, we should not require Taxpayer to use inventories (and the con-eomitant accrual method) merely because of a distortion in one year.

This argument fails on several counts. First, Page and Bradenton are part of a much different organization than they were before 1973. 24 They are now subsidiaries of a massive corporation capable of injecting large sums of money into their businesses and inventories. Thus, much of the data concerning their reporting history before 1973 is less relevant than it would have been prior to the take-over by KnightRidder. Nor does the data after 1974 prove that the inventory fluctuations will not repeat. On the contrary, in 1976 inventories rose by $100,000.

More important, ours is an annual system of accounting. Wilkinson-Beane v. Commissioner, supra, 420 F.2d at 356. A distortion in one year means an absolute loss for the government. The clear reflection of income during a few interim years does not by itself prevent future losses. Id. The substantial distortion in 1974, coupled with the new ownership by Knight-Ridder, are sufficient to support the Com missioner’s determination that inventories posed a threat to the clear reflection of income.

Taxpayer was required to use inventories pursuant to Treas.Reg. § 1.471-1 and, consequently, to adopt the accrual method of accounting pursuant to Treas. Reg. § 1.446-l(c)(2). The Commissioner did not abuse his discretion in requiring Taxpayer to switch to the accrual method for the last months of 1973 and the tax year 1974.

D. Consent of the Commissioner

Of course, section 1.446-l(c)(2) provides an exception to the use of the accrual method if the Commissioner has authorized the use of a different method. Taxpayer argues that the Commissioner is bound by his consent to Page’s and Bradenton’s use of the cash method in earlier tax years. During an audit of the 1969 returns, an I.R.S. agent considered Page’s proposal to switch to the accrual method but did not require such a change. During a subsequent audit of the 1969 and 1970 returns, the Service again failed to object to Page’s use of the cash method. The trial court found that the Commissioner had consented to the use of the method for those years. The court did not find, nor does Taxpayer argue, that the Commissioner had required Page or Bradenton to use the cash method during those years. He had merely consented to its use.

We cannot agree that the Commissioner is thereby barred from changing Taxpayer’s accounting method during a subsequent year in which it becomes apparent that the method does not clearly reflect income. This Circuit and others have long held that the Commissioner cannot so easily “waive for future years the discretionary power which the Congress has granted.” Wood v. Commissioner, 245 F.2d 888, 892 (5th Cir.1957); see Lincoln Electric Co. v. Commissioner, 444 F.2d 491, 493 (6th Cir.1971). Greater experience with the actual effects of the method or a significant change in the nature of the taxpayer’s business may convince the Commissioner that the consent given for earlier years is no longer appropriate. See R.C.A. Corp. v. United States, 664 F.2d 881, 889 (2d Cir.1981), cert. denied, 457 U.S. 1133, 102 S.Ct. 2958, 73 L.Ed.2d 1349 (1982). His decision to consent, for whatever reason, to the use of a method in one year cannot bind his hands in perpetuity. As long as he has not abused his discretion in determining that income is not clearly reflected by the taxpayer’s method, the Commissioner may require a change. As we have seen, there is no abuse of discretion in this case.

II. CLASS LIFE ELECTION

The second major issue in this appeal concerns Taxpayer’s use of the Guideline Class Life System in depreciating press equipment. The Commissioner determined that a number of Taxpayer’s subsidiaries had failed to meet the requirements for electing the system. Taxpayer responded (and the District Court held) that the subsidiaries had “substantially complied” with the requirements. We disagree and reverse the judgment of the court below.

Section 167(m)(l) of the Internal Revenue Code authorizes the Commissioner to establish depreciation class lives which taxpayers may elect to employ. 26 U.S.C. § 167(m)(l). In addition, section 167(m)(3) gives the Commissioner authority to prescribe the time, manner, and conditions of the taxpayer’s election. 26 U.S.C. § 167(m)(3). Pursuant to this statutory authorization, the Commissioner issued regulations establishing the Guideline Class Life System (pre-1971 assets) and Class Life System (post-1970 assets) as well as specific election requirements for each. See Treas.Reg. §§ 1.167(a)-ll, 12; 26 C.F.R. §§ 1.167(a)-ll, 12 (1983). Section 1.167(a)-12(e) provides that for taxable years prior to 1977, a taxpayer desiring to use the Guideline Class Life System must make an election in either of two ways. First, the taxpayer may file Form 5006 with the tax return for that year. Id. at § 1.167(a)-12(e)(3)(i). Alternatively, the election is deemed to be made if the taxpay er provides information sufficient to establish the following:

(a) Each asset guideline class for which the election is intended to apply;

(b) The class life for each such asset guideline class ..;

(c) For each asset guideline class ..., (1) the total unadjusted basis of all qualified property, (2) the aggregate of the reserves for depreciation of all accounts in the asset guideline class, and (3) the aggregate of the salvage value established for all accounts in the asset guideline class;

Id. at § 1.167(a)-l2(e)(3)(ii). 25 In either event, Schedule G of the corporate tax form also provides a box which the taxpayer must check to substantiate its election to use the Guideline Class Life method.

If a taxpayer fails to make a proper election, it is required to depreciate its equipment under the “facts and circumstances” method. 26 Both sides agree that certain subsidiaries of Knight-Ridder failed to comply literally with the election requirements. They failed to file Form 5006, they did not otherwise provide the information alerting the Commissioner to an election, and they failed to check the box in Schedule G.

Taxpayer argues, however, that the subsidiaries “substantially complied” with those requirements of Treas.Reg. § 1.167(a)-12 which go to the essence of the legislative purposes of Section 167(m). Knight-Ridder asserts that the legislative purposes are to simplify and reduce the number of depreciation systems and to reduce the number of disputes between taxpayers and the government. It then reasons that those purposes are met when it calculates its depreciation properly under the Guideline Class Life System, reports the amount accurately on its tax return, and maintains its books and records adequately to allow an audit. Any other election requirements are merely “procedural details,” unnecessary to the legislative purposes.

In support of this argument, Taxpayer cites a number of Tax Court decisions in which taxpayers had failed to meet the literal requirements for an election but the courts held that they had substantially complied. American Air Filter v. Commissioner, 81 T.C. 709 (1983); Tipps v. Commissioner, 74 T.C. 458 (1980); Columbia Iron & Metal v. Commissioner, 61 T.C. 5 (1973); Hewlett-Packard v. Commissioner, 67 T.C. 736 (1977); Sperapani v. Commissioner, 42 T.C. 308 (1964); O’Dowd v. Commissioner, 35 T.C.M. (CCU) 754 (1976), aff'd, 595 F.2d 262 (5th Cir.1979). The courts looked to see whether specific requirements related to the “essence” of the statutory and regulatory scheme. See Tipps, supra, 74 T.C. at 468; Valdes v. Commissioner, 60 T.C. 910, 913 (1973). If not, then literal compliance was not necessary for a valid election. The court in Hewlett-Packard listed a number of factors to be considered:

In ascertaining whether a particular provision of a regulation stating how an election is to be made must be literally complied with, it is necessary to examine the purpose, its relationship to other provisions, the terms of the underlying statute, and the consequences of failure to comply with the provision in question. * * *

67 T.C. at 749, quoting Valdes, supra, 60 T.C. at 913.

Applying that same analysis here, we find that the Taxpayer has overlooked certain essential purposes of I.R.C. § 167(m) and Treas.Reg. § 1.167(a)-12. First, the Congress plainly intended that each taxpayer’s election be binding. An election to come under the class life system for a taxable year may not be changed or revoked. See S.Rep. No. 92-437, 92nd Cong., 1st Sess. 49, H.R.Rep. No. 92-533, 92nd Cong., 1st Sess. 33, U.S.Code Cong. & Ad.News 1971, p. 1825. This policy is furthered by requiring a clear manifestation to the government of taxpayer’s election. Merely calculating depreciation accurately without expressly electing the system would leave room for the taxpayer to argue later that it had never intended to make an election but, rather, was applying the facts and circumstances test. The taxpayer could, therefore, argue that it was not bound to meet other conditions put on the class life system by the Commissioner, see I.R.C. § 167(m)(3), or by Congress itself. 27

A related regulatory goal is that the Commissioner actually know an election has been made. This ultimately serves the policy of minimizing disputes between taxpayers and the Internal Revenue Service. See Treas.Reg. § 1.167(a)-12(a)(l). A clear indication of the taxpayer’s election to employ the class life system removes any ground for dispute over whether the system applies and what the appropriate useful life should be. Moreover, informing the Commissioner of an election helps him to assess whether the system is being abused — for example whether the taxpayer is applying the wrong class life to a particular asset class. Alternatively, if the taxpayer has not informed the Commissioner of an election, the Commissioner can be sure that the system does not apply and can better assess the taxpayer’s analysis of facts and circumstances.

In sum, the elaborate election provisions of section 1.167(a)-12 are designed to inform the Commissioner that an election has been made and which asset classes it affects. We are not saying that each and every requirement is critical, but the goal of clearly informing the Commissioner of an election is essential.

In all but one of the “substantial compliance” eases cited by Knight-Ridder, the taxpayer’s return indicated that an election was being made even though the taxpayer had failed to comply with a minor procedural detail. For example, in Tipps, the taxpayer clearly indicated on its form that it was electing to employ accelerated depreciation pursuant to I.R.C. § 167(k). It specified the election and the property to which it would apply, failing only to supply certain “per unit information.” The court held that “Even if the per-unit information had been filed exactly as [the Commission er] desired, [he] would not have had any additional information identifying or describing the property to which the elections applied or specifying that the elections were being made.” 74 T.C. at 468. Since the taxpayer had effectively reported that it was making an election, the court held that it had substantially complied with the election requirements. Id.

Similarly, in Columbia Iron and Metal, the taxpayer indicated on its return that it was electing to take a charitable deduction; its failure to attach corporate minutes authorizing the charitable contributions did not prevent an effective election. 61 T.C. at 6-7, 9. Finally, in Hewlett-Packard, O’Dowd, and Sperapani, the taxpayer’s failure to meet certain procedural requirements did not prevent substantial compliance with election requirements, where the taxpayer in each case had clearly notified the Commissioner of its intent to make an election. See Hewlett-Packard, supra, 67 T.C. at 747-49 (taxpayer filed information specified by regulations but attached it to wrong part of the return and mailed it to wrong office of Internal Revenue Service); O’Dowd, supra, 35 T.C.M. 754 (election to have business taxed as corporation was filed, but not signed by wife or principal shareholder); Sperapani, supra, 42 T.C. at 329-331 (taxpayer filed timely election to have business taxed as partnership, but failed to attach a supporting statement). 28

By contrast, nothing in the returns of Knight Newspapers and the seven subsidiaries indicated that they were electing the Guideline Class Life System. Their returns were especially misleading, because two other subsidiaries of Knight-Ridder 29 did check the box and did file Form 5006 electing the Guideline Class Life System. Since all of the returns were filed jointly and some showed an election while others did not, the implication was that the latter group had not made the election. This implication was reinforced because four members 30 of the latter group had actually checked the box electing the Class Life ADR System (post-1971 assets), but had not checked the adjacent box for the Guideline Class Life System (pre-1971 assets). The forms clearly seemed to state that these subsidiaries were electing Class Life ADR but not Guideline Class Life. 31

Taxpayer argues, however, that the government does not need an explicit election, because it can discover the errors on audit. That argument does not answer our concerns. The Commissioner needs to know that an election has been made in order to determine whether an audit is necessary in the first place and what its scope should be. The information required by the election provisions of section 1.167(a)-12 directly serves this policy. Therefore, we cannot agree that the requirements are mere procedural details. On the contrary, they go to the essence of the regulatory scheme. In failing to indicate in any manner that it was making an election, Taxpayer failed to comply sub stantially with section 1.167(a)-12. The subsidiaries that did not meet the election requirements are not entitled to employ the Guideline Class Life System for the years in question. 32

III. ADVERTISING REBATE RESERVE

The third major issue in this case concerns whether the advertising rebate reserve used by the Detroit Free Press was an “accounting method” or simply an “improper deduction.” If the reserve was not an accounting method, then the Commissioner would be barred by the statute of limitations from challenging any rebate reserve deductions taken before 1972.

If, on the other hand, the reserve was an accounting method within the meaning of the Internal Revenue Code, then the Commissioner would not be barred. Ceasing to use the reserve in 1972 33 would constitute a change of accounting method; and I.R.C. § 481 would permit the Commissioner to make certain adjustments in taxpayer’s income to prevent a duplication of income or deductions resulting from the change. 34 26 U.S.C. § 481. In particular, if Taxpayer had deducted an item under the former method and the item would again be deducted in a future year under the new method, the Commissioner could force an adjustment to prevent the double deduction. 35 The effect is to permit the Commissioner to correct for deductions made in tax years beyond the normal limitations period. See Graff Chevrolet v. Campbell, 343 F.2d 568, 571-72 (5th Cir.1965). The seeming inconsistency between section 481 and the statute of limitations has been justified on the ground that:

When a taxpayer uses an accounting method which reflects an expense before it is proper to do so or which defers an item of income that should be reported currently, he has not succeeded (and does not purport to have succeeded) in permanently avoiding the reporting of any income; he has impliedly promised to report that income at a later date, when his accounting method, improper though it may be, would require it. Section 481, therefore, does not hold that taxpayer to any income which he has any reason to believe he has avoided, and does not frustrate the policy that men should be able, after a certain time, to be confident that past wrongs are set at rest.

Id. at 572, quoting Note, Problems Arising From Changes in Tax-Accounting Methods, 73 Harv.L.Rev. 1564, 1576-77 (1960).

Section 481 applies only if there has been a change in accounting method, however. Thus, the dispute between Taxpayer and Commissioner boils down to a single basic issue — whether a rebate reserve is an accounting method. KnightRidder contends that the reserve was merely an improper deduction. We disagree and hold that the reserve was an accounting method; therefore, the Commissioner was entitled to make the appropriate section 481 adjustments..

The Internal Revenue Code does not specifically define an “accounting method.” However, the implementing regulations provide:

A change in accounting method includes a change in the overall plan of accounting for gross income or deductions or a change in the treatment of a material item used in such overall plan____

A material item is any item which involves the proper time for the inclusion of the item in income or the taking of a deduction.

Treas.Reg. § 1.446-l(e)(2)(ii)(a); 26 C.F.R. § 1.446-l(e)(2)(ii)(a) (1983) (emphasis added); see also Treas.Reg. § 1.481-l(a)(l); 26 C.F.R. § 1.481(a)(1) (1983). 36 The essential characteristic of a “material item” is that it determines the timing of income or deductions. See Bittker, supra note 1, ¶ 105.6.2, at 124-25. 37 For example, there is a change in the treatment of a material item when a taxpayer shifts from deducting dividends when paid to deducting them in the year they are declared. See Commissioner v. O. Liquidating Corp., 292 F.2d 225 (3rd Cir.1961), cert. denied, 368 U.S. 898, 82 S.Ct. 177, 7 L.Ed.2d 94 (1961).

The change of a single item may trigger section 481 adjustments. Graff Chevrolet v. Commissioner, supra, 343 F.2d at 570-71.

In the case at bar, the Free Press’s rebate reserve was an item which affected the timing of a deduction. There is no question that a deduction would be proper in the year that rebates were actually paid to advertisers. The reserve method merely accelerated the taking of‘that deduction to the time when the amounts were originally added to the reserve. Later, when the rebates were paid, the reserve was reduced accordingly. The reserve did not determine whether or not a rebate would be deducted, but when that deduction would occur.

Taxpayer, however, refers to Schuster’s Express v. Commissioner, 66 T.C. 588 (1976), aff'd per curiam, 562 F.2d 39 (2d Cir.1977), in which the Tax Court faced a similar reserve account and concluded that it was not an accounting method within the meaning of section 481. The taxpayer had maintained a reserve for insurance expenses and deducted amounts added to the reserve. When the taxpayer abandoned the method, the Commissioner attempted to require section 481 adjustments. The Tax Court held that section 481 did not apply because the reserve was not a material item. The court argued that the reserve did not involve the timing of deductions. 66 T.C. at 596-97.

The court’s reasoning is not entirely clear. 38 However, the main concern seems to be that the taxpayer had made larger contributions to the reserve than were necessary. There was an excess balance that had never been — and might never be — offset by actual insurance expenditures. The court felt that this excess did not involve the mere timing of an otherwise proper deduction, because the deduction might never match up with a proper expense. The taxpayer could avoid income for all time rather than simply deferring it. Id. 39

Knight-Ridder would raise the same objection in our case, for the balance of the rebate reserve account never decreased after 1971. See supra at 786. We do not agree, however, that this undermines the argument that the rebate reserve involved a question of timing. To understand this, it is helpful to realize that though we talk about the timing of deductions, the basic issue is whether income is reflected and taxed. The reserve method determines when income will be taxed. When deductions are taken early (at the time money is added to the reserve), an equal amount of income is obviously not taxed. That income is taxed, however, at the later time when deductions would have been taken under a different system (i.e., at the time rebates are paid, the absence of deductions means that an equal amount of income is taxed). Most important, at the time the company ceases to use the reserve (e.g., when the company closes out its business), any remaining balance in the reserve must be included in taxable income. See Levelland Savings and Loan Assoc. v. United States, 421 F.2d 243, 246 (5th Cir.1970); Bird Management v. Commissioner, 48 T.C. 586, 595 (1967); Hawes Corp. v. Commissioner, 44 T.C. 705, 707 (1965). 40 Thus, no income is avoided altogether. Any excess deductions in earlier years are offset by an equal amount of taxable income in the final day. The question becomes one of timing, whether income is taxed when the amounts are added to the reserve or when the reserve is abandoned at the Day of Armageddon. 41 Thus, Taxpayer’s rebate reserve method was a “material item” and a method of accounting within section 481. Compare Rev.Rul. 81-93 (insurance expense reserve a method of accounting).

We note, incidentally, that the 1954 Congress intended to treat a similar statutory “reserve for estimated expenses” as an accounting method. Section 462 of the 1954 Code authorized taxpayers to keep reserves for expenses that could be estimated with “reasonable accuracy.” See S.Rep. No. 1622, supra, 1954 U.S.Code Cong. & Ad.News, at 4945. A taxpayer could establish such a reserve to cover the cost of quantity discounts, id. at 4946, and could deduct anticipated expenses in the year they were added to the reserve.

The legislative history to the accounting provisions (sections 446 and 481 42 ), indicates that the adoption or abandonment of a section 462 reserve would constitute a change of accounting method:

A change in the method of accounting ... includes a change from an accrual method without estimating expenses to an accrual method with estimated expenses, or vice versa____

Id. at 4940; H.R.Rep. No. 1337, supra, 1954 U.S.Code Cong. & Ad.News, at 4297.

Congress repealed section 462 in 1955 because the deduction of estimated expenses threatened a serious revenue loss during the transitional period. See B. Bittker and L. Stone, Federal Income Taxation 1057 (1980); Chirelstein, supra note 1, at 221. Nevertheless, the legislature’s words in the 1954 reports give an indication of Congress’s view of reserves in general and further substantiate our conclusion that the rebate reserve is a method of accounting. The Commissioner did not err in requiring Taxpayer to make adjustments when it changed that method.

CONCLUSION

In sum, we hold that

(1) the Commissioner did not abuse his discretion in requiring Page and Bradenton to adopt the accrual method of accounting;

(2) Knight Newspapers and its seven subsidiaries did not substantially comply with the requirements for electing Guideline Class Life depreciation; and

(3) the Detroit Free Press was required to make section 481 adjustments when it ceased to use its rebate reserve method.

Accordingly, this case is

REVERSED and REMANDED.

1 . See 4 B. Bittker, Federal Taxation of Income, Estates and Gifts ¶ 105.3.1, at 47-48 (1981) [hereinafter cited as Bittker]; 2 Mertens, The Law of Federal Income Taxation § 12.60 (1982) [hereinafter cited as Mertens]; M. Chirelstein, Federal Income Taxation 214 (1980) [hereinafter cited as Chirelstein ].

2 . _ ,. , . , . . Thf/istrict courl dld not fmd that the Service would have prevented such a change if voluntarlly undertaken by Page.

3 . The chart was not technically introduced as evidence; however, it was used at trial as a graphic display to clarify the testimony of Mr. Pruitt. Therefore, we will refer to the chart in describing his conclusions. Both the government and Taxpayer refer to the chart in their briefs, and Taxpayer appended a copy to its brief.

Mr. Pruitt’s calculations of income before taxes for Page and Bradenton for the years 1962-1979 were as follows:

Year Ending August 31

1962 1963 1964

Acc,rual 750,000 760,000 970,000

770,000 720,000 895,000

Difference 20,000 40,000 75,000

1965 1966 1967

Accrual 1,260,000 1,240,000 1,365,000

Cash g 1,175,000

D;fference 75 000 65i000 10,000

1968 1969 1970

Accruai 1,730,000 1,995,000 1,970,000

Cash 1,625,000 1,735,000 2,120,000

Difference 105,000 260,000 150,000

1971 1972 1973

Accrual 2,450,000 2,680,000 3,165,000

Cash 2,395,000 2,790,000 3,000,000

Difference 55,000 110,000 165,000

197 A

~QQ

Cash i,720,000

Difference 330,000

Year Ending December 31

1974 1975 1976

Accrual 1,900,000 2,125,000 2,680,000

Cash 1,400,000 2,165,000 2,665,000

Difference 500,000 40,000 15,000

1977 1978 1979

Accrual 3,095,000 3,740,000 3,675,000

Cash 3,125,000 3,605,000 3,480,000

Difference 30,000 135,000 195,000

The calculations shift to tax years ending on December 31 because Page and Bradenton changed to that tax year in 1974 in order to conform with the returns of their new parent, Knight-Ridder. This would, of course, have produced a gap period between their 1973 tax year (ending in August, 1973) and their 1974 tax year (beginning in January, 1974). They corrected for this problem, however, by filing an extra set of returns reporting income and deductions during the gap period ending December 31, 1973. Mr. Pruitt’s chart does not show the specific figures from these extra returns, but it does show what a return for the full tax year ending August 31, 1974, would have reported.

4 . Beacon Journal Publishing Company; Tallahassee Democrat, Inc.; Macon Telegraph Publishing Company; Boca Raton News, Inc.; Page (for 1973 only); Bradenton (for 1973 only); and Lexington Herald Company (for 1973 only).

5 . The Beacon Journal, Tallahassee Democrat, Macon Telegraph, and Boca Raton News all elected the Class Life ADR System (post-1971 assets) but not the Guideline Class Life System (pre-1971 assets). The Detroit Free Press and Knight Publishing Company checked both boxes, electing both systems, as would be necessary if they wished to use class life depreciation for assets put into service before 1971 as well as those put into service after that date. Page, Bradenton, and the Lexington Herald checked neither box.

6 . Taxpayer has not disputed that 18 years is the proper useful life under the "facts and circumstances” test. It did argue at trial, however, that the Commissioner should have given “effect to the prior proper use of the class system under Rev.Proc. 62-21 [a precursor to the Guideline Class Life System] in determining the remaining useful life by prorating in proportion to the useful life remaining under the prior system.” Appellee’s Brief at 14, note *. See infra note 32.

7 . 26 U.S.C. § 167(m).

8 . Treas.Reg. § 1.167(a)-12.

9 . The trial court also resolved an issue concerning the proper depreciable basis for certain real property acquired by Taxpayer. Neither side has appealed from this decision, however.

10 . The Commissioner may authorize the taxpayer to use a different accounting method "if, in the opinion of the Commissioner, income is clearly reflected by the use of such method.” 26 C.F.R. § 144-l(c)(2)(ii). See infra at 793.

11 . The account books reflect this deferral by adding the costs to the inventory account in the year in which they are incurred. They are then removed from the account in the year that the merchandise is sold. The cost of goods sold in any given tax year is determined by the following equation:

Cost of goods sold = (opening inventory + purchases during the year) - closing inventory.

Bittker, supra note 1, f 105.4.1, at 77.

12 . As the Second Circuit stated in Caldwell v. Commissioner, supra:

The use of inventories in computing income results in stating the expenses of a year’s operations in terms of the cost of the goods actually sold during that year. Thus the profit from these operations will be stated accurately only if the income from all sales made during the year is taken into consideration. This requires use of the accrual method of determining income, since the cash receipts method obviously does not reflect the actual sales made during the year where ... sales ... [are] made on credit.

202 F.2d at 112, 114.

13 . The Tax Court has held that even though the Taxpayer is already using inventories, the cash method may reflect income accurately if the inventories are very small. See Drazen v. Commissioner, 34 T.C. 1070, 1079 (1960); Brookshire v. Commissioner, 31 T.C. 1157, 1166 (1959), aff'd 273 F.2d 638 (4th Cir.1960), cert. denied, 363 U.S. 827, 80 S.Ct. 1597, 4 L.Ed.2d 1523 (1960).

Other writers, however, have suggested that whenever a taxpayer uses inventories, regardless of their size, it must use the accrual method. See Herberger v. Commissioner, 195 F.2d 293 (9th Cir.1952), cert. denied, 344 U.S. 820, 73 S.Ct. 17, 97 L.Ed. 639 (1952); Bittker, supra note 1, ¶ 105.4.1, at 83 ("the better view today is that cases [like Brookshire ] have continued vitality only in holding that taxpayers required to use inventories can use the cash method for such noninventory items as investment income and interest expense"). This view is certainly in keeping with the primary goal of inventories: matching cost and revenue. We will leave the issue for another case, however. As we discuss below, the inventory fluctuations were substantial in this case.

14 . The figures, broken down by firm and year, are as follows:

Page's Year Ending Cost of Newsprint and Ink Cash Receipts Newsprint and Ink as Percentage of Receipts

12/31/73 $ 453,564 $ 2,202,294 20.5%

12/31/74 1,309,401 7,686,729 17.0%

Bradenton’s Year Ending

12/31/73 $ 61,168 $ 585,760 10.4%

12/31/74 632,295 3,460,372 18.3%

Total $2,456,428 $13,935,155 17.6%

15 . The 15.1 figure is an average of the amounts for the two years at issue in that case:

Year Cost of Caskets Purchased Cash Basis Receipts Caskets as a % of Receipts

1963 $20,052 $128,899 15.4%

1965 17,883 121,575 14.7%

16 . In a number of other cases, the courts and the Internal Revenue Service have required mixed service and merchandise businesses to use inventories pursuant to § 1.471-1. See, e.g., McGrath & Son v. United States, 549 F.Supp. 491 (S.D.N.Y.1982) (funeral service); Rev.Rul. 74-279 (optometrist who maintains a supply of eyeglass frames); Rev.Rul. 73-485 (company that fits handicapped persons with artificial limbs and orthopedic braces).

17 . We recognize that books may serve other purposes, e.g., filling empty spaces on one’s bookshelf and lending a scholarly atmosphere to the den. Newspapers, however, can also serve residuary purposes, e.g., doubling as kindling for a holiday fire.

18 . In a variation on this theme, some cases have held that when a taxpayer is already using inventories, the Commissioner may not require a switch to accrual accounting if the inventories are insignificant. See Drazen v. Commissioner, supra note 13; Brookshire v. Commissioner, supra note 13. As discussed above, however, these rulings are suspect because they interfere with the primary goal of inventories. See supra note 13.

19 . We need not decide how large actual inventories must be to invoke the regulation if the fluctuations are not as dramatic as those in the case at bar. For example, if the inventory stood at one level for several years, it would by definition have no effect on taxable income during those years. In that case, the inventory level might have to be more substantial than a fluctuation must be to justify the application of § 1.471-1.

20 . Furthermore, in the short tax period ending December 31, 1973, see supra note 3, the inventories increased by $50,000, from $160,000 to $210,000.

21 . The $50,000 increase during the final months of 1973, see supra note 20, was also arguably substantial when compared to the $440,000 taxable income for that period. However, the ratio is smaller than in 1974, and the substantiality question is complicated by an abnormally short tax period. In any event, we need not decide the issue as to 1973, since the fluctuation in 1974 was clearly substantial.

22 . 240,000 + 1,400,000 = .171, or 17.1%.

23 . This might be a different case if the before-tax income were extraordinarily low in 1974 for reasons other than the current deduction of inventory items. For instance, if Taxpayer normally reported before-tax income of $100,000 to $250,000, but in one year reported only $12 because of high labor costs, we would be loath to require a switch to inventory accounting merely because inventories in that year were $6 (as they had been in previous years). $6 is certainly substantial relative to $12, but the dramatically low before-tax income causes us to be more careful.

That, however, is not our case. Before-tax income in 1974 for Page and Bradenton was $1,400,000 as reported on the cash method. From 1962 to 1973, their income averaged $1,647,000 per year. That is not a dramatic difference. Indeed, even compared to the period 1971-1973 (in which income averaged $2,728,000), the drop to $1.4 million is not the kind of extraordinary dip we are talking about here.

24 . The trial court found that the "nature of their business” has not changed materially. Record, Vol. II, at 221. That is true in terms of their individual operations. However, the record undeniably supports the view that their ownership changed dramatically in 1973, and that investments by the new owner can severely distort income (as occurred in 1974 with a total distortion of $500,000).

25 . These requirements are quoted from a modified version of the regulation, which was made retroactive to the taxable years in question. See T.D. 7517, 1977-2 Cum.Bull. 68, 69. The original version, which was in force during the years in question, contained the same requirements in all relevant respects and actually contained some additional requirements which were deleted by the modification. The original version provided in pertinent part:

(3) Information required. The election to apply this section for a taxable year must specify:

(i) That the taxpayer makes such election and consents to, and agrees to apply, all the provisions of this section;

(ii) Each asset guideline class for which the taxpayer makes the election;

(iii) The class life for each asset guideline class for which the taxpayer makes an election and whether the class life is determined under paragraph (b)(1) or (2) of this section;

(iv) For each asset guideline class, as of the end of the taxable year of election, (a) the total unadjusted basis of all qualified property, (b) the aggregate of the reserves for depreciation of all accounts in the asset guideline class, and (c) the aggregate of the salvage value established for all accounts in the asset guideline class;

(v) Whether the taxpayer elects to apply Revenue Procedure 72-10 on a composite asset guideline class basis in accordance with paragraph (b)(3)(iii) of this section; and

(vi) Such other information as may reasonably be required, including all or any part of the type of information specified in paragraph (f)(4) of § 1.167(1) — 11.

Treas.Reg. § 1.167(a)-12(e), T.D. 7278, 1973-2 Cum.Bull. 54, 63-64.

26 . This is true even though the taxpayer had previously been permitted to depreciate its equipment at an accelerated rate pursuant to Rev.Proc. 62-61. Rev.Proc. 62-61, a predecessor to section 1.167(a)-12, contained many of the same class lives but lacked the elaborate election requirements. When the new regulation came into force, taxpayers who had depreciated pre-1971 assets under Rev.Proc. 62-61, were required to make an election to apply the new regulation. Otherwise, they were thrown back into the "facts and circumstances” method.

27 . This is not to say that Knight-Ridder has failed to abide by any requirements in this case. We merely recognize that the policy supporting a binding election is undermined to the extent that this precedent permits future taxpayers to make an election without a clear manifestation of that decision.

28 . American Air Filter, supra, is the only Tax Court decision cited by Knight-Ridder in which the taxpayer did not make clear its intent to have a binding election. As to that case, we can only say that it does not involve an election to employ accelerated depreciation, and, in any event, it is not binding on us.

Taxpayer also cites an Internal Revenue Service private letter ruling concerning substantial compliance with Class Life ADR requirements. See National Office Technical Advice Memorandum 8404008 (Sept. 29, 1983). The ruling indicates that a taxpayer need not make an express election. For the reasons stated above, we believe that conclusion is wrong. Indeed, the government informs us that it has taken the ruling under review.

Taxpayer does not contend that it relied to its detriment on the ruling. The Service issued the ruling in 1983, at least eight years after the preparation of the tax returns involved in this case.

29 . Knight Publishing and Detroit Free Press.

30 . Beacon Journal, Tallahassee Democrat, Macon Telegraph, and Boca Raton News. See supra note 5.

31 . Taxpayer has not contended that checking the box for Class Life ADR would alert the Commissioner to an election of the Guideline Class Life System as well. Nor has Taxpayer maintained that its accountants checked the ADR box by mistake, intending to mean the other box.

32 . Because the issue has not been raised on appeal, we do not decide whether the Commissioner applied the facts and circumstances test properly in calculating the remaining useful life for Taxpayer’s press equipment.

33 . The Commissioner chose 1972 as the year that Taxpayer “ceased to use the reserve" because reserve deductions taken in 1972 and following years were nullified by I.R.S. action. Taxpayer has not argued on appeal that the Commissioner should have chosen a different date.

34 . A change of accounting method comes within section 481 regardless of whether the Commissioner or the taxpayer has initiated the change. See Grogan v. United States, 475 F.2d 15, 16-18 (5th Cir.1973).

35 . The adjustment would not necessarily occur only in the year of the accounting change. Section 481 permits some flexibility. For example, if the adjustment were large enough, it could be spread over three consecutive years. Id.

36 . Treas.Reg. § 1.481 — 1(a)(1) provides the basic discussion of I.R.C. § 481 and the meaning of a “change in accounting method." The regulation incorporates by reference the definition of "accounting method” used in Treas.Reg. § 1.446(e)(2)(ii)(a). Section 1.481 — 1(a)(1) provides:

A change in method of accounting to which section 481 applies includes a change in the over-all method of accounting for gross income or deductions, or a change in the treatment of a material item. For rules relating to changes in methods of accounting, see section 446(e) and paragraph (e) of § 1.446-1.

26 C.F.R. § 1.481 — 1(a)(1).

37 . To be sure, certain adjustments which affect timing are not treated as changes of accounting methods. Treasury Regulations § 1.446-l(e)(2)(ii)(b) specifies that there is no change of accounting method when a taxpayer (1) makes adjustments in additions to bad debt reserves or (2) makes changes in the useful life of deprecia-ble assets. 26 C.F.R. § 1.446-l(e)(2)(ii)(b). Those two cases involve minor adjustments in a method itself, however, not the wholesale abandonment of a method which affects timing.

38 . The holding of the court is muddied somewhat by the unusual burden of proof in the case.

After concluding at trial that [the Commissioner’s] intended reliance on section 481 constituted a new theory not within the breadth of the statutory notice of deficiency, [the court] ruled that [the Commissioner] must bear the burden of proof with respect to the facts of the case.

66 T.C. at 593. Thus, the Commissioner had the burden of proving that "there was any procedure or intention to restore the excessive deductions to income in future years so as to properly reflect petitioner’s total lifetime income.” Id. at 596. By contrast, in the case at bar, the section 481 argument is not a "new theory,” and the burden remains on Knight-Ridder to prove that its rebate reserve would not have adjusted for excess deductions. Taxpayer did not meet that burden. See infra at 799-800 and note 41.

39 . The court also argued that even if the reserve were an accounting method, section 481 would not apply because distortions in the taxpayer's income were not caused “solely by the change in the method of accounting.” Id. at 598. As we understand the argument, it basically restates the first point: the distortions were not caused solely by the change because even if the taxpayer had continued to use the reserve, the existence of an excess balance would show that the taxpayer had taken improper deductions.

Because the two arguments merge into one, we will not treat them separately.

40 . The cited cases describe the treatment of bad debt reserves, but the same rules would apply to a rebate reserve.

The government has argued, incidentally, that Knight-Ridder ceased to need the reserve method in 1972. Therefore, the remaining balance should have been taxed in that year irrespective of the Commissioner’s powers under section 481. The argument has some intuitive appeal but raises tricky questions about when and whether Taxpayer ceased to need the reserve. Given our resolution of the case on the section 481 ground, we need not address this alternative argument by the government.

41 . In the meantime, as long as the reserve remains in use, it contains the potential for correcting itself, reducing its balance to zero. The district court found that "additions were made to the reserve on estimated basis for anticipated liability.” Record, Vol. II, at 255. It is conceivable that in a future year, the Free Press, taking into account the existing balance in the reserve, would have estimated perfectly or underestimated the potential liability. Consequently, it would have used the entire balance in paying rebates.

By changing the method of accounting, Taxpayer artificially interrupted the process of annual adjustments and ended the possibility of reducing the reserve balance to zero.

42 . The quoted material appears in a discussion of I.R.C. § 446. As we have seen, the definition of accounting method applicable to section 446 is likewise applicable to section 481. See supra note 36.