# United States Tax Court > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/agency%3Atax-court%3Ab9e38ef349629577 ## Record - **Collection:** Agency decision - **Document type:** Agency decision ## Text United States Tax Court T.C. Memo. 2022-31 CONTINUING LIFE COMMUNITIES THOUSAND OAKS LLC, SPIEKER CLC, LLC, TAX MATTERS PARTNER, Petitioner v. COMMISSIONER OF INTERNAL REVENUE, Respondent ————— Docket No. 4806-15. Filed April 6, 2022. ————— Fred B. Weil, Christopher A. Karachale, Lawrence M. Cirelli, and Wendy Abkin, for petitioner. Melanie E. Senick and Gregory Michael Hahn, for respondent. MEMORANDUM OPINION HOLMES, Judge: Continuing-care communities are a new business for old Americans. They promise to provide housing and health-care assistance and they promise to do so for as long as their residents live. The residents pay very large sums both upfront and over time to the communities to provide for services of uncertain duration and cost. The communities must make sure they earn returns on their investments despite these uncertainties. And state governments recognize that elderly people who have given up very large amounts of money in exchange for a promise need to have someone make sure the promise is kept. This case is about how a company that owns one such community had to account for a portion of the upfront payments from its residents when it calculated its taxable income for 2008–10. The company followed generally accepted accounting principles in recognizing when Served 04/06/22 2 [*2] and how much of these payments it reported on its returns. The Commissioner says that’s not good enough. The appropriate accounting for these payments is important to companies across the nation that provide continuing care. It is not one we’ve addressed thoroughly before. Background The taxpayer here was named Continuing Life Communities Thousand Oaks, LLC during the years at issue. 1 It is a Delaware LLC with its principal place of business in Thousand Oaks, California. I. The Continuing Care Industry Continuing Life’s business is to provide housing and care to seniors even as their needs change. A new resident might need only housing and food, but as time batters away he may need more. And although Continuing Life is not a hospital, it does promise to provide for its residents’ needs all the way through skilled nursing care. The range of services that it promises costs a lot. And this is reflected in the entry fee—the initial payment that Continuing Life charges to move into the community—and in large monthly payments too. Continuing Life is no outlier—industry surveys show that the entry fees in similar communities average $402,000, with some at over $2 million, and that monthly service fees run between $2,000 and $4,000. 2 See infra pp. 8–9. California lawmakers recognized the potential for abuse here— there is a lot of money at risk, and seniors are, or may eventually become, susceptible to undue influence—so they have put the industry under strict regulation. California’s legislature noted that “tragic consequences can result if a continuing care provider becomes insolvent or unable to provide responsible care.” Cal. Health & Safety Code § 1770(b) (West 2001). California law thus requires them to provide lifetime care. It calls this a “continuing care promise,” and the contracts are called “life care contracts.” Id. § 1771. The continuing-care promise 1 In 2013, Continuing Life changed its name to University Village Thousand Oaks CCRC, LLC. In some of the documents we later quote this is abbreviated as UVTO. 2 How Continuing Care Retirement Communities Work, AARP (Oct. 24, 2019), https://www.aarp.org/caregiving/basics/info-2017/continuing-care-retirementcommunities.html (last updated Jan. 27, 2022). 3 [*3] is defined as “a promise, expressed or implied, by a provider to provide one or more elements of care to an elderly resident for the duration of his or her life.” Id. If a continuing-care community fails to fulfill a continuing-care promise, then it is subject to criminal and civil punishment and will not receive any payment for its services. Id. § 1793.5(d). 3 California isn’t alone; other states have also enacted strict regulations on continuing-care communities. See, e.g., N.J. Stat. Ann. § 52:27d-330 (West 2021) (New Jersey); 2021 N.M. Laws Ch. 56 (S.B. 152) (New Mexico); N.Y. Pub. Health Law § 4650 (McKinney 2021) (New York). Entry into the industry in California is controlled by the state’s Department of Social Services, which issues a certificate of authority for continuing-care communities to operate. Cal. Health & Safety Code § 1771.5. But continuing-care-community regulation doesn’t stop with living conditions and management; California also places minimum standards on continuing-care communities’ contractual obligations, id. § 1770(f), and requires that they provide financial statements to the residents, id. § 1771.8(f). It also requires that owners of continuing-care communities follow generally accepted accounting principles (GAAP) when they prepare those statements. See id. § 1771(a)(7). II. Continuing Life Communities Continuing Life owns and operates the continuing-care community in Thousand Oaks, California. Next door is Oakview at University Village, an assisted-living facility and skilled-nursing facility. A resident can move to Oakview if his health deteriorates and he needs additional support. Continuing Life built the community in the early 2000s and began accepting residents in 2007. It is open to individuals who are at least 62, but the average age of residents who moved in during the years at issue was approximately 82. Prospective residents choose from among 12 different floor plans that vary by their size, number of rooms, and availability of parking. Continuing Life provides one daily meal to its residents, along with other basic amenities, such as linen service and cleaning. It describes these services in a detailed form document that is central to this case—the Residence and Care Agreement, which makes what California law considered a 3 “An entity that abandons a continuing care retirement community or its obligations under a continuing care contract is guilty of a misdemeanor. An entity that violates this section shall be liable to the injured resident for treble the amount of damages assessed in any civil action brought by or on behalf of the resident . . . .” Cal. Health & Safety Code § 1793.5(d). 4 [*4] “continuing care promise” and is a life-care contract. New residents sign this Residence Agreement. 4 The Residence Agreement lays out all the rights and obligations between the residents and Continuing Life. California’s Department of Social Services must approve life-care contracts, and the parties stipulated that for all years at issue, the Department has done so. The parties also stipulated that if Continuing Life violated the Residence Agreement by failing to satisfy the obligation to provide life-time care, it would not have been able to collect any fees or payments and would have been subject to criminal fines; those individuals responsible would face imprisonment. A specific part of the Residence Agreement is called the Joinder in Master Trust Agreement University Village Thousand Oaks Master Trust (Joinder Agreement). This Joinder Agreement is important because it’s how Continuing Life funds its operations and makes its money. To put its importance into perspective requires some description of the three fees Continuing Life charges: the Contribution Amount, the Deferred Fee, and monthly fees. These three charges are not just how Continuing Life earns income but they also give an insight into the economics of its industry. The Contribution Amount ranged during the years before us from $245,000 to $570,000 and was determined entirely by a resident’s choice of floor plan—his age, health, and life expectancy played no role in determining how Continuing Life set this amount. Another interesting feature of the Contribution Amount is that residents do not pay it to Continuing Life—they instead pay it, under the Joinder Agreement, to Kenneth Cummins as trustee of the Master Trust. When a resident makes this payment, he becomes a grantor of the Master Trust. Cummins and the Master Trust acted as a third-party intermediary between the residents and Continuing Life to ensure that Continuing Life’s finances were in check and the residents’ payments were properly accounted for. Continuing Life had this roundabout set up for a couple reasons. The legal reason is that California law requires it. Cal. Health & Safety Code § 1792.6(a) (“Any provider offering a refundable contract, or other entity assuming responsibility for refundable contracts, shall maintain a refund reserve in trust for the residents.”). But the Contribution Amount also had a business purpose: It provides “permanent financing for the UVTO campus and improvement” and “protect[ing] and 4 There are a few residents with “unusual circumstances” who do not sign the Residence Agreement themselves. These outliers have no effect on the outcome of this case. 5 [*5] conserv[ing] the Master Trust property for the benefit of the residents.” The benefit to Continuing Life’s financing comes from a provision in the Master Trust itself that authorizes Cummins to use Contribution Amounts to make interest-free loans to Continuing Life. Securing these loans is a deed of trust to all of Continuing Life’s real property and improvements, current and after-acquired equity in all of the improvements, fixtures, personal property, present and future leases and rents from the property, and intangible property associated and used in connection with Continuing Life’s real property. Continuing Life used these loans to make improvements to the campus, and Cummins annually tours the property to ensure that the collateral remains sufficient to secure the loans. He reports the status of the loans and collateral to the residents’ council and the council’s budget-andfinance subcommittee, and he also holds annual town hall meetings with the residents to provide information on the administration of the Master Trust. Cummins owes fiduciary duties only to the residents and not to Continuing Life: He earns his fees and pays the Trust’s expenses out of the pooled Contribution Amounts. Apart from these relatively small expenses of managing the Trust, this arrangement meant that a resident (and his heirs) would, others things being equal, be quite secure that the nominal value of the corpus of his Contribution Amount would be preserved. The Joinder Agreement and Master Trust provide that Cummins has to repay this amount whenever a Residence Agreement was terminated. Termination comes in three ways: death, voluntary departure, and expulsion. And here we come to the key bit of Continuing Life’s income that is the subject of these motions—the Deferred Fee. Section 12 of the Residence Agreement defines the Deferred Fee and calculates it as a percentage of the Contribution Amount. Section 12.5 states: “If this Agreement is terminated under Section 12.2 or 12.4.2, you or your estate shall pay Continuing Life Communities a Deferred Fee according to the following schedule:” 6 [*6] Time Elapsed After the Execution Date Deferred Fee as a Percentage of Contribution Amount 91 days to 1 year 5% 1 year and 1 day to 2 years 10% 2 years and 1 day to 3 years 15% 3 years and 1 day to 4 years 20% Longer than 4 years 25% There is an initial 90-day cancellation period—if a new resident dies or has second thoughts and chooses to terminate the Residence Agreement, he would not have to pay any part of the Deferred Fee. After 90 days, the Deferred Fee begins to accrue at 5% a year, maxing out after 4 years at 25%. The timing of the Deferred Fee payment is important. Residents do not write checks for it themselves. It comes instead out of the Master Trust—and even then not when 90 days or one year or four years pass, but only when a resident dies or moves out and a new resident buys the unit and pays his own Contribution Amount to Cummins as trustee. Any unpaid expenses that a departed resident still owes and his Deferred Fee come out of the Contribution Amount. Cummins on behalf of the Master Trust would then pay the balance to the resident or his estate. Timing is important here, so we’ll quote at length from the relevant sections of the Residence Agreement: 12.2 Termination By Resident After Cancellation Period You may terminate this Agreement at any time after the Cancellation Period for any reason . . . . Upon such termination of this Agreement, you shall pay Continuing Life Communities a Deferred Entrance Fee as set forth in the schedule in Section 12.5 . . . . Continuing Life Communities shall withhold from your Contribution Amount the Deferred Entrance Fee, all unpaid Monthly Fees, Fees for Optional Services, other charges . . . . 7 [*7] 12.3 Termination By Continuing Life Communities After Cancellation Period 12.3.1 Right to Termination Continuing Life Communities may terminate this Agreement at any time after the Cancellation Period for good cause . . . . 12.3.3 Refund to Residents If Continuing Life Communities terminates this Agreement after the Cancellation Period, you may be entitled to a refund of amounts paid by you under this Agreement minus an amount to cover costs and the reasonable value of the services, care, and residence actually provided to you . . . . Continuing Life Communities shall withhold from your refund all unpaid Monthly Fees, Fees for Optional Services, interest and late charges due, and other charges incurred by you . . . . 12.4 Death of Resident 12.4.2 After Cancellation Period If you die after the Cancellation Period, this Agreement shall automatically terminate. Upon such termination of this Agreement, your estate or personal representative shall pay Continuing Life Communities a Deferred Entrance Fee, as set forth in the schedule in Section 12.5 . . . . Continuing Life Communities shall withhold from your Contribution Amount the Deferred Entrance Fee, all unpaid Monthly Fees, Fees for Optional Services . . . . (Emphases added.) 8 [*8] Note especially that Continuing Life gets no Deferred Fee if it expels a resident, and that the Residence Agreement speaks of the payment of the Deferred Fee by a resident who chooses to leave or who dies as a promise of future payment. From 2008 through 2010, a total of 27 residents left—15 by death, 11 by voluntary departure, and only 1 by expulsion for good cause. There is no dispute that Continuing Life did not receive any Deferred Fee from the resident that it expelled. The tables below list the residents who voluntarily terminated the Residence Agreement, and the residents whose passing terminated the Residence Agreement for them. Name Residence Agreement Signing Date Date of Termination Contribution Amount Former Resident 1 03/29/2005 07/01/2008 $475,000 Former Resident 2 10/11/2007 11/01/2008 307,000 Former Resident 3 (couple) 09/25/2007 06/30/2008 481,000 Former Resident 4 01/22/2007 02/01/2009 599,000 Former Resident 5 09/24/2008 07/16/2009 379,700 Former Resident 6 09/26/2007 08/01/2009 313,000 Former Resident 7 01/22/2007 09/01/2009 361,300 Former Resident 8 08/12/2009 12/01/2009 359,700 Former Resident 9 09/29/2009 12/01/2010 598,300 Former Resident 10 10/04/2007 08/16/2010 575,300 Former Resident 11 10/15/2009 01/15/2010 598,300 9 Name Residence Agreement Signing Date Date of Death Contribution Amount Deceased Resident 1 01/22/2007 07/06/2008 $397,500 Deceased Resident 2 09/25/2007 10/03/2009 498,000 Deceased Resident 3 10/03/2007 10/15/2009 504,750 Deceased Resident 4 10/08/2007 07/26/2009 504,750 Deceased Resident 5 02/17/2008 01/30/2010 344,400 Deceased Resident 6 10/04/2007 02/08/2010 397,500 Deceased Resident 7 03/30/2009 11/27/2009 359,700 Deceased Resident 8 (couple) 10/02/2008 04/15/2010 598,300 Deceased Resident 9 10/03/2007 09/24/2010 359,700 Deceased Resident 10 09/25/2007 11/01/2010 385,700 Deceased Resident 11 01/28/2008 08/25/2010 359,700 Deceased Resident 12 10/03/2007 11/21/2009 365,700 Deceased Resident 13 09/26/2007 02/14/2010 278,000 Deceased Resident 14 09/25/2007 04/12/2010 385,700 Deceased Resident 15 09/25/2007 12/15/2009 303,000 [*9] The final source of Continuing Life’s income is the monthly fees. Continuing Life set these fees using the community’s operating cost, the 10 [*10] prior year’s per capita costs, and other economic indicators. 5 Besides the costs to provide lifetime care, these monthly fees pay other expenses, including electricity, water, gas, and trash collection. Continuing Life itemizes optional utilities, such as cable TV, internet, and telephone services, separately from these monthly fees. The amounts of these monthly fees is fixed by the particular floor plan that the resident chooses. Continuing Life re-evaluates the monthly fees every year, and in its annual report lists any reasons for changes to them. If a resident with unpaid monthly fees dies, moves, or is expelled, Cummins would subtract the unpaid fees from the refundable portion of the Contribution Amount. One can see in this some kind of effort by Continuing Life to roughly match its initial and continuing capital costs to the interest-free use of Contribution Amounts and Deferred Fees, and its operating costs to the monthly fees. III. Accounting for Deferred Fees A. The AICPA and Position 90–8 An accounting maven will spot the issue here: In real life the probability that Continuing Life will expel a resident is low. For any longer term resident the probability that Continuing Life will actually collect the Deferred Fee is high. Most every resident dies or leaves; each has paid a very large Contribution Amount out of which the payment of the Deferred Fee is as a practical matter very well secured. But actual cash money won’t get to Continuing Life until what could well be many years after the first four years when the Deferred Fee maxes out, and the timing of any particular resident’s obligation to pay the Deferred Fee is quite uncertain. Accountants have their ways, however. The American Institute of Certified Public Accountants (AICPA) 6 has historically been the 5 According to the Residence Agreement, these economic indicators include but are not limited to “cost of purchased health care for skilled nursing and assisted living at OakView or a similar facility, insurance costs, prudent reserves, general and administrative costs, general operating costs, taxes, interest and principal on UVTO related mortgages and loans, services in kind, and depreciation and operating profits, among others.” 6 The AICPA was chartered in 1887 as the American Association of Public Accountants; changed its name in 1916 to the American Institute of Accountants, and again in 1956 to its current name. See H. Dubroff, M. Cahill, M. Norris, Tax Accounting: The Relationship of Clear Reflection of Income to Generally Accepted 11 [*11] predominant source of accounting standards. It created one of the main sources of standards for the profession, the Financial Accounting Standards Board (FASB). As the continuing-care industry took root and grew, the AICPA noticed that the industry’s accounting practices were somewhat ad hoc and thought that specific guidance was needed to “achieve uniform reporting practices.” 7 So the AICPA pondered the matter and in 1990 released the AICPA Audit and Accounting Guide Statement of Position 90–8 (Nov. 28, 1990). 8 Position 90–8 describes and dissects in detail many different accounting issues that continuingcare communities face. And it’s a discussion that has to be nuanced— different providers face different state regulations, use different form contracts, and get paid in different ways. We’ll focus only on the provisions that are relevant for this case. The key provisions are those that deal with advance fees. Position 90– 8, para. 15 defines an advance fee as a “payment required to be made by a resident prior to, or at the time of, admission.” Some continuing-care communities refund the total amount or a portion of the advance fee on the occurrence of a specified event. These amounts are called the refundable portion, and the remainder is called the nonrefundable portion. The refundable portion is credited as a liability, and the nonrefundable portion is accounted for as deferred revenue. 9 Id. paras. 20–23. For the nonrefundable portion, Position 90–8 again recognized the “wide diversity of practice exist[ing] among [continuing-care retirement communities] when accounting for nonrefundable advance fees.” Id. para. 34. Although there are eight listed methods, only two are relevant here. Paragraph 35 provides that one method recognizes nonrefundable advance fees “as revenue in the period the fees are receivable if future periodic fees can reasonably be expected to cover the cost of future services.” Paragraph 36 provides a second method, which Accounting Principles, 47 Albany L. Rev. 354, 366 n.59 (1983). To this day, the AICPA is active in “framing standards of accounting practice, defining terminology, and standardizing procedures and forms of presentation.” Id. 7 The AICPA did note that one practice that all continuing-care communities could agree on is that immediately reporting refundable advance fees as income is unacceptable. 8 FASB later adopted the AICPA’s guidance in 2009, and published it under the Accounting Standards Codification 954–430. 9 Accrual accounting uses deferred revenue as a way to keep track of money received, but not yet earned. As the business performs those services, the deferred revenue account is converted into revenue. See Boise Cascade Corp v. United States, 208 Ct. Cl. 619, 625 (1976). 12 [*12] defers recognition of nonrefundable advance fees and amortizes them into income as consideration for providing future services. This method treats the nonrefundable advance fees as future costs that “are not recoverable from other revenue sources.” 10 And, as a result, the matching principle 11 requires that the nonrefundable advance fee be deferred until the expenses arise. The AICPA came down on the side of this latter method: [N]onrefundable advance fees represent payment for future services and should be accounted for as deferred revenue . . . . Nonrefundable advance fees should be amortized [to income over future periods based on the estimated life of the resident]. Id. para. 43. There’s a subtle but important point here, which is the effect that periodic fees have on the accounting method. The AICPA specifically mentioned and considered periodic fees, and yet chose to leave them out of its adopted method. We conclude that this means that the AICPA thought that the reasoning under paragraph 36 was more convincing and decided that the periodic fees generally could not by themselves cover future costs. It might even mean that if periodic fees completely covered future operating costs, AICPA believed that the method that it adopted would still satisfy the matching principle. B. Continuing Life’s Accounting for Deferred Fees We finally get to the specific method that Continuing Life used during the years at issue. California law requires Continuing Life to follow GAAP. The Commissioner concedes that Continuing Life has followed GAAP, and the parties stipulate that Continuing Life followed Position 90–8’s guidance. When residents paid the Contribution Amount (and there can be no dispute that this amount meets the Proponents for this method argued that substantially all of the services specified in the contract have not been performed, so recognizing income under the first method would not match revenues and expenses. Position 90–8, para. 36. 10 11 The matching principle is an important financial accounting concept “which states that the revenues and related expenses must be matched in the same period to which they relate.” See Rashid Javed, Matching principle of accounting, ACCOUNTINGFORMANAGEMENT.ORG, https://www.accountingformanagement.org/matching-principle-accounting/ (last updated Oct. 20, 2021). 13 [*13] definition of “advance fee”) to Cummins, Continuing Life did not recognize any income. But as each year passed, Continuing Life amortized and recognized as income a fraction of the Deferred Fees (and there is no dispute that these are “nonrefundable advance fees”) by using the straight-line method and the actuarially determined estimated life of each resident. When the resident moved or died, Continuing Life would recognize the remaining unamortized Deferred Fee as income. Note that this also meant that Continuing Life recognized the nonrefundable amount as income before it resold the departed resident’s residence and actually got cash money from the Master Trustee. And remember as well that, throughout this process, residents were paying the monthly fees (which we have no doubt meet the definition of “periodic fees”) which covered at least some of the operating costs of the community. This accounting method had two notable effects. Because the estimated life of each resident is actuarially determined on a year-byyear basis, the method requires yearly modifications to each resident’s estimated life expectancy. And because the method amortizes income over life expectancy, it allows Continuing Life to defer recognizing the unamortized portion of the Deferred Fees until a Residence Agreement is terminated, when Continuing Life accelerates recognition of the remaining unamortized Deferred Fees. One can see the effect on Continuing Life’s Form 1065, U.S. Return of Partnership Income. For all years at issue, Continuing Life had substantial losses as its deductions were vastly greater than its gross income: It took losses of about $9.2 million in 2008, $3.15 million in 2009, and $850,000 in 2010. 12 During those years, Continuing Life recognized Deferred Fee income of only $34,188 in 2008, $420,187 in 2009, and $421,727 in 2010. IV. Audit The Commissioner audited Continuing Life, and in November 2014 he sent the notice of final partnership administrative adjustment (FPAA) for the 2008–10 tax years that proposed increasing Continuing Life’s tax bill by nearly $20 million. The parties agree on the facts and both moved for summary judgment. The only issue is whether 12 As more residents moved in, Continuing Life’s gross income began to increase: It had gross receipts of about $13 million in 2008, $16 million in 2009, and $18 million in 2010. 14 [*14] Continuing Life’s accounting for the Deferred Fees is allowed under the Code. Continuing Life is a TEFRA partnership, 13 and Spieker CLC, LLC, is its tax matters partner. 14 Any appeal would presumptively go to the Ninth Circuit. See § 7482(b)(1)(E). Discussion One way to think about tax law is to view it as a series of general rules qualified by exceptions, and exceptions to those exceptions, and exceptions to those exceptions to those exceptions. This may be a helpful way to begin to think about the tax-accounting issue we have to analyze in this case. For Continuing Life the general rule is that it gets to follow its own method of accounting. See § 446(a). To be sure, there’s an exception to this general rule for methods of accounting that do not clearly reflect income or that a taxpayer doesn’t follow consistently. § 446(b). And, as Continuing Life also points out, there’s an actual regulation that says that a “method of accounting which reflects the consistent application of generally accepted accounting principles in a particular trade or business in accordance with accepted conditions or practices in that trade or business will ordinarily be regarded as clearly reflecting income.” Treas. Reg. § 1.446-1(a)(2). It notes that the Commissioner agrees that its treatment of Deferred Fees is in accordance with GAAP standards for the continuing-care industry. It recognizes that caselaw over the decades has created an exception to this general rule to give the 13 Before its repeal, see Bipartisan Budget Act of 2015, Pub. L. No. 114-74, § 1101(a), 129 Stat. 584, 625, part of the Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA), Pub. L. No. 97-248, §§ 401–407, 96 Stat. 324, 648–71, governed the tax treatment and audit procedures for many partnerships. TEFRA partnerships were subject to special tax and audit rules. See §§ 6221–6234. TEFRA required the uniform treatment of all “partnership item[s]”—a term defined by section 6231(a)(3)—and its general goal was to have a single point of adjustment for the IRS rather than having it make separate partnership-item adjustments on each partner's individual return. See H.R. Rep. No. 97-760, at 599–601 (1982) (Conf. Rep.), 1982-2 C.B. 600, 662–63. If the IRS decided to adjust any partnership items on a partnership return, it had to notify the individual partners of the adjustment by issuing an FPAA. § 6223(a). (Unless otherwise indicated, all statutory references are to the Internal Revenue Code, Title 26 U.S.C., in effect at all relevant times, all regulation references are to the Code of Federal Regulations, Title 26 (Treas. Reg.), in effect at all relevant times, and all Rule references are to the Tax Court Rules of Practice and Procedure.) 14 Under TEFRA, a partnership designated one of its partners as the tax matters partner to handle its administrative issues with the Commissioner and manage any resulting litigation. § 6231(a)(7). 15 [*15] Commissioner some kind of discretion in determining whether a particular accounting method “clearly reflects income,” but insists that its method still clearly reflects income. For the Commissioner to disagree is therefore an abuse of that discretion. The Commissioner has a different perspective. He agrees that the general rule is that a taxpayer gets to follow its own consistent method of accounting. He agrees too that there is an exception for taxpayers whose method of accounting does not clearly reflect income. But he differs on who gets to decide whether a taxpayer’s accounting method clearly reflects income—the Commissioner can point to a long line of cases that say that he gets to decide whether a particular method of accounting for income clearly reflects income. He acknowledges that there’s an exception to this exception when his decision is an abuse of discretion, but argues that Continuing Life has not shown that it was. Both parties argue that the question of the effect of a taxpayer’s following GAAP in its tax accounting is an old one. It is also an old question that has metamorphosed into several different methods—like the sometimes jumbled geology of a roadside cut, these methods are not perfectly defined and are often themselves the compressed sediment of different approaches to statutory interpretation. The most important case in the field and certainly one that shows in its analysis all the layers relevant to this problem remains Thor Power Tool Co. v. Commissioner, 439 U.S. 522 (1979). In some ways, Thor would seem to have been an easy case: The taxpayer customized its own way of accounting for what it perceived to be the loss in value of its inventory of excess spare parts for products that it no longer made. Id. at 527–28. Thor produced “distinguished” accounting professionals at trial who testified that it had followed GAAP. Thor Power Tool Co. v. Commissioner, 64 T.C. 154, 165 (1975), aff’d, 563 F.2d 861 (7th Cir. 1977), aff’d, 439 U.S. 522 (1979). We found as a fact that Thor had followed GAAP, and the Commissioner didn’t fight the point too much. Id. That’s probably because there were some other serious weaknesses in Thor’s case. For one thing, there was little doubt that its method of accounting violated a specific regulation. Thor, 439 U.S. at 535. And then there was the major problem that Thor couldn’t really explain the value that it did put on this inventory—it didn’t sell any of it and didn’t keep records of its closing inventory, relying instead on “a well-educated guess.” Id. at 536. 16 [*16] The outcome for a taxpayer like this is not in doubt. But the Supreme Court’s opinion turned out to be rich in interpretive ambiguity. It had at least four distinct grounds. The first was the obvious textual argument—Thor’s accounting violated a valid regulation that governed the specific question of inventory accounting at issue. Id. at 535. Regulations have the force of law, and can be trumped only by the Code or the Constitution. See Adams Challenge (UK) Ltd. v. Commissioner, 154 T.C. 37, 64 (2020). But the Court didn’t stop there. Instead it also said that GAAP and tax accounting have different purposes, and so the Code and regulations shouldn’t be read to let a taxpayer argue that because he follows GAAP on a particular question he should presumptively win: [T]he presumption petitioner postulates is insupportable in light of the vastly different objectives that financial and tax accounting have. The primary goal of financial accounting is to provide useful information to management, shareholders, creditors, and others properly interested; the major responsibility of the accountant is to protect these parties from being misled. The primary goal of the income tax system, in contrast, is the equitable collection of revenue; the major responsibility of the Internal Revenue Service is to protect the public fisc. Thor, 439 U.S. at 542. And then, with both text and purpose mixed in, the Court also held that the Commissioner has an unusually broad power of discretion to set aside a taxpayer’s method of accounting “if, ‘in [his] opinion,’ it does not reflect income clearly.” Id. at 540. If in the Commissioner’s opinion GAAP doesn’t pass muster for tax purposes, then he has the discretion to “prescribe a different practice without having to rebut any presumption running against the Treasury.” Id. This unusual standard—not just the presumption of correctness that the Commissioner gets whenever he issues a notice of deficiency, Rule 142(a); Welch v. Helvering, 290 U.S. 111, 115 (1933)—but a seemingly heightened measure of discretion reversible only for its abuse, is nowhere in the Code or regulations. It is, however, found in caselaw that has stood for nearly a century in one form or another and, as a fourth distinct reason for its holding, the Court in Thor also relied on 17 [*17] the consistency of the result it reached with this caselaw. 439 U.S. at 540–42. This has left the lower courts with rich veins to chisel at and choose from. It has also meant that tax cases that address questions of tax accounting resemble less traditional sculpture and more an assemblage. Some go for the textual option and ask first to see what the Code says, then what valid regulations say, then whatever subregulatory guidance entitled to some level of deference says. Peninsula Steel Prods. & Equip. Co. v. Commissioner, 78 T.C. 1029, 1037–43 (1982); Photo-Sonics, Inc. v. Commissioner, 42 T.C. 926 (1964), aff’d, 357 F.2d 656 (9th. Cir. 1966); Shasta Indus., Inc. v. Commissioner, 52 T.C.M. (CCH) 190, 197 (1986). Reasoning by analogy has a place, but it’s a place that helps a court understand and not bypass hammering away at the meaning of the Code and regulations. Some courts try to reason from the purpose of tax accounting, and ask whether a particular instance of applying GAAP would leave a cashrich taxpayer with the opportunity to defer tax on his hoard in some improper way. Am. Auto. Ass’n v. United States, 149 Ct. Cl. 324, 330 (1960), aff’d, 367 U.S. 687 (1961); Auto. Club of Mich. v. Commissioner, 20 T.C. 1033, 1046–47 (1953), aff’d, 230 F.2d 585 (6th Cir. 1956), aff’d, 353 U.S. 180 (1957). Some courts reason in the common-law fashion by analogy to earlier cases—even to the point of not quoting Code or regulations, Highland Farms, Inc. & Subs. v. Commissioner, 106 T.C. 237, 250–51 (1996) (not citing regulations); Stendig v. United States, 843 F.2d 163 (4th Cir. 1988) (not citing Code or regulations); Auburn Packing Co. v. Commissioner, 60 T.C. 794, 800 (1973); and some yield to general statements of the Commissioner’s discretion on the question of whether a particular method of accounting “clearly reflects income,” RLC Indus. Co. & Subs. v. Commissioner, 98 T.C. 457, 491 (1992), aff’d, 58 F.3d 413 (9th Cir. 1995). There are few cases like Thor, where all these approaches end up at the same destination. And there are few courts that will just pick one approach to the exclusion of others. We will begin with what seems to be the dominant approach of courts today—a focused attention on the text of the relevant law. 18 [*18] I. A. Textual Analysis Section 446, GAAP, and a Taxpayer’s Regular Method of Accounting A taxpayer must compute taxable income under the “method of accounting on the basis of which the taxpayer regularly computes his income in keeping his books.” § 446(a). The Code provides four permissible accounting methods: cash receipts and disbursements; accrual; any method prescribed by chapter 1 of the Code; or a combination of the above methods that is prescribed by regulation. § 446(c). But regardless of the accounting method used, all accounting methods must “clearly reflect[] income.” § 446(b); Treas. Reg. § 1.4461(a)(2). “Clearly” as used in the statute means “plainly, honestly, straightforwardly and frankly, but does not mean ‘accurately’ which, in its ordinary use, means precisely, exactly, correctly, without error or defect.” Huntington Sec. Corp. v. Busey, 112 F.2d 368, 370 (6th Cir. 1940) (analyzing 1934 Code section 41, predecessor of section 446). Since at least the late 1950s there has been a regulation which provides that consistent compliance with GAAP in accordance with accepted conditions or practices in a trade or business “will ordinarily be regarded as clearly reflecting income.” Treas. Reg. § 1.446-1(a)(2). This leads us to the question of what effect GAAP compliance really has—or in other words, what does “ordinarily” mean? Neither statutes nor regulations provide any guidance. At oral argument we asked both parties for their interpretation of “ordinarily” as used in the regulations. Continuing Life argues that we should treat the words “ordinarily” and “generally” as a safe harbor from the Commissioner’s discretion to change its accounting method. But Continuing Life doesn’t cite any authority for this argument; nothing in the regulations supports this argument, and caselaw clearly contradicts it. The Commissioner argues that we should take “ordinarily” at face value and according to its plain meaning. He cites other instances where courts use the word “ordinarily”. But these cases interpret “ordinarily” to “indicate a certain flexibility of application rather than an undeviating practice,” Ralston Steel Car Co. v. Commissioner, 53 F.2d 948, 950 (6th Cir. 1931), and that “[o]rdinarily does not mean always,” Alison v. United States, 344 U.S. 167, 170 (1952). In Thor, 439 U.S. at 540, the Supreme Court seems to have interpreted “ordinarily” to be a statement of probability and not 19 [*19] presumption: “The Regulations embody no presumption; they say merely that, in most cases, generally accepted accounting practices will pass muster for tax purposes. And in most cases they will.” This has not proved to be a fruitful source of help for lower courts that need to apply the language of section 446 and its regulations. In tax-accounting cases where GAAP conflicts with regulatory language, Thor makes the answer easy, and consistent with a textual analysis: There are some situations where a regulation conflicts with GAAP; these are unusual, which is another way of saying that they are not “ordinary”. But what about the many cases where there is no regulation on point, and a taxpayer consistently follows GAAP in his accounting? Thor tells us that “ordinarily” is not a presumption rebuttable only by a showing that GAAP conflicts with a regulation or Code section. But our decisions after Thor edge us closer to an answer. We’ve held after Thor that compliance with GAAP is at least one factor we should look for in figuring out whether an accounting method clearly reflects income, and we must answer every question of whether an accounting method clearly reflects income as a question of fact which might vary from case to case. Ansley-Sheppard-Burgess Co. v. Commissioner, 104 T.C. 367, 371 (1995); RLC Indus., 98 T.C. at 492; Peninsula Steel Prods. & Equip. Co., 78 T.C. at 1045; Garth v. Commissioner, 56 T.C. 610, 618– 19 (1971); Hosp. Corp. of Am. & Subs. v. Commissioner, 71 T.C.M. (CCH) 2319, 2331 (1996). We take a prongified approach—listing some factors to squint at and answering the ultimate question after we’ve looked at each of those factors. We summarized the state of the law in RLC Indus., 98 T.C. at 502 (citations omitted): In a post-Thor environment respondent has been found to have appropriately exercised her discretion where a taxpayer’s method of accounting conflicted with the regulations. Similarly, where a taxpayer’s method was contrary to accounting principles, did not conform to industry practice, was not used for tax and financial reporting, and/or was not reliable, respondent was found to have appropriately exercised her discretion. In this case, the parties stipulate that Continuing Life has consistently applied GAAP. This is not quite the same as stipulating that its treatment of the Deferred Fees is accepted industry practice, 20 [*20] and we don’t have much expressly on this record about this, but under Position 90–8 this accounting method was one of two main views on the proper accounting method. We can infer from this that Position 90–8 states common industry practice. We note that our one other relevant case that involves a continuing-care community also followed the same AICPA accounting guidelines. See Highland Farms, 106 T.C. at 247. We can also look at how the Commissioner himself has said he construes “ordinarily”. 15 In GCM 39586 he used a slightly different list of factors: Consideration will be given to a variety of factors which include the validity of the method in matching income and expense, the consistent use of the method, the materiality of the item in dispute to the taxpayer’s overall income, the conformity of the disputed method with GAAP, and the economic realities of the transaction viewed on an annual rather than a transactional basis. I.R.S. Gen. Couns. Mem. 39586 (Dec. 3, 1986). Some of these overlap what we said in RLC we should look at— conformance with GAAP and a taxpayer’s consistency in its accounting—but some are different. Yet on the undisputed facts of this case, some of these additional factors also favor Continuing Life: The expenses that Continuing Life incurs because of its continuing-care promise are expenses that it incurs over the entire lifespan of each of its residents, yet it is entitled to the Deferred Fees only when residents depart from the community. We can conclude from this that Continuing Life’s method matches income and expenses better than the accelerated treatment that the Commissioner proposes. We can also conclude from this that Continuing Life’s method, even when viewed on an annual basis, looks like a better match than the Commissioner’s because it recognizes income each year that a resident continues to live in the community and thus impels Community Life to incur expenses on that resident’s behalf. The one remaining factor that we see identified in the Commissioner’s subregulatory guidance—the materiality of the item 15 This kind of subregulatory guidance is entitled to Skidmore deference—the deference a court owes to persuasive argument. See United States v. Mead Corp., 533 U.S. 218, 234 (2001); see also Skidmore v. Swift & Co., 323 U.S. 134 (1944). 21 [*21] compared to the taxpayer’s overall income—is one whose relevance here is quite unclear. Deferred Fees are without doubt material to Continuing Life’s bottom line, but their treatment as an accrued item by a taxpayer following the accrual method doesn’t mark them as out of the ordinary (as it might, for example, for a cash-method taxpayer who uses accrual accounting for a particular item of material expense). We conclude that the undisputed facts here show that there is no reason to conclude that Continuing Life’s use of GAAP accounting for the Deferred Fees takes it out of the ordinary rule that an accounting method consistent with GAAP accounting “clearly reflects income” under section 446. That is not the end of our textual analysis. For the Commissioner argues that, even if Continuing Life’s embrace of GAAP lets it fall within the general (or “ordinary”) case of clearly reflecting income, it still runs afoul of the Code and some of the particular rules that govern the recognition of income by taxpayers who use accrual accounting. B. Section 451 and Rules for Inclusion in Income In accrual accounting the regulation tells us, “income is includible in gross income when all the events have occurred which fix the right to receive such income and the amount thereof can be determined with reasonable accuracy.” Treas. Reg. § 1.451-1(a). The key inquiry is about when a taxpayer has a fixed “right to such compensation.” Id. “[I]f, in the case of compensation for services, no determination can be made as to the right to such compensation or the amount thereof until the services are completed, the amount of compensation is ordinarily income for the taxable year in which the determination can be made.” Id. This all-events test is the foundation of accrual accounting and a “fundamental principle of tax accounting.” 16 United States v. Hughes Props., Inc., 476 U.S. 593, 600 (1986) (quoting United States v. Consol. Edison Co. of N.Y., 366 U.S. 380, 385 (1961)). We look for when the taxpayer has a fixed right to income, not whether there has been actual payment. Schlude v. Commissioner, 372 U.S. 128, 137 (1963). The right to income is fixed when there is an unconditional right to receive 16 Although the all-events test rule was in the regulations during the years at issue, Congress gave it its own subsection in 2017. Tax Cuts and Jobs Act of 2017, Pub. L. No. 115-97, § 13221, 131 Stat. 2054, 2113 (codified at section 451(b)(1)(C)). 22 [*22] payment. Hallmark Cards, Inc. & Subs. v. Commissioner, 90 T.C. 26, 32 (1988). Determining when a taxpayer’s right to payment becomes “unconditional” is a blurry-line test. The inquiry can be confusing and often requires a close encounter with the mystical side of accounting. We must peer into the facts to see what “constitutes the very heart of the transaction,” and not events that are merely “ministerial” or “formalit[ies].” Id. at 32–33; cf. PGA Tour, Inc. v. Martin, 532 U.S. 661, 700 (2001) (Scalia, J., dissenting) (exploring Platonic essence of golf). We don’t need to consider possible contingencies to payment, but instead we look at the “existence or nonexistence of legal rights or obligations.” Id. at 34. The earliest of the following dates is typically when the right to income is fixed: the date the payment is received; the date the payment is due; or the date of performance. Schlude, 372 U.S. at 137; Johnson v. Commissioner, 108 T.C. 448, 459 (1997), aff’d in part, rev’d in part, 184 F.3d 786 (8th Cir. 1999); Harkins v. Commissioner, 81 T.C.M. (CCH) 1547, 1550 (2001). There is no dispute here that Continuing Life receives the Deferred Fee when the trustee closes out a client’s account after death or departure. We know that the Deferred Fee is due on the date that a client’s unit is reoccupied. But when has Continuing Life performed the services that entitle it to receive the Deferred Fee? This is the heart of the parties’ dispute. Continuing Life argues that the Residence Agreement provides that residents pay Deferred Fees only when they die or move out. Because it has an obligation to provide care for their entire lives, it also contends that its performance ends only when a resident dies or moves out. This means that Continuing Life’s right to the Deferred Fee also becomes fixed and definite only when a resident dies or moves out. California’s own regulation of this industry is important here. The Supreme Court has held that state law can fix a liability for accrual accounting purposes. Hughes, 476 U.S. at 601 (“Nevada Gaming Commission’s regulations fix liability”); see also Commissioner v. Indianapolis Power & Light Co., 493 U.S. 203, 205 (1990). But see Morning Star Packing Co. v. Commissioner, 120 T.C.M. (CCH) 259, 263 (2020) (vague references in contracts to comply with “all laws” not fixed and definite). We think state law can also help us fix the date a taxpayer has performed its obligations. Continuing Life drafted the Residence Agreement to comply with California law to win the Department of 23 [*23] Social Services’ approval. See Cal. Health & Safety Code § 1787(c). The Residence Agreement fits squarely within California’s definition of a continuing-care promise, as a promise to provide care “for the duration of [the resident’s] life.” Id. § 1771(a)(10). If Continuing Life abandons this obligation, it opens itself to fines, and its managers to prison. See id. § 1793.5(d) and (e). And, more importantly from our perspective, it would not receive any Deferred Fees. The Commissioner doesn’t dispute that Continuing Life has to provide lifelong care to its residents, but argues that the Residence Agreement’s schedule fixing the amount of Deferred Fees that Continuing Life earns each year is what really fixes its right to those fees. From his perspective, it is the passage of time and not the provision of services that entitles Continuing Life to the Deferred Fees. We disagree. The Deferred Fee schedule fixes only the Deferred Fee amount: Section 12.5 of the Residence Agreement states that “[if] this Agreement is terminated [by voluntary termination or death of the resident], you or your estate shall pay [Continuing Life] a [Deferred Fee] according to the following schedule.” Section 12.5 lacks any language that would oblige a resident or the trustee to pay that amount at the time the amount is fixed. There is no language in the Agreement that characterizes the date on which the amount that a Deferred Fee is fixed as the date it is earned. There is likewise nothing in the Residence Agreement that makes a resident liable to pay any part of the Deferred Fee when the amount is fixed. By the terms of the Agreement the resident pays the Deferred Fee only when he departs or dies. If we say that Continuing Life’s right to income is fixed with the Deferred Fee schedule, then any obligations after the Deferred Fee amount is maxed out would end up being “ministerial” or “formalities”. Hallmark Cards, Inc., 90 T.C. at 32–33. And this turns on the question of whether the lifetime care obligation is the “essential service that Continuing Life provides.” Identifying a contract’s essential object is not a new problem. Courts have seen it before in cases arising from the timing of payments held in escrow accounts or similar arrangements. In Iler v. Commissioner, 37 T.C.M. (CCH) 783 (1978), the taxpayer was a contractor. He made a deal with the Kentucky highway department that entitled him to periodic payments, but with a portion of the agreed price held in a “retainage” account in the taxpayer’s name and administered by a bank acting as custodian. Id. at 784. The Commissioner argued that this made it income to the taxpayer when the 24 [*24] bank received it. But we held that a contractual provision that allowed the department to seize the account for nonperformance up until it finally accepted the project was such a “substantial condition” that it meant deposit of a fixed percentage in an account with the taxpayer’s name on it did not amount to receipt. Id. at 786. Iler featured a taxpayer who used cash accounting, so we focused there on whether the retainage was constructively received. But a district court used the same reasoning for an accrual taxpayer in Southern Family Insurance Co. v. United States, 733 F. Supp. 2d 1290 (M.D. Fla. 2010). In that case the taxpayer was an insurance company that accepted Florida’s offer to write property-casualty policies for homeowners who otherwise would have been relegated to a state-run joint underwriting pool. The incentive was a bonus payment that Florida paid into an escrow account each year—but that the insurer could withdraw only if its policies remained in effect for three years and a state audit confirmed that they had. Id. at 1291–92. The IRS argued that under accrual principles, this meant that the insurance company had to report as taxable income the money put into the escrow account in the year of deposit, when the amount of the bonus payment was fixed. Id. at 1292. The district court disagreed. It began by quoting the regulation with which we began this section—“in the case of compensation for services, no determination can be made as to the right to such compensation or the amount thereof until the services are completed, the amount of compensation is ordinarily income for the taxable year in which the determination can be made.” Id. at 1295. It then recited the numerous contingencies that the taxpayer had to meet before it could win release of the escrowed funds: “Simply put, no determination could be made as to Southern Family’s right to the takeout bonuses or the amount thereof until after the three-year escrow period and the completion of the audits.” Id. at 1297. We cannot ignore Continuing Life’s continuing obligation to provide lifetime care under California state law. As we said in Hallmark, 90 T.C. at 34, “[t]he fact that . . . petitioner knows with absolute certainty that in the next instant these rights will arise cannot compensate for the fact that . . . they do not exist.” We think that argument is even stronger here. Continuing Life may know the exact amount of Deferred Fees, but it hasn’t yet earned them. 25 [*25] The Commissioner’s backup argument is that Continuing Life’s obligation to provide services is only a “condition subsequent” and not a “condition precedent” to its receipt of the Deferred Fees. The distinction is an easy one to state: A condition precedent is one that must be met before a fixed right to income arises, while a condition subsequent ends an existing right to income but does not preclude the accrual of income. Keith v. Commissioner, 115 T.C. 605, 617 (2000); Charles Schwab Corp. v. Commissioner, 107 T.C. 282, 293 (1996), aff’d, 161 F.3d 1231 (9th Cir. 1988); Harkins, 81 T.C.M. (CCH) at 1550. The distinction can be important because, although a condition subsequent may take away the right to receive income, one would ignore it for purposes of the all-events test. Keith, 115 T.C. at 617. These definitions state the consequences of their characterization but don’t help much in figuring out whether a particular condition is “precedent” or “subsequent”. For that we have to go to caselaw. In Harkins, 81 T.C.M. (CCH) at 1548, the taxpayer ran a movie theater and had a contract with Pepsi to advertise and market Pepsi’s products. The contract between the taxpayer and Pepsi provided that the taxpayer had to meet certain marketing obligations during separate six-month periods; and if he met those obligations, then Pepsi would pay for that period within 60 days after its end. Id. at 1551. If, however, he failed to meet these obligations during those 60 days, then he forfeited the payment amount. Id. at 1552. We held that the continuing obligation during those 60 days was a condition subsequent because the payment was fully earned at the end of the six-month period and “not contingent on Pepsi’s investigating the theater company’s obligation.” Id. at 1551. In Keith, 115 T.C. at 607, the taxpayer was in the business of financing, selling, and renting real estate. We held that an accrualmethod taxpayer needed to recognize the entire sale price on the date a sales contract was executed. Id. at 618–19. The taxpayer didn’t receive the full amount, but we noted that the buyer’s obligation to pay the full purchase price was unconditional and the taxpayer's right to that amount was fixed on the date of execution. Id. The possibility that the buyer would default was a condition subsequent because it didn’t affect the taxpayer’s right to the sale price. Id. at 617. Some cases teach us that a condition is a condition subsequent when the occurrence of that condition has no effect on whether the right to income was fixed or earned. Id.; Harkins, 81 T.C.M. (CCH) at 1550. We can ignore conditions subsequent in deciding whether there is a fixed right to payment. For example, in Harkins, if we ignored the 60-day 26 [*26] period, and in Keith, if we ignored anything beyond the contract execution date, the taxpayer would still have had a fixed right to payment. In contrast, the condition precedent in these cases was a condition necessary for any right to payment to even exist. See Keith, 115 T.C. at 618; Harkins, 81 T.C.M. (CCH) at 1551. We couldn’t ignore the event which marked the completion of that condition. With that in mind, we turn to the parties’ arguments. Continuing Life argues that its lifetime care obligation is the condition precedent to receiving a Deferred Fee. It argues that the completion of the lifetime care obligation is the earliest that it could possibly recognize a Deferred Fee as income. See Schlude, 372 U.S. at 136; Johnson, 108 T.C. 448; Harkins, 81 T.C.M. (CCH) at 1550. It reiterates that the Residence Agreement and California state law require Continuing Life to provide care for the life of a resident for it to have a fixed right to a fixed amount of a Deferred Fee. The Commissioner again argues that it is the contractual payment schedule that fixes Continuing Life’s entitlement to the Deferred Fees. He implicitly argues that the yearly incremental additions to those Deferred Fees is the condition precedent, and that any later events such as a breach of the agreement by Continuing Life are conditions subsequent. Continuing Life, he says, has a fixed right to the entire Deferred Fee that is fixed in exchange for providing care for only those first four years. This explains that the “essential service” which Continuing Life provides is that first four years of care. Sure, Continuing Life has to keep providing services after that for residents that don’t leave if it wants eventually to get paid the Deferred Fees, but why couldn’t the indefiniteness of that term for the provision of care make it a condition subsequent? He’d analogize the situation to the realestate company that has to recognize income on execution of the sales contract, but might have to forfeit that income if it doesn’t deliver good title at closing. There are a few answers. The first is again to rely on the text of the regulation, which speaks not of conditions precedent and subsequent but more plainly says that, in cases where a taxpayer’s right to compensation for services requires that those services be completed, “the amount of compensation is ordinarily income for the taxable year in which the determination can be made.” Treas. Reg. § 1.451-1(a). (The 27 [*27] “determination” here would be that Continuing Life has fulfilled its obligation to provide lifetime care.)17 A second answer is that this argument requires us to split Continuing Life’s single lifetime commitment to its residents into two parts: the first being the initial four years, and the second being the indefinite remainder of a resident’s stay. The Commissioner doesn’t explain why we should do this, nor does he point us to any cases where what would seem to be a single legal obligation can be split in two to enable part to be classified as a condition precedent and part to be classified as a condition subsequent. See Gen. Dynamics Corp. & Subs. v. Commissioner, 74 T.C.M. (CCH) 632, 651 (1997) (rejecting argument that taxpayer divide single long-term contract into four parts). The Commissioner distinguishes these cases by observing that the probability that Continuing Life will not uphold its end of the deal after four years and the probability that it will not be paid are both very low. He argues that caselaw teaches that a key characteristic of conditions subsequent is that the probability that they might occur is also very low. Perhaps low probability is the mark of a condition subsequent. In Charles Schwab Corp., 107 T.C. at 286, we had to distinguish between the “settlement date” and the “trade date” in securities trading. The “trade date” was the date the day the trade was executed, while the “settlement date” occurred only after the taxpayer performed certain functions such as recording, figuration, confirmation, comparison, and booking. Id. at 286–87. We held that the “trade date” was the condition precedent and the “settlement date” was a condition subsequent because all functions after the “trade date” were ministerial acts. Id. at 293–94. The “trade date” was the essential service; the taxpayer determined the purchase price and commission amount on the trade date, and the taxpayer’s client couldn’t cancel the trade after the order was submitted on the trade date. Id. at 292. The “settlement date” was a condition subsequent because the possibility that the trade was going to be canceled was “too indefinite or contingent for accrual.” Id. at 294. We did not, however, focus exclusively on the low probability 17 The attentive reader will remember that Continuing Life does not argue that it recognizes Deferred Fees as income only when a resident departs or dies, but amortizes it over each resident’s expected life, with any unamortized amount recognized in the year of departure. See supra p.13. That situation is governed by a sentence one can find a little later in the same regulation: “Where an amount of income is properly accrued on the basis of a reasonable estimate and the exact amount is subsequently determined, the difference, if any, shall be taken into account for the taxable year in which such determination is made.” Treas. Reg. § 1.451-1(a). 28 [*28] that something would go haywire between the trade and settlement dates. We focused more on when the right to dividend income accrued. That right turned on the trade date, and we described the remoteness of cancellation after the trade date only as proof that executing the trade was the essential service that Schwab provided. Id. at 292. We do think that the Commissioner is right that there’s no genuine dispute that the probability that Continuing Life will not in the end receive the Deferred Fees is low. Continuing Life terminated only one Residence Agreement during the year at issue. It had eleven residents move out and fifteen who died, and it collected the Deferred Fees from them. While we do agree that these facts show that it is unlikely that Continuing Life won’t receive a Deferred Fee from a resident, we have to hold that a possibility’s remoteness does not itself create a condition subsequent. We also found a very old case that is very close to this one on this point. In 1924, long before the dawn of antibiotics and the modern welfare state, a man named Victor Gauss lost his wife and three children to tuberculosis. His only surviving child was his son William. William was severely afflicted—“William was at that time 32 years old, but had the mentality of a 3-year old child. He weighed 89 pounds, had been losing weight, would not eat, and required forced feeding.” Norbury Sanatorium Co. v. Commissioner, 9 T.C. 586, 587 (1947). Victor himself was 69 and, as we found, “was anxious to make some arrangement by which the proper care of William might be assured during William’s lifetime, even though he (Victor) should predecease him and leave a negligible estate. William was his only remaining obligation.” Id. Victor found a private hospital that said it could provide his son with lifetime care. He promised to pay a monthly fee and to leave in trust for the sanatorium a portfolio of bonds worth $28,000. 18 The sanatorium would get the entire corpus of the trust in exchange for providing William all “necessary medical attention and render such other services as might reasonably be expected . . . as long as the said William Gauss lives.” Id. at 588. The contract went on to provide that the sanatorium would lose its right to this corpus if it “should mistreat 18 This is the equivalent of about $460,000 today. Inflation Calculator, US Inflation Calculator, https://www.usinflationcalculator.com/ (last visited Feb. 28, 2022). 29 [*29] the said William Gauss, or should willfully neglect him.” at 588–89. Id. Victor’s wishes for his son seem to have been fulfilled. He died in 1931; his son lived until 1944 and never left the care of the sanatorium. Periodic inspections showed his son as well cared for as the medical science of the time allowed. The sanatorium, all agreed, was entitled to what was left in the trust when William died. We concluded back in 1947 that the initial settlement of the trust did not make the sanatorium its beneficiary. William was the real beneficiary of the trust; it was for the purpose of obtaining proper care for him during his lifetime and after the death of his father, that the trust was created. As compensation for its services rendered to the trust in caring for William, [the taxpayer] was to receive the current trust income; and as additional compensation and an inducement to petitioner to comply faithfully with its undertaking . . . was to receive the trust corpus. Id. at 594. We found then that the requirement that the sanatorium provide lifetime care meant that it had to complete “its undertaking to properly care for William during his lifetime” before it recognized the trust corpus as income. Id. We therefore rejected the argument that the sanatorium received the trust’s corpus subject to being divested of its beneficial ownership upon the occurrence of a condition subsequent, viz., its failure to furnish proper care to William during William’s lifetime. Id. at 593. Standards of care for the infirm change as the decades pass. The Master Trust is much longer than the three pages or so that Mr. Gauss drafted a century ago. State regulation has taken over the tasks that used to be provided by third-party inspectors sent by a private trustee. But the human inclination to care for the infirmities of old age or debilitating illness while one can still do so remains, and the sums people set aside and the web of promises that they spin show us that it is the open-ended provision of those services that is essential to such contracts. 19 19 The Commissioner also argues that the conditional language “if” and “should” in a contract create a condition subsequent. The case that he cites, Harkins, 30 [*30] The text of section 446 and its regulation mean that Continuing Life’s adherence to GAAP in its treatment of the Deferred Fees clearly reflected income. The text of section 451 and its regulation mean that Continuing Life’s promise to provide lifetime care also means that it did not have to recognize income when the amount of those Fees was fixed by the passage of time. If we could stop here we would hold that on the undisputed facts of this case Continuing Life was not entitled to the Deferred Fees, although fixed in amount by the Residence Agreement after four years, until it had finished its job of providing care for its residents. II. Deferred Fees and the Purpose of Tax Accounting Thor states that the “vastly different objectives” of financial and tax accounting mean that “any presumptive equivalency between [them] would be unacceptable.” Thor, 439 U.S. at 542–43. While financial accounting looks to “provide useful information,” tax accounting is concerned with the “equitable collection of revenue.” Id. at 542. The Second Circuit identified the key distinction: “Tax accounting therefore tends to compute taxable income on the basis of the taxpayer’s present ability to pay the tax, as manifested by his current cash flow, without regard to deductions that may later accrue.” RCA Corp. v. United States, 664 F.2d 881, 888 (2d Cir. 1981). Our decision in Straight v. Commissioner, 74 T.C.M. (CCH) 1457 (1997), shows that we sometimes rely on this difference between the purposes of financial and tax accounting to decide cases. In Straight, the taxpayer produced two accounting experts who testified that its accounting method complied with GAAP and AICPA guidance, and matched revenues with expenses. Id. at 1464. The Commissioner didn’t dispute those claims or produce any experts on this topic, and we accepted those conclusions as truth. Id. Despite that, we ruled in favor of the Commissioner, stating GAAP compliance didn’t matter because tax and financial accounting have different objectives. See id. (citing Thor, 439 U.S. at 540–44). These differences simply aren’t present here. Continuing Life residents pay their Contribution Amount to Cummins as Trustee. Cummins has the authority to make interest-free loans to Continuing 81 T.C.M. (CCH) at 1551, doesn’t say this. Although the contract in Harkins contains the conditional language “if” and “should”, we held that the conditional part of the contract was a condition subsequent because the right to income had already accrued; we didn’t place any weight on the conditional language. Id. 31 [*31] Life, but at no point does Continuing Life have any “control” over these funds as income, and it does not have the “ability to pay” taxes with these loans—it has to use them to improve the community’s capital plant. We also don’t see how GAAP’s treatment of Deferred Fees in Position 90–8 would allow Continuing Life’s management to get looseygoosey with their inclusion into Continuing Life’s income. Position 90– 8 requires actuarial determinations of lifespans for individuals whose age is known. The population involved is relatively small and actuarial determination of life expectancy has a long history and is as objective as any numbers relating to a human population can be. Seeing neither imprecision in their computation nor any ability to pay tax out of Deferred Fees that remain in the hands of a third-party trustee, we see no way in which Continuing Life’s reporting of its income from Deferred Fees contradicts the purpose of tax accounting’s treatment of income recognition. III. Deferred Fees and Caselaw We’ve discussed Continuing Life’s accounting for Deferred Fees and how it complies with the Code and regulations, as well as how the differing objectives of financial and tax accounting should not cause us to question the normal rules of statutory interpretation that support Continuing Life’s position. But that may not yet be enough—unusually for a question of tax law, there are lines of precedent that are sired by the Code but barely acknowledge their parentage before reasoning analogically in a common-law fashion. The closest set of facts to ours is in Highland Farms, where we began our analysis with an acknowledgment that section 446(a) tells taxpayers to compute their taxable income using their regular accounting method, that the taxpayer used the approved accrual method of accounting and “kept its books regularly in accordance with this method.” Highland Farms, 106 T.C. at 250. But we immediately noted that this was not enough, and analyzed the facts in light of caselaw that analyzed deposits and advance payments. See id. at 251. Highland Farms is precedential, so we must follow or distinguish it. We also agree with the parties that it is the only other case on the books that discusses the tax treatment of income to a taxpayer who owned a continuing-care community. The key issue was the appropriate tax treatment of “entry fees” paid by residents and kept in a segregated 32 [*32] account owned by Highland Farms. Much like the Deferred Fees here, the entry fee that a particular resident owed was fixed in amount by the passage of time. See id. at 244. According to its contracts, Highland Farms was entitled to 20% of an entry fee at the end of each of a resident’s first five years. At the end of each year, it would move that portion of a resident’s fee from the segregated account to its general account and move it in its books from the “advance deposit” line to “income”. The Commissioner viewed all the entry fees as prepaid rent and wanted to tax them in the year of receipt. But we held that, as long as a resident could leave the community and demand repayment of the unearned portion of the entry fees, Highland Farms “had ‘no unfettered “dominion” over the money at the time of receipt.’” Id. at 252. And then we held that “[o]nly the nonrefundable or nonforfeitable amounts each year constitute income.” Id. The Commissioner reasonably sees similarities to Continuing Life’s situation—it too charges residents a big upfront payment and becomes eligible for a percentage of that payment over the first few years of residence (although, as we’ve stressed, only if it upholds its part of the deal by providing continuing care for the rest of a resident’s life if necessary). He asks us to be astonished at his moderation in demanding taxation only of that fixed amount of the Deferred Fees year by year, and not (as he argued in Highland Farms) all at once. Continuing Life demurs. It argues that there is a crucial difference here—only Cummins, the Trustee, has “dominion” over the Deferred Fees until they are paid out. This, it argues, is of decisive importance because it makes the other advance-payment cases that we relied on in Highland Farms distinguishable. Who’s right? We can begin with the trio of advance-payment cases that accounting aficionados all know—Schlude, American Automobile Association, and Automobile Club of Michigan. These cases all dealt with the treatment of prepaid income and whether a taxpayer that used accrual accounting could defer recognition of that income. The Supreme Court rejected deferral in all these cases, and held that the taxpayer’s accounting was “artificial” in that the advance payments were related to services performed only upon the customers’ demand without relation to fixed dates. Schlude, 372 U.S. at 135; Am. Auto. Ass’n, 367 U.S. at 694; 33 [*33] Auto. Club of Mich., 353 U.S. at 189. For example, in Schlude, the Supreme Court denied deferral of income from prepaid dance lessons that had to be taken during a designated period but not on a fixed schedule. See Schlude, 372 U.S. at 130. These accounting systems didn’t clearly reflect income, the Court held, and so the Commissioner didn’t abuse his discretion in requiring these taxpayers to recognize the whole amount of prepaid income upon receipt. Schlude, 372 U.S. at 136; Am. Auto. Ass’n, 367 U.S. at 698; Auto. Club of Mich., 353 U.S. at 189– 90. We distinguished these cases in Highland Farms because Highland Farms’ residents could control the amount of refunded entry fees—if they left within the first few years, they got partial refunds. If they didn’t leave, they didn’t get refunds. See Highland Farms, 106 T.C. at 244. We could thus analogize Highland Farms to cases where a utility company demanded deposits from its customers—customers who controlled whether the utility could ever take the deposits as its own by either keeping current onF their electric bills or building good enough credit that the utility no longer needed to have a deposit on hand. Indianapolis Power, 493 U.S. at 210–11; see also Kan. City S. Indus., Inc. v. Commissioner, 98 T.C. 242, 262 (1992) (taxpayer did not have sufficient rights in the deposits for the deposits to be taxable income upon receipt because the customer controlled whether his deposit would be refunded); Oak Indus., Inc. & Subs. v. Commissioner, 96 T.C. 559, 571–72 (1991) (customer deposit with taxpayer for any future unpaid fees, equipment damage, and so forth not includible in income); Houston Indus., Inc. & Subs. v. United States, 32 Fed. Cl. 202, 212 (1994) (taxpayer’s obligation to repay to its customers all overrecoveries received precludes the receipts’ inclusions in income), aff’d, 125 F.3d 1442 (Fed. Cir. 1997). But we don’t think that any of these analogies fits here, because there can be no genuine dispute that Continuing Life didn’t have “dominion” over any of the Deferred Fees. Unlike utilities, lessors, or Highland Farms, Continuing Life did not get the Deferred Fees in its hands subject to an obligation to refund them. It simply didn’t get the Deferred Fees at all until the Trustee paid them over, and it didn’t get the right to those fees until it had fulfilled its promise to provide lifetime care to its residents. Even this distinction, however, isn’t the end of the argument. There is another line of cases that analyze what might be income to an accrual taxpayer when money is paid to a trustee or escrow agent. In 34 [*34] Angelus Funeral Home v. Commissioner, 47 T.C. 391, 392 (1967), aff’d, 407 F.2d 210 (9th Cir. 1969), the taxpayer sold “pre-need” funeral services for an upfront payment and with small monthly payments until the total outstanding balance was paid. One of the contracts we looked at provided that the total amounts paid would be held in an irrevocable trust, and deposited in a bank, trust company, or savings-and-loan association. Angelus couldn’t withdraw any amount from this account until it fully performed its services, when it then earned what had already been paid. Id. at 392–93. At some point, Angelus changed this contract to increase its control over the funds. Id. at 393. The new contract provided that Angelus could deposit the paid amounts anywhere and into multiple accounts, and that Angelus could withdraw these amounts to use as collateral or to pay for capital improvements or real property. Id. In exchange for this, Angelus paid its customers 10% of the total annual payments made under the contract. We distinguished the earlier and later versions of the contract. We held that Angelus did not recognize any income under the first version of its contract because it was a true trustee and had no rights to the money. Id. at 395. But we also held that Angelus did recognize income upon receipt under the second version because it didn’t impose any restraint or limitations on Angelus’s ability to use the funds. Id. at 398. This made it look like an advance-payment case. Id. at 399 (citing Schlude; American Automobile Association, and Automobile Club of Michigan). 20 In Miele v. Commissioner, 72 T.C. 284, 289–90 (1979), the taxpayer was a lawyer who held his clients’ funds in trust in a segregated account. The issue was when he had to recognize these funds as income. Id. at 288. We held that he did not need to when he received the funds because they were still the clients’ even though held in trust. Id. at 290. We disagreed with the Commissioner’s characterization of them as advance payments for future services. Id. at 289. However, once he performed services, they became his income even though the funds were still held in a segregated trust account. Id. at 290–91. We found that he constructively received them because under the all-events test he had a right to them after he had performed his services. Id. 20 The Ninth Circuit noted the unusual lack of reliance on any Code section in our analysis: “[T]he Commissioner did not rely, and he does not now rely, on the power given him under 26 U.S.C. § 446(b) . . . Nor does Angelus assert that its ‘method of accounting’ (26 U.S.C. § 446(a)) is correct and should therefore be followed. In short, this case is not an accounting case.” Angelus, 407 F.2d at 212. 35 [*35] Continuing Life likewise never had dominion over or an entitlement to the Deferred Fees. It never had dominion over them because Cummins held these amounts in trust for the residents. Although Cummins was authorized to make interest-free loans to Continuing Life, it was obligated to repay those loans to the Trust when a resident died or moved out. Continuing Life did not take its Deferred Fee from the loan, but instead it paid any loans back to Cummins, and then Cummins paid the Deferred Fee to Continuing Life. In this situation, we can’t say that Continuing Life had any rights to the Deferred Fees when it borrowed from the Trust. But there is also a line of cases that might favor the Commissioner. It begins with Commissioner v. Hansen, 360 U.S. 446 (1959). In Hansen, the taxpayer sold cars and lent its customers the money to pay. Id. at 448. It got this money from a financing company. After selling a car, it sold the loan to the financing company and guaranteed repayment of the loan. Id. The contract between the taxpayer and financing company provided that the financing company would pay the taxpayer a large percentage of the purchase price, but would withhold a portion as security for the taxpayer’s performance of its obligation to guarantee payments from the car buyers. Id. Everyone agreed that the portion of the purchase price that the taxpayer received right away was income, but the taxpayer did not want to recognize as income the portion that the financing company didn’t have to pay over until the car buyer paid off the loan. Id. at 449. The Commissioner argued that the taxpayer should recognize the whole amount of the finance company’s purchase price, including the portion retained as security. Id. The Supreme Court ruled in favor of the Commissioner, finding that the taxpayer had a fixed right to the full purchase price. Id. at 466. It didn’t matter that the taxpayer might not actually receive the money for many years. The key fact was that on sale of its cars it had a fixed right to the payment. Id. at 466–67. That fixed right might in the end mature into cash or into repayments under the guaranty that the taxpayer had made to the financing company. Id. at 465. But in either of these scenarios, the car’s full purchase price was for the taxpayer’s benefit. See also Johnson, 108 T.C. at 481 (similar analysis of portion of cars’ purchase price held by third-party in reserve to guarantee vehicle service contracts). Stendig, 843 F.2d 163, and Bolling v. Commissioner, 357 F.2d 3 (8th Cir. 1966), are very similar to Hansen. Both cases featured home 36 [*36] builders that had deals with third parties that financed its deals. Stendig, 843 F.2d at 163; Bolling, 357 F.2d at 5. Both home builders had to deposit a percentage of either rent, Stendig, 843 F.2d at 164; or the sale price of a home, Bolling, 357 F.2d at 5, in segregated accounts to ensure their performance on guarantee obligations to the third party. And both taxpayers ended up losing to the IRS on the question of whether those segregated amounts—which they might well not get in hand for many years—were current income. See Stendig, 843 F.2d at 165–66; Bolling, 357 F.2d at 6. As in Hansen, the key point was that the money would either eventually be received by the taxpayer or be paid to the third party in fulfillment of the taxpayer’s obligation to that third party. Cases like these are not perfectly analogous to Continuing Life’s. The key distinction that we see is that there is no equivalent to the continuing-care promise that the taxpayers in those cases owed to their customers. In Hansen and Bolling, once the taxpayer sold a car or house to a customer, it owed him no further duty. It did owe a duty to a financing company—namely, guaranty of the customer’s payments on the car or home—but it had a fixed right to the money that its customer had paid, whether that right took the form of cash to be received in the future or payment of its own guaranty to a third party. In Stendig, the taxpayer may have had some kind of future obligation to its tenants (depending on whether the leases were monthly or for some longer term), but the rent it received was monthly and was exchanged for a month’s tenancy, and not an open-ended promise of care. Its right to a particular month’s rent was likewise a fixed right. We therefore hold that, even if we reasoned by analogy to precedent without recourse to the text of the Code or regulations, Continuing Life’s accounting for the Deferred Fees was correct. IV. The Commissioner’s Discretion We’ve decided that the text of the Code and regulations, tax accounting principles, and caselaw are on the side of Continuing Life. But the Commissioner still has one exceptionally strong argument: on questions of tax accounting, there is solid precedent that says we must uphold his determination unless we find it an abuse of discretion. We begin again with Thor. In the course of rejecting Thor’s argument that compliance with GAAP establishes a presumption that an accounting method clearly reflects income, the Court cited Treasury 37 [*37] Regulation § 1.446-1(a)(2). Thor, 439 U.S. at 540. This regulation includes the remarkable sentence: “However, no method of accounting is acceptable unless, in the opinion of the Commissioner, it clearly reflects income.” Federal courts have recently been reminded that we are to interpret regulations using “all the ‘traditional tools’ of construction.” Kisor v. Wilkie, 139 S. Ct. 2400, 2415 (2019) (quoting Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 843 n.9 (1984)). If we just applied the ordinary plain meaning of “opinion”, then this case (and nearly all others in this dark corner of tax law) would become easy—the Commissioner wins with proof of what’s in the notice of deficiency or FPAA. Look at the FPAA here. It’s the Commissioner’s opinion. It says Continuing Life’s method of accounting doesn’t clearly reflect income. That means it’s not acceptable. This is where things again get puzzling. Section 446 does require deference to the opinion of the Commissioner, but makes that deference conditional: “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.” 21 At least one other court has noticed the contradiction between the unconditional statement in the regulation and the plainly conditional requirement of the Code. In Mulholland v. United States, 28 Fed. Cl. 320, 335 (1993), aff’d without published opinion, 22 F.3d 1105 (Fed. Cir. 1994), the Court of Federal Claims took a close look at the Code and regulation: [T]he statute does not provide that the decision—whether the taxpayer’s income is clearly reflected—shall be measured only by “the opinion of the Secretary,” as does the regulation. Instead, we read it to merely grant the Commissioner/Secretary the discretion to make his determination as to whether reported income is clearly reflected, but does not preclude the court, at trial, from making its own de novo determination as to whether income is clearly reflected as reported. 21 Section 7701(a)(11)(B) defines “Secretary” to include not only the Secretary of the Treasury but also his delegates, who include the Commissioner and IRS employees. 38 [*38] (Emphases in original.) See also Hewlett-Packard Co. & Subs. v. United States, 71 F.3d 398, 402–03 (Fed. Cir. 1995) (similar “opinion of the Secretary” language in section 471(a) would, if taken literally, “make the Commissioner’s exercise of discretion in this area virtually unreviewable, no matter how serious the legal or factual errors upon which it rested”). We do note that neither party questioned the validity of this regulation with the usual reference to step one of Chevron. But we also don’t think we can overturn decades of precedent in this area by applying this regulation according to its plain terms—its conflict with the language of the Code is too plain. We can take some comfort in this conclusion because there don’t seem to be any cases in which a holding depends on deference to the Commissioner’s “opinion”, even though there are opinions where the sentence is cited as an additional ground to defer to the government in questions of whether an accounting method clearly reflects income. See, e.g., Van Raden v. Commissioner, 71 T.C. 1083, 1119 (1979) (Chabot, J., dissenting), aff’d, 650 F.2d 1046 (9th Cir. 1981). There is also some good authority that perhaps we should read “opinion” to mean something similar to “discretion”. The Supreme Court in Thor seemed to conflate the two words. The same paragraph which quotes regulation § 1.446-1(a)(2) concludes with the Court’s saying that “if the Commissioner, in the exercise of his discretion, determines that [GAAP does not pass muster as a clear reflection of income], he may prescribe a different practice without having to rebut any presumption running against the Treasury.” Thor, 439 U.S. at 540 (emphasis added). “Discretion” and its possible abuse are very familiar to administrative lawyers. Congress routinely delegates functions to executive agencies, and those agencies exercise discretion in performing those functions. Disgruntled persons may seek judicial review, and that review is aimed to uncover “abuses of discretion.” This term itself is well defined, and courts know that they have to look at an agency’s findings and conclusions to see if they are arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with the law. See 5 U.S.C. § 706(2)(A); Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 41 (1983); Fargo v. Commissioner, 87 T.C.M. (CCH) 815, 817 (2004), aff’d, 447 F.3d 706 (9th Cir. 2006). And when courts look to see if an agency abused its discretion, they look at the whole record or parts of it cited by a party. 5 U.S.C. § 706. A court asked to decide if an agency has abused its discretion must usually review how the agency 39 [*39] exercised its discretion on the basis of the administrative record compiled by the agency. We review only the rationale that the agency uses, and don’t come up with one of our own or let the Commissioner’s attorneys come up with one of their own. 22 This is how our own Court reviews whistleblower awards, 23 and, at least in cases appealable to certain circuits, notices of determination in collection-due-process cases. See, e.g., LG Kendrick, LLC v. Commissioner, 146 T.C. 17, 35 (2016), aff’d, 684 F. App’x 744 (10th Cir. 2017); Jones v. Commissioner, 104 T.C.M. (CCH) 364 (2012); see also Keller v. Commissioner, 568 F.3d 710, 718 (9th Cir. 2009); Murphy v. Commissioner, 469 F.3d 27, 31 (1st Cir. 2006); Robinette v. Commissioner, 439 F.3d 455 (8th Cir. 2006). It is also exceedingly common in the judicial review of other agencies’ work. See Alfred C. Aman Jr. & William T. Mayton, Administrative Law 437 (3d ed. 2014); see also United States v. Carlo Bianchi & Co., 373 U.S. 709 (1963). But that is not at all how we review exercises of the Commissioner’s discretion in contesting or changing a taxpayer’s method of accounting. Over the decades that we’ve been reviewing the Commissioner’s work we never ask for an administrative record, and we don’t confine ourselves to the rationale given by the IRS at the end of an audit. We don’t even ask whether there’s some particular clearly erroneous factfinding or mistake of law or irrational application of law to facts. We instead treat abuse of discretion as a heightened standard of review, and use phrases like “the taxpayer bears a heavy burden of proof,” or “we do not interfere unless the Commissioner’s determination is arbitrary, capricious, clearly unlawful, or without sound basis in fact or law.” Ewing v. Commissioner, 122 T.C. 32, 39–40 (2004) (collecting authorities), vacated, 439 F.3d 1009 (9th Cir. 2006). The Ninth Circuit, to which any decision in this case is presumptively appealable, treats our conclusions about whether the Commissioner has abused his discretion in changing an accounting 22 This is the Chenery doctrine, an administrative-law principle that says “a reviewing court, in dealing with a determination or judgment which an administrative agency alone is authorized to make, must judge the propriety of such action solely by the grounds invoked by the agency.” SEC v. Chenery Corp., 332 U.S. 194, 196 (1947). 23 In Kasper v. Commissioner, 150 T.C. 8, 23 (2018), we explained that due to the Chenery doctrine, we can uphold the IRS Whistleblower Office’s (WBO) determination only on the grounds it actually relied on when making its determination. This means that the WBO must clearly set the grounds on which it made its determination, so that we do not have to guess. 40 [*40] method as a question of fact. “In reviewing the Tax Court’s finding, we can reverse only if its determination is ‘clearly erroneous.’” Sandor v. Commissioner, 536 F.2d 874, 875 (9th Cir. 1976); see also Cole v. Commissioner, 586 F.2d 747, 749 (9th Cir. 1978). Other circuits disagree. The Sixth Circuit characterizes the question as one of “ultimate fact” and so reviewable de novo. See Ford Motor Co. v. Commissioner, 71 F.3d 209, 212 (6th Cir. 1995). The Second Circuit takes a third approach—it has held that the issue in cases involving whether an accounting method clearly reflects income is “not whether [an] accounting method adequately reflected income, but whether the Commissioner abused his discretion in determining that it did not. The latter question is one of law.” RCA Corp., 664 F.2d at 889. On the other hand, the Eighth Circuit found that the question of whether “a particular method of accounting resulted in a clear reflection of income is a conclusion of law, or at least a mixed question of law and fact, subject to be de novo review.” Wal-Mart Stores, Inc. & Subs. v. Commissioner, 153 F.3d 650, 657 (8th Cir. 1998). We are deciding the parties’ respective summary-judgment motions, so this particular divergence of analyses is not quite present here. As we analyzed the problem in the first section above, we ask if there is any genuine dispute that Continuing Life’s accounting for Deferred Fees clearly reflected income. We concluded that it did not, supra p. 28, but must admit that we didn’t do so on the basis of any administrative record. We have to be frank that if we were to decide these motions without reference to the interplay of GAAP compliance and the text of the Code and regulations, the purpose of the Code’s rules on tax accounting, or to analogous caselaw—if we were in other words to judge purely on the reasonableness of the Commissioner’s exercise of discretion in this case on a blank slate—we would be hard pressed to say without a trial that either the Commissioner or Continuing Life was unreasonable. And, if those cases that give the Commissioner considerable discretion in this area were pushed to their extreme, the Commissioner would win. But what is the source of this discretion that is so widely acknowledged to exist? Here things get curiouser and curiouser. We return for a last time to Thor. The Court there noted that deference to a taxpayer’s choice of accounting method is limited to cases “where the Commissioner believes that the accounts clearly reflect the net income.” Thor, 439 U.S. at 541 (quoting Lucas v. Am. Code Co., 280 U.S. 445, 449 41 [*41] (1930)). And that a taxpayer who wants to overcome the Commissioner’s rejection of his accounting method must show that the rejection was “plainly arbitrary.” Id. at 533 (quoting Lucas v. Kan. City Structural Steel Co., 281 U.S. 264, 271 (1930)). We ourselves observed in RLC that these two cases from 1930 were the earliest mentions of the Commissioner’s discretionary power. See RLC, 98 T.C. at 491. What makes this curious is that those old cases did not say that the Commissioner had this discretion, but rather that “much latitude for discretion is thus given to the administrative board charged with the duty of enforcing the act.” Lucas v. Am. Code Co., 280 U.S. at 449. That “administrative board” was our predecessor, the Board of Tax Appeals. Even by the late ‘20s, the BTA was independent of the Bureau of Internal Revenue (the IRS’s old name). 24 And the Supreme Court of that era specifically disclaimed any need to defer to the Commissioner when the Board itself had spoken: “[T]here is no reason for thinking that Congress considered the Commissioner to be better qualified for making determinations under section 327 and 328 [sections calling for valuation of mixed classes of property under a long-repealed excess-profits tax from 1919].” Williamsport Wire Rope Co. v. United States, 277 U.S. 551, 565 (1928). A close reading of the cases from that era shows that they were not deferring to the Commissioner, they were deferring to us. In the landmark case of Dobson v. Commissioner, 320 U.S. 489, 505 (1943), the Supreme Court referred to the “mischief of overruling the Tax Court in matters of tax accounting.” And that “whatever latitude exists in resolving questions such as those of proper accounting . . . exists in the Tax Court and not in the regular courts; when the court cannot separate the elements of a decision so as to identify a clear-cut mistake of law, the decision of the Tax Court must stand.” Id. at 501–02. In a particularly flattering passage, the Court explained the basis for its deference: It deals with a subject that is highly specialized and so complex as to be the despair of judges. It is relatively better staffed for its task than is the judiciary. Its members not infrequently bring to their task long legislative or 24 The Revenue Act of 1924, ch. 234, § 900(a), (k), 43 Stat. 253, 336, 338, established the Board of Tax Appeals to permit taxpayers to challenge determinations made by the IRS. In 1942 Congress passed the Revenue Act of 1942, ch. 619, § 504(a), 56 Stat. 798, 957, which renamed the Board the “Tax Court of the United States.” 42 [*42] administrative experience in their subject. . . . Individual cases are disposed of wholly on records publicly made, in adversary proceedings, and the court has no responsibility for previous handling. Tested by every theoretical and practical reason for administrative finality, no administrative decisions are entitled to higher credit in the courts. Id. at 498–99. The persuasiveness of Justice Jackson’s prose caught Congress’s attention. In 1948 it amended section 7482 to state that “[t]he United States Courts of Appeals . . . shall have exclusive jurisdiction to review the decisions of the Tax Court . . . in the same manner and to the same extent as decisions of the district courts in civil actions tried without a jury,” 25 and Dobson deference was no more. Sic semper transit gloria mundi. The cases that Justice Jackson collated and explained in Dobson live on, however, in a peculiar way. They described why it made sense to defer to a body with specialized expertise that developed a record and reasonably explained the result it reached. Once Congress decided to treat us as more of a court than an administrative agency, we lost any deference to our exercise of discretion in matters of tax accounting. But that deference to agency expertise did not disappear. It came to rest— through decades of citations to these old cases—with the Commissioner. The Commissioner, however, does not have to explain why he disagrees with a taxpayer’s method of accounting, and he does not have to justify that disagreement with an administrative record. He just has to issue a notice of deficiency. See QinetiQ US Holdings, Inc. & Subs. v. Commissioner, 845 F.3d 555, 559–60 (4th Cir. 2017), aff’g 110 T.C.M. (CCH) 17 (2015). And so we have the peculiarity of discretion in the Commissioner to change accounting methods that courts can review for abuse of discretion in a de novo deficiency proceeding unbound and unjustified by any record that the Commissioner prepares. This evolution is beyond our power as a trial court to change. The law is settled that the Commissioner has discretion to change a taxpayer’s accounting method. But in this case we must reach a decision 25 See Act of June 25, 1948, ch. 646, § 36, 62 Stat. 869, 991 (amending 26 U.S.C. § 1141(a)). 43 [*43] as to whether Continuing Life’s accounting of the Deferred Fees clearly reflected income. The voluminous law in this area directs us to the text of the Code and regulations, to the purpose of distinctions between tax and financial accounting, to the exercise of common-law reasoning by analogy to treat similar cases similarly, and finally to defer in some fashion to the determination of the Commissioner. V. Conclusion Textualism would lead us to hold for Continuing Life. The purpose of these sections of the Code and regulations, and more broadly, the purpose of distinguishing tax and financial accounting is in no way in conflict with the conclusion that a textual analysis leads us to. And Continuing Life’s accounting for the Deferred Fees fits snugly into the pattern of similar cases. That leaves deference to the Commissioner as the best argument for ruling in his favor. That he has discretion to change accounting methods is undoubtedly true; but the history of how that discretion came to be weakens its power to overcome text, purpose, and analogy. We will grant Continuing Life’s motion for summary judgment and deny the Commissioner’s motion. An appropriate order and decision will be entered. --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/agency%3Atax-court%3Ab9e38ef349629577. Public record. Not legal advice.