What isn't a change in method of accounting?

Introduction

What Is a Method of Accounting?

A method of accounting is a rule that governs when an item of income or expense will be taken into account for federal income tax purposes. The paradigmatic example in the regulations is that a change from consistently claiming current deductions in the year assets are purchased to depreciating the cost of those assets is a change in method of accounting. Treas. Reg. [sections] 1.446-1(e)(2)(ii)(b).

Why Do Different Methods of Accounting Exist?

Because it is not always clear as a theoretical matter when income is earned or when expenses are incurred, it is helpful as an administrative matter to have each taxpayer select clear rules that its employees can follow instead of engaging in daily debates about whether the day's events increased or decreased the wealth of the taxpayer. The decision to expense or depreciate a long-lived asset is too simple to illustrate the point. Here is a harder example:

For $120 million, you sell a customer the right to buy up to 120 million units of your product at today's price during the next 10 years. Will you --

* Include the entire amount in this year's income?

* Include the entire amount in Year 10 when you know whether you won or lost the bet?

* Include $1 million a month for the next 120 months?

* Include $1 per unit as units are actually delivered?

* Revalue the contract every time the price of your product rises or falls, and recognize loss or gain as the buyer's rights become more or less valuable?

You probably won't choose that last treatment, but whatever you choose will become your method of accounting for this $120 million contract.

Why Does It Matter If a Treatment Is Called a "Method of Accounting"?

First, section 446(b) of the Internal Revenue Code grants the IRS broad discretion to decide whether or not a "method of accounting" clearly reflects the taxpayer's income, and to force the taxpayer to change methods that do not clearly reflect their income. Although there are exceptions,(1) the courts rarely hold that the Internal Revenue Service abused its discretion in this area, or in the closely-related area of inventory accounting methods under section 471(a). If anything, the courts seem to give the IRS even more leeway with respect to inventory methods.

Second, section 446(e) requires the taxpayer to obtain the Commissioner's consent before changing any method of accounting. The trouble is that a large company probably changes numerous small accounting methods every time it hires a new cost accountant or updates its accounting software. The company does not want to stifle efforts to improve its cost accounting. But any "improvement" in accounting is likely to be a method of accounting for which the company should have obtained the Commissioner's permission before extending the improvement to its tax accounting. And revenue agents have a nasty habit of deciding whether an improvement is a change in method based on whether it gives rise to a positive or negative adjustment to income, and whether it helps or hurts the taxpayer to go back and fix the problem for prior years.

Third, section 481 and the applicable revenue procedures generally allow a taxpayer several years to take into the income the effect of an unfavorable change in accounting method. This "spread period" can significantly ease the pain of the change. On the other hand, the amount of the adjustment under section 481 is special in a way that often hurts taxpayers -- it takes into account closed tax years back to at least 1954, and back to the taxpayer's first use of the method if the taxpayer wants a spread period.

Why Is the Commissioner's Permission Required to Change a Method of Accounting?

The main justification for consent is to allow the Commissioner to prevent omission or duplication of income by making an appropriate transition adjustment when a method is changed. Going back to the prepayment example, the Commissioner wants to make sure that even if you change methods after six years, you eventually report all $120 million of income. This is why the transition adjustment that the Commissioner is permitted to make under section 481 takes into account closed years.

Consent may also promote consistency and prevent taxpayers from flip-flopping among permissible methods that become more or less advantageous over time. The IRS gets particularly excited when it thinks the taxpayer is getting the benefit of hindsight.

Finally, consent gives the Commissioner notice of changes and an opportunity to assess the propriety of the new method before it is adopted, instead of during audit. Taxpayers, however, adopt their initial methods of accounting without oversight, so this justification is not terribly convincing.

When Isn't Permission Required?

This article lists 10 arguments that might work -- and some of the cases you can cite(2) -- in different situations. Taxpayers should bear in mind that most of these arguments can also be used to show that a treatment is not a method that triggers the Commissioner's clear reflection authority. But it is easier to focus on the permission issue because a couple of extra arguments are available there.

Not Just Timing

Accounting Methods Are "Just Timing"

Treas. Reg. [sections] 1.446-1(e)(2)(ii)(a) provides that a change in method of accounting includes a change in the overall plan of accounting for income or deductions or a change in the treatment of any material item used in that overall plan that involves the proper time for including an item in income or taking a deduction. Rev. Proc. 92-20, 1992-1 C.B. 685, is the IRS's manual for making changes in methods of accounting. It provides that in determining whether a practice involves the proper time for including an item in income or taking a deduction, the relevant question is generally whether the practice permanently changes the amount of taxable income over the taxpayer's lifetime. If the practice does not permanently affect the taxpayer's lifetime taxable income, but does change the taxable year in which taxable income is reported, it involves timing and is considered a method of accounting. 1992-1 C.B. at 688. See also Rev. Proc. 91-31, 1991-1 C.B. 566, 567.

Therefore, You Do Not Need Permission If its "Not Just Timing"

Under Treas. Reg. [sections] 1.446-1(e)(2)(ii)(b), deducting as interest or salary expense a payment to an officer that was really a dividend is not a method because dividends paid are never deductible. Deducting as a business expense an amount that was really a personal expense is not a method because personal expenses are never deductible. Other obvious examples of items that are "not just timing" are mistakenly including in income an item that is exempt from tax under section 103 or section 115.

The regulations also say that errors in the computation of the foreign tax credit, net operating loss, percentage depletion, or investment credit are not methods of accounting. Although the IRS may argue that these four items were included in the regulations as examples of "mathematical or posting errors in the computation of tax liability," that argument does not explain the coincidence that all four are permanent items. The Court of Federal Claims favored this reading in Travelers Insurance Co. v. United States, 28 Fed. Cl. 602, 614 (Fed. Cl. 1993) (the foreign tax credit computation takes place only after taxable income has been computed and all timing matters have been resolved).

If the practice affects whether another tax credit is available, the permanent loss of that credit is not just timing. Although the IRS may argue that credits never affect lifetime taxable income, the rationale of Travelers should extend to tax credits not mentioned by the regulations.

If the practice involves the proper year to take a loss or deduction, the Code and regulations provide that the loss or deduction must be taken in the proper year. See Treas. Reg. [sections] 1.461-1(a)(3) (the expenses, liabilities, or loss of one year generally cannot be used to reduce the income of a subsequent year); I.R.C. [sections] 165(a) and Treas. Reg. [sections] 1.165-1(d) (losses are deductible only in the year they are sustained). This is not just timing because the courts deny permanently taxpayers' claims for losses if the claims are not made for the year the loss was actually sustained. See, e.g., L.B. Maytag v. Commissioner, 32 T.C. 270 (1959) (taxpayer claimed loss deduction upon abandonment of last of five oil and gas leases that it considered a single asset; the court held taxpayer should have deducted losses as each lease was abandoned and permanently denied the deductions for the other four leases).

If the practice affects whether income is reported as ordinary or capital gain or loss, the rate differential for gains and possible expiration of capital losses is not just timing. If the capital loss limitation does not apply (e.g., because this is a section 1231 asset) and the rate differential for the taxpayer and year at issue is expressed as a capital gains rate (instead of an exclusion of part of the capital gain), the IRS might argue that the differential does not affect lifetime taxable income. It would be interesting to see how a court would react to this sort of permanent difference in tax.(3)

Characterization 

Income or Recovery of Basis?

In Underhill v. Commissioner, 45 T.C. 489 (1966), the taxpayer had for many years purchased notes at a discount and reported any principal payments received on the notes as part recovery of basis and part taxable income (recovery of the discount). In 1961, the taxpayer stopped reporting any taxable income until his basis was fully recovered, relying on Phillips v. Frank, 295 F.2d 629 (9th Cir. 1961), which held that where the taxpayer owns a "speculative" note and may never recover its basis, principal payments are reported as return of basis until the taxpayer's basis is recovered.

Because Mr. Underhill had not asked for consent, the IRS asked the Tax Court to deny his change in treatment of principal payments received on loans purchased before 1961. But the Tax Court refused, saying:

The issue before us is the extent to which payments received by petitioner are taxable or nontaxable -- i.e., the character of the payment -- not the proper method or time of reporting an item the character of which is not in question. Petitioner should no more be precluded from reporting his payments on the correct basis than a taxpayer who has previously been reporting nontaxable income as taxable income would be required to continue to do so because of his prior error.

45 T.C. at 496. Judge Tannenwald was not willing to deny the taxpayer the substantive legal result in Phillips v. Frank just because the taxpayer had not asked for the Commissioner's permission.

The judge also said that the proper tax accounting was not a matter of the taxpayer's choice; rather, it followed from the substantive characterization of the notes as either speculative or not. This "element of choice" rationale may seem compelling, but it is very hard to square with the regulations under section 446. After all, there is no element of choice in whether a taxpayer expenses or depreciates the cost of an asset (the regulations' paradigm of a change in method). Nevertheless, the element of choice can be seen in other cases, such as Standard Oil Co. (Indiana) v. Commissioner, 77 T.C. 349 (1981), acq. on other issue, 1989-2 C.B. 1 (discussed below under the heading "Error").

Lease or Sale?

In Coulter Electronics, Inc., 59 T.C.M. (CCH) 350 (1990), aff'd in unpublished opinion, 943 F.2d 1318 (11th Cir. 1991), a manufacturer of medical equipment offered its products to hospitals and laboratories for sale or lease. Most of the leases entered into by the taxpayer were assigned to a commercial bank. On its tax returns for 1974-1978, the taxpayer reported its leases to customers as sales of equipment with financing, and reported the assignments to the bank as sales of the customers' notes. When the IRS audited those years, however, the taxpayer notified the agent that its leases to customers and assignments to the bank should both have been reported as leases, not sales, for tax purposes.

The agent conceded that the customer contracts should be treated as leases for tax purposes, but maintained that the assignments to the bank should be treated as sales. The Tax Court agreed with the taxpayer that its relationship with the bank was more like a lease or secured financing than a sale. Government counsel argued that, even so, the taxpayer should not be permitted to change its method of accounting for the assignments because it did not ask for permission. The Tax Court disagreed, citing Underhill and Standard Oil: "Although there is a timing consequence to the outcome of the characterization, it is automatically determined by the characterization and no change of accounting within the meaning of section 446 is involved." 59 T.C.M. (CCH) at 365. Again, the court was unwilling to deny the taxpayer the proper substantive legal result merely because it had not asked for permission.

The IRS held likewise in Letter Ruling No. 9307002 (October 5, 1992). When a taxpayer argues that the lease or sale issue is not a method of accounting, however, the IRS has sometimes cited Cubic Corp. v. United States, 74-2 U.S. Tax Cas. (CCH) [paragraph] 9667 (S.D. Cal. 1974), aff'd on other grounds, 541 F.2d 829 (9th Cir. 1976). But in Cubic the district court held that the taxpayer had properly characterized its leases on its original returns, and it rejected as a substantive matter the taxpayer's request to recharacterize the leases as sales for tax purposes. The court was particularly impressed by the parties' having structured the transactions as leases because of concerns about the legal enforceability of financed sales to Cubic's customers.

Having rejected the taxpayer's argument on the merits, the court added that "[p]laintiff cannot prevail herein for another reason." Having reported the transactions as leases on its original returns, the court said, the taxpayer could not change that accounting method without the Commissioner's consent. 74-2 U.S. Tax Cas. (CCH) at 85,149-50. In light of its holding on the merits, the district court's statements about the change in accounting method were clearly dicta.

The weight of the district court's statements were weakened further when the Ninth Circuit affirmed the result in Cubic in a per curiam opinion that does not even mention the change-in-method issue. The appellate court was impressed that the parties had structured the transactions as leases to avoid the legal restrictions on the buyers' ability to purchase Cubic's goods on credit. In fact, a look at the briefs reveals that the government decided to abandon the change-in-method argument on appeal. In footnote 11 of its brief, government counsel conceded to the appellate court that change in method was not the right way to argue a characterization case, and the district court's statements about change in method were dicta anyway.(4)

With Cubic cut down to size, the biggest problem in this area is a revenue procedure that the IRS published last year. Rev. Proc. 95-38, 1995-34 I.R.B. 25, was published to resolve many of the disputes that the IRS was having with rent-to-own dealers about whether their arrangements with customers were leases or sales. Because a dealer that rents appliances or furniture to consumers with an option to purchase might expect half of the customers to eventually buy the leased property and the other half to eventually buy the leased property to the dealer, these cases were particularly hard to characterize as either sales or leases. Rev. Proc. 95-38 notes that, as a legal matter, each transaction should be characterized as a sale or a lease based on all of the facts and circumstances of that transaction. As a practical alternative, the revenue procedure allows dealers to avoid the expense and uncertainty of trying to characterize each transaction by treating all contracts that contain certain terms as leases (but without the benefit of section 1231 if the property is sold at a gain).

Section 7 of Rev. Proc. 95-38 states that the treatment offered is a method of accounting, and dealers who wish to change from their current treatment (e.g., as sale transactions) to the treatment offered by 95-38 must seek the Commissioner's permission. 1995-34 I.R.B. at 26. This should not be read as inconsistent with Coulter Electronics, which teaches that the legal characterization of a contract under all the facts and circumstances is not a method. Rev. Proc. 95-38 is not a legal characterization of the contracts under all the facts and circumstances; it is a very different set of characterization standards that the IRS National Office is offering to taxpayers who come in and ask to use it. By changing the set of standards it uses to evaluate whether a contract is a lease or a sale, taxpayer may well be changing methods of accounting when it adopts the bright line rules of Rev. Proc. 95-38. That does not mean that mischaracterizing a contract under the common law standards for evaluating a contract is a method of accounting.

Inventory or Depreciable Asset?

* Spare Computer Parts. In Diebold, Inc. v. United States, 16 Cl. Ct. 193 (1989), aff'd, 891 F.2d 1579 (Fed. Cir. 1989), cert. denied, 498 U.S. 823 (1989), a manufacturer of automatic teller machines (ATMs) sold banks service contracts under which the taxpayer agreed to keep the ATMs running. In order to minimize downtime, Diebold designed the ATMs around "modules" that could be replaced quickly to restore the ATM to service without searching for malfunctioning part. Diebold manufactured and maintained a pool of replacement modules for use by its service fleet.

On its original returns, Diebold treated the pool of replacement modules as inventory. Diebold later filed amended returns claiming depreciation and investment tax credit (ITC) with respect to the replacement modules. Diebold claimed the modules were depreciable capital assets, not inventory, because they were not held for sale to customers. The Claims Court and the Federal Circuit never reached the substantive issue whether the modules were inventory or depreciable assets, holding that Diebold could not change accounting methods from inventory treatment to depreciable asset treatment without the Commissioner's consent.

The IRS quickly adopted the Diebold decision as proof that a taxpayer can adopt a method of accounting by characterizing a transaction. What the IRS ignores, however, is that the two courts never addressed the fact that, if Diebold was right that the modules were depreciable, gain upon disposition of the modules could have been capital gain and given rise to a permanent difference in tax. In fact, the parties never briefed this issue, or the similar issue of the permanent loss of ITC. Under the 1993 Travelers case, the Federal Circuit might have held for the taxpayer on one or both of these "not just timing" arguments.

On the substantive issue whether replacement computer modules are inventory or depreciable assets, the courts have upheld other taxpayers' ability to claim depreciation and ITC. See Honeywell, Inc., 64 T.C.M. (CCH) 437 (1992), aff'd, 27 F.3d 571 (8th Cir. 1994); Hewlett-Packard Co. v. United States, 71 F.3d 398 (Fed. Cir. 1995), rev'g Apollo Computer, Inc. v. United States, 32 Fed. Cl. 334 (1994). Setting aside the possibility that Diebold might have come out differently if the briefs had focused on the ITC and the possibility of capital gain, the IRS's successful denial to Diebold of a substantive treatment that other taxpayers have been allowed shows the power of section 446(e). Diebold was also unlucky that the other cases came after its case. Unlike Mr. Underhill, Diebold did not have the chance to argue that it should not lose the correct substantive legal treatment that other courts had allowed, just because it did not ask for permission.

* Natural Gas Transportation and Storage. In Transwestern Pipeline Co. v. United States, 225 Ct. Cl. 399 (1980), the Court of Claims held, per curiam, that "line pack" gas that was needed to pressurize the taxpayer's pipeline, but 80 to 97 percent of which would be lost permanently whenever the pipeline was abandoned, was properly accounted for as a depreciable capital asset, instead of as inventory.

In ARKLA, Inc. v. United States, 765 F.2d 487 (5th Cir. 1985), cert. denied, 475 U.S. 1064 (1985), the taxpayer sought to extend the rationale of Transwestern Pipeline to "cushion" gas needed to pressurize the taxpayer's storage reservoir. In ARKLA, only five percent of the cushion gas would be lost permanently when the reservoir was abandoned. This led the Fifth Circuit to allow depreciation and ITC on the nonrecoverable cushion gas, but to require the recoverable cushion gas to be accounted for as inventory.

In Pacific Enterprises v. Commissioner, 101 T.C. 1 (1993), the Tax Court held that recoverable and nonrecoverable cushion gas are both capital assets. Judge Cohen reasoned that because the recoverable cushion gas would not be recovered until the pipeline was abandoned, it seemed clear that, during the years at issue, that volume of gas was not held for sale to customers. Unfortunately, the company had overestimated the amount of "working gas" (i.e., inventory) in two of its reservoirs and, during the years at issue, had reclassified 8.8 billion cubic feet of natural gas from working gas to cushion gas (i.e., capital asset). The Tax Court held that although the taxpayer's characterization was substantively correct, the reclassification was a change in accounting method that required the Commissioner's consent.

Pacific Enterprises is unique because the judge held for the taxpayer on the substantive characterization issue, but denied the taxpayer the benefit of that substantive holding because it failed to ask the IRS for permission to reclassify its gas. The judge distinguished a number of pro-taxpayer change-in-method cases on the grounds that they did not involve inventory items. The regulations are more strict, she said unconvincingly, with respect to changing the identification of inventory items.

One explanation of Pacific Enterprises is that Judge Cohen had just decided in Honeywell that rotable spare parts were capital assets and not inventory. She would have had a very hard time distinguishing the substance of the cushion gas situation, and yet, she was uncomfortable disagreeing with the Fifth Circuit's decision in ARKLA. So in the opinion, she did not highlight her disagreement with the Fifth Circuit's substantive holding, and then she used the change-in-method issue to reach the same bottom line as the Fifth Circuit, but by a different path. Whether or not this explanation is correct, Pacific Enterprises is a bad precedent with which taxpayers now have to cope.

Income or Loan?

In Commissioner v. Indianapolis Power & Light Co., 493 U.S. 203 (1990), the Supreme Court was forced to decide whether deposits received from utility customers should be characterized as income when received, or as loans repayable to the customers at a later date. The IRS contended that because the deposits were often used to pay for electricity (when a customer either defaulted on its bills or elected to apply the deposit to its last bill when terminating service), and because the taxpayer in any event controlled the deposited funds for an extended period, the deposits should be reported as income when received. The Supreme Court disagreed, holding that because the taxpayer was obliged to repay the deposits (when a customer either satisfied a credit test or elected to take back the deposit when terminating service), the taxpayer did not have the sort of "complete dominion" over the deposited funds that would be necessary to characterize the deposits as income. The court decided to characterize the deposits as loans, like tenants' security deposits.

The Supreme Court's opinion covers only the substantive legal issue -- how should the deposits be characterized? -- and makes no reference to any change in accounting method by Indianapolis Power & Light. There were, however, some other utility companies that had reported deposits as income when received. And many utilities had accounted for deposits as loans, but had agreed on audit or at Appeals to start reporting the deposits as income. After the Supreme Court's decision, these taxpayers all wanted to go back to reporting their deposits as loans.

The IRS quickly published Rev. Proc. 91-31, 1991-1 C.B. 566, which characterized the prepaid income-versus-loan issue in Indianapolis Power & Light as a change in accounting method and graciously allowed taxpayers to start accounting for deposits as loans after the Supreme Court's decision. The IRS said that unless a taxpayer complies with the terms of Rev. Proc. 91-31 or applies for a change in method in the traditional way, the taxpayer may not account for deposits consistent with the Supreme Court's decision.

That would be a fun case to litigate. A case that involves no omission or duplication of income -- only a timing difference -- and the IRS tries to rely on its revenue procedure to keep a taxpayer from following the Supreme Courts decision. The IRS has a legitimate interest in an orderly transition after a Supreme Court decision, but it will be difficult for government counsel to look any court in the eye and say --

We know the Supreme Court concluded that, as a matter of substantive law, this was a loan. But the Commissioner did not say "Simon says this was a loan," so the Commissioner's statutory right to keep the gate on changes in accounting method trumps the courts' constitutional power to say what the substantive law is.

The other side of that oral argument would be more fun.

Income or Contribution to Capital?

In Saline Sewer Co., 63 T.C.M. (CCH) 2832 (1992), the taxpayer treated customer connection fees as nontaxable contributions to capital and did not claim depreciation deductions with respect to assets acquired with the nontaxable contributions. The IRS tried to recharacterize the connection fees as taxable income. The court agreed with the taxpayer that "section 481 does not apply in this case because the issue with respect to the fees is a question of whether or not the fees are taxable, not a timing question as to when the fees should be considered income."

It is hard to say why the court came out where it did. It is possible the court concluded that the fee income issue and the depreciation issue both gave rise to permanent differences in the taxpayer's lifetime income because it looked at the two separately, instead of together. On the other hand, perhaps the court's decision was based on the judge's visceral reaction to the Commissioner's attempt to turn a characterization issue into a change-in-method issue. Taxpayers may need to wait for the next characterization case to learn whether the court would have come out the same place if it were clear that there was no change in the taxpayer's lifetime income if the fee income and the depreciation deductions were analyzed together.

Stock Sale or Sales Discount?

In Sun Microsystems, Inc., 66 T.C.M. (CCH) 997 (1993), the Tax Court held that stock warrants granted to purchasers of computer workstations represented a trade discount extended to the purchaser, not a sale of stock. The seller was therefore allowed to reduce its gross sales by its cost of honoring the warrants.(5)

Convergent Technologies, Inc., 70 T.C.M. (CCH) 87 (1995), presented facts very similar to Sun Microsystems, except that the taxpayer initially amortized the value of the warrants as if they were sales of stock and later amended its returns to treat the cost of honoring the warrants as sales discounts that reduced gross sales. The IRS argued that the taxpayer's amended returns were unauthorized changes in method of accounting because the change from amortization to deduction was one of timing. Tax Court Judge Tannenwald agreed that the case would have presented a timing question if the IRS had attempted to return the taxpayer to its original method of accounting. But the situation changed when the IRS urged the court to hold that the warrants represented nondeductible sales of stock. "Under these circumstances, we think that respondent's action changes the focus of this case from one of timing, as respondent contends, to one of characterization of the item, as petitioner contends. Consequently, consent of respondent is not required."

What Is the Property?

My law firm represented a taxpayer that incurred a casualty loss to timber property. In a similar case, Westvaco Corp. v. United States, 639 F.2d 700 (Ct. Cl. 1980), the court had held that the single identifiable property damaged or destroyed by a casualty was the timber depletion block, not the part of the block that was physically affected by the casualty. When our taxpayer began in 1980 to claim casualty loss deductions limited to its basis in the depletion block, government counsel threatened to assert that this was a change in method of accounting that required consent. He argued that even after Supreme Court cases, the IRS has to consent to taxpayers changing to the treatment the Court approves. We thought it would be interesting to watch him explain to a judge that, even after the Court of Claims said that the depletion block was the definition of the property for tax purposes, the Commissioner could ignore that result by refusing to consent to taxpayers' requests to change their method.

Note that if government counsel had been correct that the definition of the property is a method of accounting, taxpayers like L.B. Maytag (discussed above under the heading "Not Just Timing") would never permanently lose their loss deductions by misdefining the property that was abandoned, lost, or damaged by casualty. Section 481 would be available to restore losses that should have been claimed in prior (even closed) years. So it is possible that when this argument is eventually made to a court, it will be made by a taxpayer, not the IRS, and the IRS will be torn about whether to contest the breadth of the taxpayer's proposed definition of "change in method of accounting."

Error

A change in method of accounting does not include correction of mathematical or posting errors, or errors in the computation of tax liability (such as errors in computation of the foreign tax credit, net operating loss, percentage depletion, or investment credit). Treas. Reg. [sections] 1.446-1(e)(2)(ii)(b). The IRS thinks taxpayers overuse the word "error" in this context, but most of the cases are attributable to errors.

Under the regulations, an error is not the same as an erroneous method of accounting. Again, the paradigmatic example in the regulations is that a change from consistently claiming current deductions in the year of purchase to depreciating a class of depreciable assets is a change in method of accounting because the difference is only timing. The Commissioner's permission is required despite the fact that the taxpayer's old method is clearly wrong and the Code prescribes the only permissible depreciation life and method.(6) The courts have held likewise in a number of cases.(7)

In other cases, however, the courts have held that taxpayers' errors were not changes in methods, either because the taxpayer never intended to reverse the effect of the error (i.e., the difference was "not just timing"), or the taxpayer's error was inadvertent, or the taxpayer incorrectly applied its chosen method of accounting.

An Error That Will Not Reverse

In Schuster's Express, Inc. v. Commissioner, 66 T. C. 588 (1976), nonacq. 1978-2 C.B. 4, aff'd without published opinion, 562 F.2d 39 (2d Cir. 1977), a taxpayer deducted estimated (instead of actual) costs of purchasing insurance for, and self-insuring against, the risk of damage done to customers' goods during transit. The taxpayer conceded that it had to correct its error and deduct only actual costs for 1968-1970, but the IRS wanted to use a section 481 adjustment to recapture the excess deductions claimed in 1966 and 1967, which were closed years. The Tax Court held that the taxpayer's error was not a method, so that section 481 could not be used to reopen the closed years.

Here, the court said, there was no indication that the taxpayer ever intended to reconcile its estimated deductions to its actual costs. The court's suspicion was supported by the evidence -- the difference between the estimated and actual costs grew every year. Because large companies do not often take deductions that are never reconciled to reality, Schuster's Express may not come up very often.

Inadvertent Errors

North Carolina Granite Corp. v. Commissioner, 43 T.C. 149 (1964), involved a taxpayer that properly maintained inventories for purposes of computing taxable income, but failed to take into account its inventories in computing the "taxable income limitation" on its percentage depletion deductions. The Tax Court held that the taxpayer made a mistake in computing the taxable income limitation, but that "[t]he fact that petitioner may have repeated its mistake for a number of years does not transform it into a `method of accounting." 43 T.C. at 167. Even if the mistake were a method, the court said, it is the kind of method that the taxpayer can change without consent because the general requirement of consent yields to the specific prescription of a method of accounting or computation for the percentage depletion deduction. 43 T.C. at 168.

Other "inadvertent error" cases include Buddy Schoellkopf Products, Inc. v. Commissioner, 65 T.C. 640 (1975), nonacq. 1981-2 C.B. 3 (which is discussed below under the heading "First Year") and Korn Industries, Inc. v. United States, 209 Ct. Cl. 559 (1976) (where a furniture manufacturer failed to include the cost of 3 out of 14 materials in its finished goods inventory, the Court of Claims concluded that the taxpayer's failure to capitalize the 3 costs was not a method of accounting, but an inadvertent error).

The IRS announced in Rev. Rul. 77-134, 1977-1 C.B. 132, that it will not follow Korn Industries, which is also a case where the error was inconsistent with the taxpayer's chosen method of accounting.

Errors That Are Inconsistent with the Taxpayer's Chosen Method of Accounting

In Gimbel Brothers, Inc. v. United States, 535 F.2d 14 (Ct. Cl. 1976), the taxpayer accounted for most of its charge accounts on the installment method, but used the less favorable accrual method for one type of charge account that the IRS did not think properly fit the definition of "installment sale." The IRS later changed its mind, and the taxpayer applied the installment method to the last type of charge account. The Court of Claims found that once the taxpayer elected to use the installment method, the Code required the taxpayer to use that method for all its installment sales. Thus, the court allowed the taxpayer to amend its returns to correctly report the installment sales that the taxpayer had improperly reported on the accrual method.

The government made an alternative argument that Gimbel Brothers had adopted an erroneous method of accounting by consistently reporting some installment sales on the accrual method, and needed permission to change from its erroneous method. The Court of Claims rejected that argument, citing two revenue rulings that were later superseded and revoked by Rev. Rul. 90-38, 1990-1 C.B. 57. Although the IRS might argue that Gimbel Brothers did not survive Diebold, there is no such suggestion in the Federal Circuit's references to Gimbel Brothers in Diebold. 891 F.2d at 1582.

The real fight in Gimbel Brothers was over the scope of the taxpayer's method of accounting for charge accounts. The IRS argued that Gimbel Brothers had established an accrual method of accounting for one type of charge account. The taxpayer won the case by convincing the court that the installment sale rules required all charge sales that met the definition of "installment sale" to be accounted for under the installment method.

In Standard Oil Co. (Indiana) v. Commissioner, 77 T.C. 349 (1981), acq. on other issues, 1989-2 C.B. 1, the taxpayer consistently capitalized platform costs that could have been deducted as intangible drilling costs (IDCs). But the taxpayer had deducted other IDCs. The court held that, having elected to deduct (as opposed to capitalize) its IDCs, the taxpayer had to deduct all IDCs. Thus, the court held that the taxpayer's consistent failure to deduct certain costs that qualified as IDCs was not a method of accounting, but an error that could be corrected without consent.

Like Gimbel Brothers, the real fight in Standard Oil was over the scope of the method of accounting the taxpayer had chosen. The taxpayer prevailed by convincing the court that it was required to deduct all costs that met the definition of "intangible drilling cost," not just the types of costs that it recognized as IDCs when it filed its returns.

How far can this argument be extended? Treas. Reg. [sections] 1.461-1(a)(3) says that the expense, liability, or loss of one year generally cannot be used to reduce the income of a subsequent year. Could an accrual method taxpayer argue that this regulation requires it to deduct all costs in the proper year so that, as the court found in Standard Oil, any deviation from that rule is an error instead of a method of accounting? The Tax Court said in North Carolina Granite that the general must yield to the more specific, but how does the general rule in the section 461 regulations measure up against the statement in the section 446 regulations that a taxpayer needs permission to change a method except as otherwise expressly provided by the Code or regulations? Are the section 461 regulations express enough?

This line of cases is ripe for expansion because most taxpayers occasionally make errors that are inconsistent with their chosen method of accounting. Because the IRS has been aggressively asserting that corrections are changes in method, there is sure to be more case law in this area. What, then, distinguishes the cases that taxpayers won from the cases (e.g., in footnote [7]) that taxpayers lost? Is it possible that some judges decide whether or not a treatment is a method after they decide whether they want to allow it, based on the substantive law and the equities? Maybe taxpayers only succeed in claiming "error" when the judge thinks the mistake was inadvertent, as opposed to an attempt to flip-flop between treatments when the taxpayer discovers that another treatment would save tax. In most of the cases taxpayers won, there was less risk that the taxpayer was using hindsight because it appeared that the taxpayer's "erroneous" treatment never would save tax.

Current Adjustment

Certain practices that do involve the proper time for taking a deduction are "traditionally corrected by adjustments in current and future years." Treas. Reg. [sections] 1.446-1(e)(2)(ii)(b) gives two examples -- an adjustment to a reserve for bad debts under the section 166 regulations (which is no longer available to most taxpayers), and an adjustment to the useful life of a depreciable asset under the section 167(b) regulations (which is no longer available for most assets, which are depreciated under section 168 or section 197).

This exception for practices that are traditionally corrected by adjustments for current and future years can sometimes arise, however, and should apply to items of income, as well as to deductions, such as --

* The nonaccrual-experience method for service providers under section 448(d)(5) and Temp. Reg. [sections] 1.448-2T.

* Long-term contract accounting under section 460 and Treas. Reg. [sections] 1.460-6 (and before that under Treas. Reg. [sections] 1.451-3).

* The special method of accounting for trading stamps and coupons under Treas. Reg. [sections] 1.451-4.

* The income forecast method of depreciation under Rev. Rul. 60-358, 1960-2 C.B. 68.

* Cost depletion under section 611(a) and Treas. Reg. [sections] 1.611-2(c)(2).

It is not clear whether anything short of a statutory provision, regulation, or revenue ruling would satisfy the IRS that a practice is "traditionally" corrected by adjustments in current and future years. But consider inventory shrinkage under Dayton Hudson and the other cases now pending before the Tax Court. It seems likely that when the Tax Court approves the deduction at year end of estimated shrinkage, it will endorse adjustments to those deductions when physical inventories are taken and when estimates are made for the next year.

Change In Facts

A change in method of accounting does not include a change in treatment resulting from a change in underlying facts. Treas. Reg. [sections] 1.446-1(e)(2)(ii)(b). These "Changes in Fact" are as common as the previous category, "Current Adjustments," are rare. The regulations give two examples,(8) but they do not fairly represent the versatility of the change-in-facts doctrine.

Changing the Date of Sale

The Tax Court has shown that it will respect legitimate changes in facts, even if the changes were recommended by the tax department. In Decision, Inc. v. Commissioner, 47 T.C. 58 (1966), acq. 1967-2 C.B. 2, the taxpayer had always billed advertisers in November for advertising space in its job directories to be published the following February. In 1963, the taxpayer decided to change the terms of its arrangement for the 1964 directories, notifying advertisers that it would not bill them until January of 1964. The IRS argued that this was a change in method of accounting that left expenses, but no gross income, to be recognized in 1963. The Tax Court held that Decision, Inc. was free to change the date of accrual by changing the terms on which it dealt with its customers:

Although the change had consequences in the annual determination of income, such consequences were not produced by the accounting system.... To sustain [the IRS's change-in-method argument] would have the effect of denying a business the right to determine the terms of sale of its product without clearing the matter with the Commissioner of Internal Revenue, clearly an odious propagation of the tentacles of the Government anemone.

47 T.C. at 64.

The Tax Court held similarly in Hallmark Cards, Inc. v. Commissioner, 90 T.C. 26 (1988). The taxpayer had historically sold Valentine cards to retailers before year end, but changed its contracts in 1958 to provide that Hallmark would retain until January 1 title and risk of loss with respect to Valentine goods delivered before year end. This new arrangement meant that retailers would not have to include in ending inventory or pay personal property tax on, the Valentine goods. It also meant that Hallmark could defer recognizing income from the sale of those goods until January 1. Citing the "odious propagation" language from Decision, Inc., the Tax Court held once again that a taxpayer is free to change the time when taxable income accrues by changing the terms of its sales. There was no change in method, only a change in facts. 90 T.C. at 35.

Although the court said, in footnote 6 of its Hallmark opinion, that it expressed no opinion on the tax consequences of a situation where the taxpayer deliberately manipulated the terms of sale in order to prevent income from accruing, the opinion's reliance on Decision, Inc. speaks for itself. It is not clear why it should even matter whether the tax department suggested a change to the terms of sale. In any event, it would be rare for tax considerations alone to drive a change in the terms of sale. On the other hand, if a taxpayer gets too aggressive, the Commissioner might assert that the new result, however carefully the new terms of sale dovetail with the all-events test, does not clearly reflect income.

Changing Estimating Formulas

In The Baltimore & Ohio R.R. v. United States, 221 Ct. Cl. 16 (1979), the Court of Claims found an entirely different type of change in underlying facts -- a change in estimating formulas. The taxpayer depreciated its track under the retirement-replacement-betterment (RRB) method of accounting, which requires the taxpayer to estimate the value of used rail. Because there was no market for trading used rail, however, the taxpayer from 1918 until 1963 estimated the value of its used rail at 75 percent of the price of new rail. It then appears that when the Tax Court approved a new formula in Chesapeake & Ohio Ry. v. Commissioner, 64 T.C. 352, 392 (1975), the taxpayer amended its returns back to 1955 to use the new formula, which increased its depreciation deductions. The IRS conceded that the new formula more accurately estimated the value of the taxpayer's used rail, but asked the court to hold that changing formulas was a change in method requiring consent. The court refused, saying that the value of the used rail was a fact, not a method. Once the fact of value was determined, it would be plugged into the RRB accounting method. But a change in the formula was at most a change in an underlying fact. 221 Ct. Cl. at 24-25.

This is another case where the court resisted the IRS's invitation to deny a taxpayer the better treatment -- the substantive treatment allowed by another court -- merely because the taxpayer failed to ask the IRS for permission. A similar use of the "change is underlying facts" doctrine appears in ESCO Corp. v. United States, 750 F.2d 1466 (9th Cir. 1985), where the taxpayer's improvement of its formula for estimating accrued workers' compensation claims was held not to be a change in method, but rather a change in underlying facts.(9)

The Taxpayers' Ability to Change the Facts

The example that opens this article lists five different ways a taxpayer might choose to account for a $120 million prepayment. If the business folks would allow some fairly modest differences in the economics, the tax department might endeavor to write that contract with the customer so that any one of those five treatments could be used on the taxpayer's return:

* If the taxpayer had expiring NOLs and wanted to include the entire amount in this year's income, the contract might be drafted to come under the Supreme Court's decision in Schlude.

* If the taxpayer wanted to defer the entire amount until Year 10 when it knows whether it won or lost the bet, a prepaid forward contract could be prepared.

* Including $1 million a month for the next 120 months should not be hard to write.

* If the taxpayer wanted to include $1 per unit as units are actually delivered, the transaction could be structured as an advance payment for goods under Treas. Reg. [sections] 1.451-5.

* And if the taxpayer really wanted to revalue the contract every time the price of its product rises or falls, and recognize loss or gain as the buyer's rights become more or less valuable, the taxpayer might be made to qualify as a dealer in forward contracts under section 475.

Some of these contracts might be harder to write than others, but the taxpayer's ability to change the facts ex ante is a powerful tool. And remember, a taxpayer cannot establish a method of accounting for a particular type of transaction before it has reported that type of transaction on its tax returns, and the taxpayer is permitted to choose its method of accounting for a new type of transaction without the Commissioner's consent.

Inconsistency 

In most instances, a method is not established without consistent treatment. Treas. Reg. [sections] 1.446-1(e)(2)(ii)(a) provides that, although a method of accounting may exist "without the necessity of a pattern of consistent treatment of an item, in most instances a method of accounting is not established for an item without such consistent treatment." Treas. Reg. [sections][sections] 1.446-1(e)(2)(iii), Examples (4), (6), (7), and (8) are explicit that the taxpayer is changing a practice that it has consistently applied for three years, many years, or "always." The other four examples in the regulations are not explicit, but imply consistent use of the taxpayer's old practice.

Does Inconsistency Mean That a Practice Is Not a Method of Accounting? The Two-Return Rule for Improper Treatments

Rev. Proc. 92-20, 1992-1 C.B. 685, 688 provides that the incorrect treatment of an item is the same way on two consecutively filed tax returns constitutes consistent treatment, and therefore a method of accounting, under the regulations. Rev. Proc. 92-20 cites Rev. Rul. 90-38, 1990-1 C.B. 57, which relies on Diebold for the proposition that a taxpayer cannot change a consistent method of accounting using an amended return.

Rev. Proc. 92-20 states that if the taxpayer treats an item incorrectly on one tax return, however, the taxpayer has not adopted a method of accounting, and can correct the error on an amended return. 1992-1 C.B. at 688. See also Robert M. Foley v. Commissioner, 56 T.C. 765 (1971), acq. 1972-2 C.B. 2, and Rev. Rul. 72-491, 1972-2 C.B. 104 (where the taxpayer was permitted to correct an improper treatment in Year 1 on an amended return filed before the return for Year 2 was filed).

When the change involves the time for claiming a deduction or loss, consider the interplay between the IRS's attempts to prevent taxpayers from changing methods by amended returns and Treas. Reg. [sections] 1.461-1(a)(3), which provides that deductions may not be claimed for a later taxable year if they should have been claimed in an earlier taxable year. Perhaps Treas. Reg. [sections] 1.461-1(a)(3) takes precedence, not only because it is a regulation, but because of the statement in Treas. Reg. [sections] 1.446-1(e)(2)(i) that permission is required to change a method, "[e]xcept as otherwise expressly provided" by the Code or regulations.

The One-Return Rule for Permissible Methods

The two-return rule in Rev. Proc. 92-20 does not apply if the taxpayer's treatment of the item was permissible. Even if a permissible treatment is used only once, the taxpayer cannot change to another permissible treatment without consent. Rev. Rul. 90-38, 1990-1 C.B. 57, 58. The one-return rule for permissible methods under Rev. Proc. 92-20 and Rev. Rul. 90-38 is consistent with Foley, but contrary to Buddy Schoellkopf (discussed below under the heading "First Year"). Absent unusual facts, it is hard to persuade a court to allow changes to permissible methods because of the risk that the taxpayer is using hindsight to lower its tax bill.

First Year

There is an important exception to the consistency rules of Rev. Proc. 92-20 and Rev. Rul. 90-38 if the IRS challenges an improper method for the first year it was used. If the taxpayer has used an improper treatment on two consecutive returns, but the IRS disallows the treatment for the first year it was used by the taxpayer, the disallowance prevents the taxpayer from adopting the incorrect treatment as a method of accounting. Because the taxpayer has not adopted a method of accounting, it is free to adopt any permissible method for the first year. Rev. Rul. 72-491, 1972-2 C.B. 104. This exception reflects the facts and reasoning in Silver Queen Motel v. Commissioner, 55 T.C. 1101 (1971), acq. 1972-2 C.B. 3. The IRS's failure to discuss this part of Rev. Rul. 72-491 in Rev. Rul. 90-38 implies that the IRS agrees that this exception survived the Federal Circuit's decision in Diebold.

What Kinds of Cases Can You Argue Under the Theory of Rev. Rul. 72-491?

What if the IRS disallows the taxpayer's improper treatment for the second or third year it was used by the taxpayer, but the first year is still open? Treas. Reg. [sections] 601.601(d)(2)(v)(e) and the front of each volume of the Cumulative Bulletin warn that revenue rulings do not bind the IRS unless the pivotal facts are substantially the same. But would a court allow the IRS to avoid the spirit of Rev. Rul. 72-491 by challenging the taxpayer's improper treatment for the second or third year, and taking the position that the taxpayer had therefore adopted an improper method and needed permission to change? In light of the IRS's two-return rule for improper methods, a court should simply apply the rationale of Silver Queen and hold for the taxpayer if the IRS challenges an improper treatment in the second year it was used. It would be very interesting to see the IRS challenge an improper treatment in the third year it was used, while the first year is still open.

What if the IRS disallows the taxpayer's permissible method for the first year it was used? For example, what if the IRS asserts that the permissible method does not clearly reflect the taxpayer's income? Again, a court should apply the rationale of Silver Queen and allow the taxpayer to pick another permissible method because the IRS's challenge prevented the taxpayer from adopting the method that did not clearly reflect its income. This example illustrates, however, an apparent inconsistency between the Silver Queen rationale and the plain language of section 446(b): "If no method of accounting has been regularly used by the taxpayer, or 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." Silver Queen walks a fine line between the taxpayer's right to pick its initial method of accounting, and the IRS's right to pick a new method of accounting if the taxpayer has not used consistently a proper method of accounting. The court read the depreciation regulations as providing that the IRS may pick the new method only if the taxpayer refuses to select a proper method even after the IRS disallows the improper method. Rev. Rul. 72-491 and the IRS's acquiescence in Silver Queen can be read as conceding that section 446(b) should be read the same way. And in Convergent Technologies, the Tax Court recognized that the Silver Queen rationale is not limited to depreciation cases.

The Courts Tend to Give the IRS a Hard Time Whenever the First Year Is Still Open

In the IRS's favorite case, Diebold, the taxpayer was trying to treat its spare parts as depreciable assets after it had accounted for spare parts as inventory in two closed years. The Federal Circuit mentioned this fact in distinguishing its holding in Diebold from the taxpayer victories in Gimbel Brothers and Standard Oil. 891 F.2d at 1582. Diebold tried to explain that the spare parts were held for use under a different type of service contract during the closed years, but the court was not convinced: "Section 446(e) prohibits taxpayers from unilaterally amending their tax returns simply because they have discovered that a different method of accounting yields a lower tax liability than the method they originally chose." 891 F.2d at 1583.

Why should it matter if the first year is open? Is it possible that the courts are offended by the IRS's trying to stick taxpayers with improper methods of accounting, even if all years are still open, when the improper method benefits the fisc? What legitimate interest does the IRS have in defending the improper method in this sort of case? There is no risk of omission or duplication of income, and no need for a section 481 adjustment, if all of the years are still open. There is less risk that taxpayers are flip-flopping (as they could among permissible methods) because another method has now become more advantageous. And the Commissioner is not objecting to the appropriateness of the new method. In fact, the new method is better than the old and could have been adopted by the taxpayer in the first year without the Commissioner's oversight. Thus, of the three reasons for the consent requirement -- preventing omission or duplication of income, promoting consistency (and preventing flip-flopping), and providing the IRS with the opportunity to assess the propriety of the method -- the only reason to regulate first year changes from improper methods is if there is evidence that the taxpayer is using hindsight.

This raises an interesting possibility. How would a court react if a taxpayer that is regularly audited as part of the Coordinated Examination Program served up a proposed change to an improper method for the first year it was used? See Rev. Proc. 94-69, 1994-2 C.B. 804 (allowing CEP taxpayers 15 days to offer adjustments of items that were not disclosed on the return). Would the court apply the Silver Queen principle because the first year is in play, or would it limit the Silver Queen principle to cases where the revenue agents find and disallow the improper method without the taxpayer's assistance?

In One Case, a Court Gave the IRS a Hard Time Even Though the First Three Years Were Closed

In Buddy Schoellkopf Products, Inc. v. Commissioner, 65 T.C. 640 (1975), nonacq. 1981-2 C.B. 3, the taxpayer purchased a used aircraft and used an improper double declining balance method of depreciation for three years. The agent switched the taxpayer to the straight-line method in Year 4, but the taxpayer argued that it should be permitted to use the 150-percent declining balance method that it could have claimed when the plane was placed in service. The Tax Court agreed with the taxpayer, reasoning that the change from the erroneous method to a proper method of depreciation was not a change in method of accounting, or was a change that did not require the Commissioner's consent. The court said that the fact that the first three years were closed and the taxpayer's error could only be corrected beginning with the fourth year was not a sufficient reason to not apply the principle of Silver Queen.

What principle was the court referring to? Apparently not the principle that the IRS prevented the taxpayer from adopting a method by challenging it for the first year. Buddy Schoellkopf Products arguably stands for the proposition that in this area, as in many others, the outcome of a court case will turn on which party -- the taxpayer or the IRS -- tried to cherry-pick the system ex post. How would the case would have come down if, instead of trying to use the change-in-method doctrine to put the taxpayer on the straight-line method, the government had asked the Tax Court to allow a section 481 adjustment to recapture the benefit of the 200-percent declining balance method that the taxpayer claimed during the closed years? Perhaps no differently. See Korn Industries, cited above under the heading "Error."

Materiality 

A change in method of accounting includes a change in the overall plan of accounting for gross income or deductions or a change in the treatment of any 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. [sections] 1.446-1(e)(2)(ii)(a).(10)

Although some de minimis test would help in many real-world situations, the regulations do not use the normal dictionary or accounting definition of "material." It seems that, under the regulations, any item that has even a minute effect on the timing of the taxpayer's lifetime taxable income is a "material item." Reportedly, certain practitioner groups asked the IRS to add this unusual definition of "material" to the regulations in 1970, and the IRS obliged. Is it possible that these folks had clients with exposure in open years earlier than 1954, and wanted to protect the oldest years under section 481?

Is there any room to argue that a more traditional notion of materiality should apply? The courts have not historically been inclined to import the accountants' notion of materiality into the tax law. And courts probably assume that taxpayers do not choose to incur the time and cost to litigate over truly "immaterial" items.

In Witte v. Commissioner, 513 F.2d 391, 393 n.4 (D.C. Cir. 1975), rev'g 31 T.C.M. (CCH) 1137 (1972), the United States Court of Appeals for the D.C. Circuit surveyed the case law on the materiality of accounting method items and held that a change in the way the taxpayer reported $88,000 of gain over three taxable years was material. In Southern Pacific Transp. Co. v. Commissioner, 75 T.C. 497, 684 (1980), the Tax Court declined to create an exception from the regulations' definition of "material item" for a change in the treatment of $122,000 of expenses over three taxable years.

Materiality is an argument that rarely works as well as it should.

New Business

Treas. Reg. [sections] 1.446-1(d) provides that where a taxpayer has two or more separate trades or businesses, a different method of accounting may be used for each trade or business, as long as the method used for each business clearly reflects the income of that business. Because taxpayers do not need the Commissioner's consent to adopt (as opposed to change) accounting methods, there is sometimes an opportunity to avoid the need to ask permission if a new activity can be treated as a separate trade or business that has not yet adopted a method of accounting for similar items. The trick is to get comfortable that the new business is truly separate from the old business (which long ago adopted the method that you do not want to apply to this new activity). This determination is based on all the facts and circumstances, but at a minimum, the regulations require:

* Separate books and records, and

* No shifting of profit or loss between the businesses.

It is easiest to get comfortable in the consolidated group context. Tears. Reg. [sections] 1.1502-17(a) provides that each member of a consolidated group is entitled to adopt its own accounting methods. Again, because taxpayers do not need the Commissioner's consent to adopt accounting methods, there is sometimes an opportunity to form a new subsidiary by transferring assets tax-free under section 351. This technique can work, and is a key tool for dealing with certain inventory problems that can arise.

The IRS knows about this technique, and is currently thinking about trying to extend its victories in Hamilton Industries v. Commissioner, 97 T.C. 120 (1991), and Kohler Co. v. United States, 34 Fed. Cl. 379 (1995), to cover transfers of inventory under section 351. In those two cases, the IRS convinced the courts that inventory purchased at a bargain should not be included in the same LIFO inventory with physically identical goods purchased later at a market price. It is not clear that the courts will see the transfer of appreciated inventory under section 351 as equivalent to a bargain purchase.

Automatic

Some changes in method are permitted without consent. For example, when the IRS publishes final regulations that require many taxpayers to change their methods of accounting for an item, the IRS usually publishes a revenue procedure granting automatic consent for taxpayers to change from their current method to the method required by the new regulations.(11) The IRS generally attaches certain terms and conditions to the automatic consent, and taxpayers who do not fit or do not want the terms and conditions can generally request consent under the normal procedures.

When a court decides a case that either allows or requires many taxpayers to change their methods of accounting for an item, the IRS often publishes a notice or revenue procedure granting automatic consent for taxpayers to change from their current method to the method allowed or required by the court.(12) This practice is discussed above under the heading "Characterization."

Occasionally, the IRS publishes a regulation or revenue procedure that allows many taxpayers to make a change in method without consent because the IRS feels it can adequately protect its interests using terms and conditions, and it does not want to process all of the Forms 3115.(13)

There is no complete catalog of these "automatic changes," so taxpayers need to search for whether there are any automatic change regulations or revenue procedures that they might have used without knowing it. A taxpayer may not have dotted every i and crossed every t, but the existence of an automatic change could help win the day. A good place to start is with the "automatic change" procedures that the IRS publishes each year in the first revenue procedure of the year. See, e.g., Rev. Proc. 96-1, 1996-1 I.R.B. 8, 33-34.

Concluding Thoughts: Where Should We Go from Here?

What do these 10 arguments teach us about changes in method? That the IRS's authority in this area is incredibly broad. Frankly, it is tough to win these issues at IRS. They like to adhere to their revenue procedures because they fear that any unusual result will be jumped on by other taxpayers who want similar treatment. The Claims Court is a much friendlier forum than the IRS, and the Tax Court is probably better still.

What needs to change?(14) First, there should be a definition of immaterial changes to screen out small changes that could safely be made without advance consent.(14) The standard could be:

* Quantitative, like a percentage of taxable income over the last three years;

* Qualitative, like the materiality standard under GAAP;

* Protected against "nibblers" by writing a cumulative change rule to limit a taxpayer's ability to make several small (related or unrelated) changes that grow large in the aggregate; or

* Protected against "the camel's nose" so that a taxpayer does not make a change without consent for the year before it becomes material (e.g., because the taxpayer knew that it was about to expand a particular line of business).

The only thing the IRS could not require is disclosure of the changes that a taxpayer made without permission because this new standard would be for changes so small that the tax department does not even know they were made. But if new taxpayers can adopt huge methods without consent, there must be some change so small that an existing taxpayer can be trusted to make it without advance consent.

Second, changes from improper methods to proper methods of accounting should be encouraged. On a substantive level, these situations are reminiscent of the automatic change revenue procedures that the IRS issues after a new regulation or court decision is published. The IRS took a step in the right direction in Rev. Proc. 92-20, when it formalized the "audit protection" for taxpayers who request a change in method. Now, when a taxpayer comes forward with a request to change a method, it knows that a revenue agent cannot propose that the taxpayer make the change retroactively for all open years. How could the IRS encourage other changes?

When a revenue agent proposes a change in method for a year under audit, the taxpayer knows that even if it agrees to the change, there is some residual risk that a subsequent audit team will question either the new method or the company's changing methods without the permission of the National Office. The Court of Federal Claims' very recent decision in Travelers Insurance Co. v. United States, No. 262-89T (March 21, 1996), reduced this risk somewhat by holding that the IRS abused its section 446(b) discretion when it challenged a method that the IRS had twice approved on audit. Still, if the taxpayer does not want to bear the residual risk (or if the agent does not want the responsibility for approving the change without formal authority), the taxpayer has to continue using the old method for at least this year while applying for permission from the National Office. Even if the National Office permits the change, the taxpayer will have to amend this year's return and any intervening year's returns. The only thing more frustrating is to change the method without delay, only to have the National Office not approve the change that the revenue agent proposed -- and this has happened on more than one occasion.

This can be a real Catch-22, not to mention expensive, for a taxpayer that is making investment or other business decisions each day based on its assumption about the timing of income or deductions for tax purposes. Why not formally delegate to case managers the authority to change a method of accounting, so that taxpayers are encouraged, not discouraged, to settle these issues on audit? Section 10.14 of Rev. Proc. 92-20 implies that an Appeals officer with delegated settlement authority can approve a change in method of accounting as part of a settlement. The Commissioner has been giving CEP case managers more power; why not give them this power?(15) In addition to removing the residual risk that might make a taxpayer less inclined to agree to an agent-proposed change, delegating this authority to the case manager might allow the audit team to approve an occasional taxpayer-proposed change.

Third, most changes in method could be permitted without consent if the taxpayer applied a "cut-off" transition to the new method. In a cut-off, items arising after the change are accounted for under the new method, while the old method continues to govern items arising in prior years. Because there is no need for a section 481(a) adjustment, the need for the Commissioner to consent to the change in advance is reduced considerably. Cut-offs are already used in affecting LIFO and long-term contract changes -- areas that carry the potential for some of the largest section 481(a) adjustments. And cut-off is the de facto technique that taxpayers use when they change a method of accounting without telling the tax department.

Until changes like this are made, taxpayers and tax practitioners will continue to see huge change-in-method disputes on a regular basis, and taxpayers and the IRS will expend huge sums fighting over an issue that should not be this hard.

(1) In one recent case, Chief Judge Smith of the Court of Federal Claims held that the IRS abused its discretion in claiming that the taxpayer's method did not clearly reflect its income and in claiming that the agent's method did clearly reflect income. Perhaps most remarkably, Judge Smith held that the court was entitled to apply de novo review of the issue whether the taxpayer's method clearly reflected its income. Travelers Insurance Co. v. United States, No. 262-89T (March 21, 1996).

(2) Two caveats apply to the cited cases. First, the cases support the general propositions for which they are cited, but in an intensely factual area like this one, finding the case that has similar facts to your case can be the difference between victory and defeat. Fortunately, there are hundreds of cases in this area from which to choose. Second, most of the cases in this area set forth alternative arguments for their holdings. Each case is categorized according to the argument for which it is best suited, although in truth many could be said to support three or four different arguments.

(3) Would a court react any differently if Congress expressed the capital gains rate for the taxpayer and year at issue as an exclusion? It is difficult to imagine that Congress thought about change in method when deciding whether to draft the capital gains preference as a lower rate or an exclusion.

(4) The government's position on brief seemed to echo a decision made by the IRS and Treasury six years previously when new regulations were issued under section 446(e). Prop. Reg. [sections] 1.446-1(e)(2)(ii)(b) referred to a change from lease treatment to sale treatment as a change in method, but that example was not included in the final regulations.

(5) See also Computervision International Corp., T.C. Memo 1996-131, holding that Sun's purchaser must treat the redemption of the warrants as a trade discount that reduced its basis in the goods.

(6) See also Treas. Reg. [sections][sections] 1.446-1(e)(2)(iii), Example (1) (taxpayer with inventories improperly using cash method; change to accrual method is a change in method of accounting); Example (2) (accrual method taxpayer improperly deducting real estate taxes when paid; change to accrual of real estate taxes is a change in method of accounting); Example (6) (taxpayer improperly failing to capitalize any overhead costs into inventories; change to allocating overhead costs to inventories is a change in method of accounting); Example (7) (taxpayer improperly discounting cost of ending inventory value by 20-percent reserve for price changes; change to properly valuing inventories at cost is a change in method of accounting); and Example (8) (taxpayer improperly using base stock inventory system; change to a proper inventory system is a change in method of accounting).

(7) See, e.g., Fruehauf Trailer Co. v. Commissioner, 42 T.C. 83 (1964), acq. 1965-2 C.B. 5, aff'd, 356 F.2d 975 (6th Cir. 1966) (valuing used trailer inventory at $1 per unit not permitted; change to LCM was a change in method of accounting); Peoples Bank & Trust Co. v. Commissioner, 50 T.C. 750 (1968), aff'd, 415 F.2d 1341 (7th Cir. 1969) (deducting estimated accrued interest on December 31 on deposits that would only be credited interest if they were still on hand the following May 1 not permitted; change to all events test was a change in method of accounting); Primo Pants Co. v. Commissioner, 78 T.C. 705 (1982) (valuing materials and work in process at percentage of cost, finished pants at a percentage of selling price, and not allocating labor and overhead costs to finished goods not permitted; change to proper LCM was a change in method of accounting); Wayne Bolt & Nut Co. v. Commissioner, 93 T.C. 500 (1989) (taxpayer's perpetual inventory included only 10 percent of the goods on hand; change to physical inventories was a change in method of accounting giving rise to a section 481 adjustment that included 20 closed years); Pacific Enterprises v. Commissioner, 101 T.C. 1 (1993) (taxpayer's reclassification of 8.8 billion cubic feet of gas from inventory to capital asset was a change in accounting method; court seemed to assume that "error" is not an available argument in the case of misidentified inventory).

(8) Treas. Reg. [sections][sections] 1.446-1(e)(2)(iii), Example (3) (accrual method taxpayer properly deducting unvested vacation pay in the year it is paid; changing to the accrual method when the taxpayer adopts a completely vested vacation pay plan is not a change in method of accounting); and Example (4) (taxpayer allocating indirect overhead costs to inventory using a fixed percentage of direct costs; changing the percentage when the ratio of indirect to direct costs changes is not a change in method of accounting).

(9) But see Pacific Enterprises v. Commissioner, 101 T.C. 1 (1993) (discussed in the text under the heading "Characterization"), which held that a taxpayer's engineers' reclassification of 8.8 billion cubic feet of gas from inventory to capital asset was a change in accounting method, not a change in facts. FERC Order 636, however, may largely moot the holding in Pacific Enterprises. By preventing pipeline companies from holding gas for sale, FERC may have created a nationwide change in facts that allows pipeline companies to reclassify gas without permission.

(10) A change in the treatment of any material item used in the overall plan for identifying or valuing items in inventory is also a change in method of accounting. Treas. Reg. [sections] 1.446-1(e)(2)(ii)(c). Although the definition of "material item" is not repeated, it seems prudent to assume that the definition applies here, too.

(11) See, e.g., Rev. Proc. 95-19, 1995-1 C.B. 664, one of five revenue procedures that allow taxpayers to adopt the final interest capitalization regulations under section 263A(f); Rev. Proc. 94-28, 1994-1 C.B. 614, one of three revenue procedures that allow taxpayers to adopt the final OID regulations under sections 1272-1275; and Rev. Proc. 93-48, 1993-2 C.B. 580, which allows taxpayers to adopt the final notional principal contract regulations under section 446.

(12) See, e.g., Notice 95-57, 1995-45 I.R.B. 12, which allowed banks in the Eighth Circuit to change to the cash method of accounting for stated interest on short-term loans after the Security Bank of Minnesota decision; Rev. Proc. 91-31, 1991-1 C.B. 566, which characterized the prepaid income-versus-loan issue in Indianapolis Power & Light as a change in accounting method and allowed taxpayers to start accounting for deposits as loans after the Supreme Court's decision; and Rev. Proc. 80-5, 1980-1 C.B. 582, which "allowed" taxpayers to conform their accounting for obsolete inventory to the Supreme Court's Thor Power Tool Co. decision.

(13) See, e.g., Treas. Reg. [sections] 1.442-1(c), which allows corporations to change their year end without consent if they meet certain conditions; Rev. Proc. 90-63, 1990-2 C.B. 664, which allows taxpayers to change to their choice of three methods of amortizing package design costs; and Rev. Proc. 88-15, 1988-1 C.B. 683, which allows some taxpayers to change from LIFO to FIFO without consent.

(14) Hal I. Gann & Roy E. Strowd, Jr., Methods of Changing Accounting, TAX NOTES, April 10, 1995, at 289.

(15) See Delegation Order No. 247, Authority of Examination Case Managers to Accept Settlement Offers and Executive Closing Agreements on Industry Specialization Program (ISP) and International Field Assistance Specialization Program (IFASP) Issues, Manual Transmittal, 1229-202 (April 5, 1996).