MEMORANDUM OPINIONKROUPA, Judge: This case is one of seven pending affected item proceedings involving separate allocation of section 6662 accuracy-related penalties under section 6015(c) (separate liability allocation). 2 The taxpayers in each of the pending cases were investors in Hoyt cattle partnerships subject to the provisions of the Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA), Pub. L. 97-248, sec. 402, 96 Stat. 648.
The parties in each of the pending cases have agreed that this case and Andrews v. Commissioner, T.C. Memo 2010-230, also filed today, will be used to present the penalty allocation issue to the Court. The parties chose Andrews and Malsom because they represent two distinct scenarios. The requesting spouse in this case is the spouse who earned the higher income during the years at issue while the requesting spouse in Andrews earned the lower income. Nevertheless, the parties agree that the same computational methodology should be used in all cases for consistency. We will therefore apply the methodology established in Andrews to this case.
The parties filed a Stipulation of Settled Issues at trial that resolved all issues except the computation of petitioner's separate liability allocation. The parties agree that Dennis Malsom (petitioner) and his late spouse (Mrs. Malsom) are liable for the accuracy-related penalties for taxable years 1989 through 1996 (the years at issue), and they agree that petitioner is entitled to the separate liability allocation for each year. They also agree that the computational methodology established in Estate of Capehart v. Commissioner, 125 T.C. 211 (2005) (Capehart Estate methodology), should be applied to determine petitioner's separate liability allocation. There is no dispute as to the liability for each year at issue. The parties disagree, however, on how the liabilities and the penalties should be allocated when there are multiple sets of computational adjustments to petitioner's liability for each year. There are two sets of computational adjustments for each year because petitioner and his late spouse were partners in multiple TEFRA partnerships, the Tier 1 and Tier 2 partnerships, during the years at issue. 3 The Tier 2 partnerships were also partners in the Tier 1 partnership. Petitioners reported partnership losses from both the Tier 1 and Tier 2 partnerships on the joint returns for the years at issue.
The parties' disagreement focuses on the timing of the computations. We dealt with this same issue in Andrews. Respondent's proposed computations make the Tier 1 and Tier 2 computational adjustments for each year in two separate steps. In contrast, petitioner's proposed computations make the Tier 1 and Tier 2 computational adjustments in one step. Applying the computations in one step as petitioner proposes results in one deficiency and one penalty for each of the years at issue. Applying the computations in two steps as respondent proposes results in two separate sets of deficiencies and penalties for each year. The allocated amounts are also different when the computations are made in two steps rather than together in one step. We agree with respondent that the computations for each TEFRA partnership must be made separately before the liabilities and penalties are allocated under the Capehart Estate methodology.
I. Separate Liability Allocation and Capehart Estate MethodologyWe begin with an overview of separate liability allocation. Generally, taxpayers filing joint Federal income tax returns are jointly and severally liable for all taxes due. See sec. 6013(d)(3). A spouse (requesting spouse) may elect to have the liability limited to his or her proportionate share of the liability, however, if the spouses are divorced, legally separated, or living apart for the 12 months preceding the election. Sec. 6015(c)(1), (3)(A)(i). The separate liability allocation must be made no later than two years after the Secretary has begun collection activities with respect to the electing spouse. Sec. 6015(c)(3)(B).
Liability allocation is based on the items that gave rise to the deficiency (erroneous items). Erroneous items are allocated to each spouse as though each had filed a separate return for the taxable year, subject to some exceptions. Sec. 6015(d)(3)(A). Erroneous items are reallocated under the tax benefit exception to the extent one spouse received a tax benefit on the joint return and the other spouse did not. Sec. 6015(d)(3)(B). The requesting spouse's proportionate share of the deficiency is based on his or her proportionate share of the erroneous items. See sec. 6015(d)(1), (3)(A); sec. 1.6015-3(d)(4)(i)(A), Income Tax Regs.
The requesting spouse's allocable share of the accuracy-related penalty is based on his or her allocable share of the underpayment. Sec. 1.6015-3(d)(4)(iv)(B), Income Tax Regs. The underpayment is reduced, however, by the requesting spouse's share of excess withholding credits, estimated tax payments, and other payments that were frozen by the Commissioner instead of being refunded to the taxpayer. See sec. 6664 (a); sec. 1.6664-2(a), (d), Income Tax Regs. The requesting spouse's share of the underpayment is then multiplied by 20 percent to determine his or her allocable share of the accuracy-related penalty. Capehart Estate, supra at 225-226.
II. Applying the Computational AdjustmentsWe now turn to the parties' proposed computations of petitioner's separate liability allocation. Petitioner claims that the amount of the liability should be the same regardless of whether the computational adjustments are made in one step or multiple steps. He analogizes the assessment of computational adjustments in TEFRA proceedings to the computation of tax deficiencies to support his claim. We note, however, that normal deficiency procedures do not apply to computational adjustments. Sec. 6230(a)(1); see sec. 6230(a)(2)(A). We do not find petitioner's analogy helpful.
Petitioner's proposed computations also ignore the partnership procedures set forth in TEFRA. See secs. 6221 through 6231. The TEFRA rules provide for separate proceedings at the partner and partnership levels. See GAF Corp. & Subs. v. Commissioner, 114 T.C. 519, 524 (2000). The computational adjustments at issue stem from separate partner and partnership level proceedings. The Tier 1 computational adjustments are partnership-level adjustments and the Tier 2 computational adjustments are partner-level adjustments of Tier 1 partnership items. Accordingly, the TEFRA rules require that the Tier 1 partnership-level computational adjustments be made separate from the Tier 2 partner-level computational adjustments.
Petitioner's proposed computations also ignore the 1-year limitation on making computational adjustments. See sec. 6229 (a), (d). Decisions were finalized in the Tier 1 partnership proceedings in 2005, or almost a year before the decisions were finalized in the Tier 2 partnership proceedings. The Tier 1 computational adjustments must therefore be made before the Tier 2 computational adjustments to comply with the Tier 1 limitations period.
III. Applying the Capehart Estate MethodologyWe now turn to the application of the Capehart Estate methodology to make the separate liability allocation. We have already determined that the Tier 1 and Tier 2 computations for each taxable year should be made separately and the Tier 1 computations should be made first. The Capehart Estate methodology should also be applied separately to each set of computations. This means that the tax benefit exception should be applied twice. The tax benefit exception should be applied once at the Tier 1 partnership level and again at the Tier 2 partnership level.
Petitioner's proposed computations, however, apply the tax benefit exception only to the Tier 1 erroneous items and ignore the tax benefits petitioner received from the Tier 2 Hoyt income and partnership losses. We find that respondent's proposed computations properly apply the tax benefit exception separately to the Tier 1 and Tier 2 erroneous items.
IV. Allocation of the PenaltiesWe now turn to the allocation of the penalties. We find that the accuracy-related penalties, which are based on the Tier 1 and Tier 2 underpayments, must also be separately allocated for each set of computations. We further find that respondent's computations properly account for frozen refunds before separately allocating the Tier 1 and Tier 2 penalties.
In reaching our holding, we have considered all arguments made, and, to the extent not mentioned, we conclude that they are moot, irrelevant, or without merit.
An appropriate order will be issued.