equal-amount inclusion. Only the rate-bracket differential between the parties, if any, results in a revenue loss, and this loss is self-limiting, as the greater the amount paid to the payee, the higher the tax bracket that will apply to it, until further income-shifting would not produce a revenue loss. Moreover, in the context of a payor in a significantly higher tax bracket than the payee--which is the very context where it would seem that income-shifting would be at its most extreme and therefore result in the most lost revenue--any divorce “bonus” is more illusory than real. This is because that combination is just the combination that enjoys a significant “marriage bonus” under current law if the couple files a joint return,10 and this marriage bonus is lost on the divorce. Even with income-shifting, the parties are not likely better off taxwise by much, if any, in the aggregate under the more onerous schedule for single filers that must apply to them after the divorce. As mentioned above, our current system for taxing transfers in divorce is the result of history and political compromise as much as grand theory. Part I below will recount that history, since it's difficult to understand how we got to where we are today without an understanding of it. It will also necessarily introduce the various ways to think about these transfers, as it's difficult to discuss how the law evolved without an introduction to these thoughts at the same time. Part II will continue the discussion of what works in the current system, what is fundamentally flawed, and what should be done about it and why. Part III will examine two inextricably related problems that need addressing. 9 Malman, supra note 4, at 410-12. 10 Joint filing can generate a “marriage penalty on two-earner couples," with the married couple paying more in tax than they would if they were able to file separately under the rate schedule for unmarried individuals. The marriage penalty is at its peak when each spouse earns the same amount of income. On the other hand, joint filing can generate a “marriage bonus,” with the joint-return tax lower than would arise under separate filing using the individual rate schedule, if one spouse earns significantly more income than the other. See generally Lawrence Zelenak, Marriage and the Income Tax, 67 S. CAL. L. REV. 339 (1994) (recounting this history and advocating abandonment of the joint return in favor of individual filing for all). More married couples obtain a benefit from the marriage bonus than are subject to the marriage penalty. About 51% of married couples enjoy a marriage bonus, whereas about 42% of married couples incur a marriage penalty. In 1996, those enjoying a marriage bonus paid, as a group, $33 billion less in taxes than they would have paid if single, while those incurring the marriage penalty paid, as a group, $29 billion more in taxes than they would have paid if single. Thus, $4 billion of net revenue that would have been collected by the Treasury if all individuals filed separately was lost. See For Better Or Worse: Marriage and the Federal Income Tax, CONGRESSIONAL BUDGET OFFICE (June 1997). I. The History A. Prior to 1942 The early statute did not specifically address how to treat cash payments (or in-kind property transfers) between divorcing spouses. Hence, it devolved upon the Supreme Court to determine the status of such payments and transfers. The Court tackled cash payments way back in 1917 in Gould v. Gould11 and in-kind property transfers in 1962 in United States v. Davis,12 discussed in Part B.ii., below. 12 The parties in Gould divorced in 1909, and the court ordered Mr. Gould to pay Mrs. Gould $3,000 each month for the rest of her life "for her support and maintenance.' The issue was whether Mrs. Gould must include these payments in her gross income. The Supreme Court first quoted the predecessor to § 61 as follows: B. That, subject only to such exemptions and deductions as are hereinafter 14 The Court then wrote four paragraphs, which comprised the opinion's entire reasoning. Because of their brevity, I quote them in full. In the interpretation of statutes levying taxes it is the established rule not to extend Joseph M. Dodge, J. Clifton Fleming, Jr., & Deborah A. Geier, FEDERAL INCOME TAX: DOCTRINE, STRUCTURE, AND POLICY 191 n.11 (2d ed. 1999). The situation in which incomeshifting in divorce under an inclusion/deduction system can lose the most revenue is when a high-earning spouse makes payments to a low-earning spouse after divorce, which is precisely the situation where the divorced couple will lose their prior marriage bonus, since after the divorce they must file using the rate schedule for individuals in order for the payments to be eligible for the inclusion/deduction system that generates the income-shifting. See I.R.C. § 71(e). to enlarge their operations so as to embrace matters not specifically pointed out. As appears from the above quotations, the new income upon which subdivision 1 In Audubon v. Shufeldt ..., we said: "Alimony does not arise from any business The net income of the divorced husband subject to taxation was not decreased by 15 As made clear in the first quoted paragraph, one major ground relied upon by the Court was a rule of statutory interpretation borrowed from tariff law that revenue statutes are to be narrowly construed. No such canon of statutory interpretation survives today with respect to the income tax. Moving to the exegesis of the terms of the statute itself, the Court seemed to conclude that, because alimony is not similar to the items specifically listed in the statute as taxable, alimony must not have been contemplated by Congress as coming within the ambit of the provision. This canon of statutory construction sometimes goes by the Latin name of ejusdem generis, meaning "of the same kind." Precisely the same reasoning, accepted by the Gould Court, was again used by a taxpayer in 1955 in arguing (in part) that punitive damages are not includable in gross income because they are not like the items specifically listed as taxable. That time, the taxpayer lost in the seminal case of Commissioner v. Glenshaw Glass.10 Rejecting the argument premised on the notion of ejusdem generis, the Glenshaw Glass Court instead stressed the catchall language at the end--“income derived from any source whatever"--in concluding that 16 punitive damages were includable in gross income by the recipient. The 1955 Court stated that this catchall language required inclusion of all "undeniable accessions to wealth, clearly realized, and over which the taxpayers have complete dominion."17 (Indeed, it's an interesting exercise to think about whether the Gould Court might have decided differently if the case had arisen in the later era, after Glenshaw Glass.) Notice that this form of analysis focuses on whether alimony comes within some notion of "income." If the analysis is viewed this way, each party is analyzed independently, and a payment of alimony could conceivably be taxable to both the payor and the recipient. For example, if we view the matter from the recipient's side alone, and if alimony is considered as coming within the Glenshaw Glass notion of "income" as an undeniable accession to wealth, etc., then it would be includable by the recipient. At the same time, the payor earning wages from which the alimony was paid would have to include the wages in gross income, since compensation for services rendered is specifically listed as "income" in § 61(a)(1). Moreover, the payor would arguably be denied a deduction for the payment under a strict definition of "income" in the familiar Schanz-Haig-Simons sense, under which only outlays incurred to produce includable income are properly deductible (with personal consumption outlays being nondeductible, and thus taxed). Since the payment of alimony is not an outlay incurred directly to produce the payor's wages, it would be nondeductible (and thus remain in the tax base of the payor). Thus, the payment would be taxed twice, much as amounts paid by an employee from 17 18 Id. at 431. 18 In general, a tax on "income" reaches amounts saved and amounts spent on personal consumption. Amounts that are saved are taxed since the Internal Revenue Code generally disallows outlays constituting "capital expenditures," such as the purchase of an asset. See I.R.C. § 263. That is to say, outlays that result merely in a change in the form in which wealth is held, rather than a diminution in wealth, are not deductible. Current "expenses" are the opposite of a "capital expenditure" in that current expenses represent a real diminution in wealth in the year spent. If that expense is incurred in income-producing activity, then the expense is generally deductible. See, e.g., I.R.C. §§ 162 and 212. If, on the other hand, the expense is spent on personal consumption, it is generally nondeductible. See I.R.C. § 162(a). The only personal consumption expenses that are deductible are those that Congress has specifically allowed, usually as an incentive to engage in certain desirable behavior. See, e.g., I.R.C. §§ 170 (charitable contribution deduction), 163(h)(3) (home mortgage interest deduction). The reason why expenses incurred to produce includable income (as opposed to expenses incurred to buy personal consumption) should be deductible under an income tax is to avoid double taxation of the same dollars to the same taxpayer. You can think of it in the following way: If income-producing expenses were not allowed as deductions, then those expenses would create basis (previously taxed dollars). That basis should offset any includable income produced by that outlay, which would result in inclusion of only the "net" receipt in income. But the tax system doesn't work this way. Instead, I.R.C. § 61 requires the inclusion of every dollar of "gross" receipts. In order, therefore, to prevent the double taxation of those gross receipts to the extent of the outlays incurred to produce them, those outlays must be deductible. See generally Dodge, Fleming, & Geier, supra note 10, at 30-33 and 39-55. his wages to his housecleaner for cleaning his house are taxed twice (once to the employee and once to the housecleaner). In the housecleaner example, the amounts constitute wages to both, and the payment to the housecleaner would clearly be a nondeductible personal expense of the payor. The only means by which to differentiate the alimony payment from the housecleaner payment would be to argue that alimony does not really purchase any personal consumption for the payor and thus should be removed from his tax base via a deduction. This means of analysis would define personal consumption not by looking to whether the payment directly contributed to income production of some kind--which defines personal consumption by default"--but rather by looking to what is actually purchased with the payment and determining whether it affirmatively qualifies as "personal consumption." 20 Perhaps we don't have to resolve that definitional dilemma, however, for there is another way to view this issue, which was also hinted at by the Gould Court and which is, I think, the more appropriate way to think about the payment. Rather than analyzing the tax consequences to each of these taxpayers independently of the other, i.e., determining whether the receipt qualifies as “income” to the recipient and whether the payment qualifies as a deductible one to the payor under an “income" analysis because it does not purchase discretionary personal consumption, we could view both taxpayers together. In an intact marriage, by analogy, amounts earned by one spouse and paid to another are ignored for tax purposes (i.e., they are neither includable by the recipient nor deductible by the payor), whether or not the couple file a joint return or file separate returns using the rate schedule for married couples filing separately.21 Therefore, the amounts 19 This is the method currently used in the Internal Revenue Code. Amounts can be deducted only if they are specifically described in a Code section containing the words, "there shall be allowed as a deduction ...," and, as described supra note 18, most such provisions allow deduction only for income-producing expenses. 20 One example of this mode of analysis would be to argue that certain types of mandatory payments should not be considered as purchasing personal consumption and thus should be deductible. Payments of certain state and local taxes can be analyzed this way, as can alimony and child support. See generally Joseph M. Dodge, THE LOGIC OF TAX 123 (1989). 21 See I.R.C. § 1 (describing the various filing statuses and their related rate schedules). Some couples file separately in order to lower significant state income taxes that use a progressive rate schedule. Filing separately might allow the couple to use the lowest state rate brackets twice, which can more than offset the increased Federal tax burden that may arise from using the more onerous schedule (compared to the rate schedule applicable to single taxpayers) applicable to married couples filing separately. Moreover, if each member of the couple earns approximately half of the couple's aggregate income, the aggregate federal tax should be roughly the same, regardless of whether they file a joint return or file separately, since the joint return mechanism produces a tax "exactly twice the tax that would be due if the [rate schedule for married filing separately] were applied to 50% of the couple's aggregate taxable income." Dodge, Fleming, & Geier, supra note 10, at 188. Another reason that some married couples file separately is to enable one to deduct more medical expenses, personal casualty losses, or miscellaneous itemized deductions, all of which are deductible only to the extent that they exceed certain percentages of adjusted gross income, than he or she would otherwise be able to |