22

are taxed only once between the two. We could reason that the amounts should continue to be taxed only once, even though the family is no longer intact, because of the clear and direct relationship of the payments to the former legal relationship of the parties (or the continuing legal relationship, in the case of a paternity payment to support a child after a divorce or otherwise outside of marriage). These payments would not have been made but for the prior legal relationship. Unless the government wishes to affirmatively discourage divorce via the tax laws (a very unwise possibility, in my view), there does not seem to be a persuasive reason to tax such payments more onerously outside marriage (by taxing them twice) than within it (by taxing them once.) Viewed this way, the question is not whether the amount conceptually constitutes "income" to both but rather who should be taxed on what is concededly income to someone? That is to say, this question can be viewed not as a "what is income?" question" but rather a “to whom should income be taxed?" question. The income could be taxed to the recipient by requiring the recipient to include it and allowing the payor to deduct it. Conversely, it could be taxed to the payor by allowing the recipient to exclude it and disallowing the payor a deduction. The government, in this scenario, becomes a mere stakeholder in the matter: The amounts will be taxed to someone (once); the only question is which taxpayer's marginal rates should apply?

The Gould opinion obliquely picked up on this perspective when it stressed that “[t]he net income of the divorced husband subject to taxation was not decreased by payment of alimony under the court's order; and, on the other hand, the sum received by the wife on account thereof cannot be regarded as income arising or accruing to her within the enactment."23 The thought was written as a single sentence, with a semi-colon connecting the two independent clauses. This form seems to imply a causal link, i.e., that another reason why the amount was not includable by Mrs. Gould was precisely because it was not deductible by Mr. Gould. The Court's implication was that the amount should not be taxed to both, and since the Court cannot create nonstatutory deductions, it exercised its power to define the contours of "gross income" to ensure that the amount was not taxed twice by holding that the recipient need not include the payments in gross income. The Court would presumably have come to the same conclusion with respect to child support, if any had been at issue.

24

deduct. See I.R.C. §§ 67, 165(c)(3) & (h), and 213. Finally, filing separately allows the couple to avoid joint and several liability. See I.R.C. § 6013(d)(3).

22 This conforms the treatment of support payments within the intact family to the treatment of gifts within a family, which are also excludable by the recipient and not deductible by the donor. See I.R.C. §§ 102, 162. Mandatory and legally enforceable support payments would not otherwise qualify as "gifts," which are generally defined for income tax purposes as transfers made out of “detached and disinterested generosity." See Comm'r v. Duberstein, 363 U.S. 278 (1960).

23

24

245 U.S. at 154.

In order to take a deduction, a taxpayer must find a Code section that contains the words "there shall be allowed as a deduction" and satisfy each of the requirements specified there. He must also survive the gauntlet of provisions that take away "otherwise allowable deductions." No court, the Supreme Court included, has any power to create a deduction. Only

With respect to support payments, the one-tax approach has continued to this day. The issues here have been how to decide between the two whom to tax and how to differentiate support payments (alimony and child support) from property settlements, a question that first arose with the 1942 legislation.

B. From 1942 to 1984

1. Support Payments

In the midst of World War II, the highest federal income tax marginal rate under the regular tax and a special surtax exceeded 90% in order to finance the war effort. Under such a rate structure, a payor of nondeductible alimony of any significant size could have little left upon which to live once he paid his alimony and then the tax on his income, including the alimony. The legislative history underlying the Revenue Act of 194226 reflected this concern.

25

The existing law does not tax alimony payments to the wife who receives them,
nor does it allow the husband to take any deduction on account of alimony
payments made by him. He is fully taxable on his entire net income even though
a large portion of his income goes to his wife as alimony or as separate
maintenance payments. The increased surtax rates would intensify this hardship
and in many cases the husband would not have sufficient income left after paying
alimony to meet his income tax obligations.

Congress can do that. On the other hand, because present-day I.R.C. § 61 contains the circular definition that gross income means "income from whatever source derived," the Supreme Court has long exercised its power to construe the contours of “gross income" by creating common law regarding what constitutes gross income. See, e.g., the cases developing the assignment-ofincome doctrine (discussed infra notes 252-301 and accompanying text). See generally Deborah A. Geier, Some Meandering Thoughts on Plaintiffs and Their Attorneys' Fees and Costs, 88 TAX NOTES 531 (2000) (discussing the interplay between the Court's power to craft the contours of gross income and its inability to create deductions in the context of the treatment of attorneys fees and costs that, though deductible, are subject to deduction restrictions that could be avoided if the plaintiffs could exclude from gross income the portion of taxable litigation awards to the extent paid to their attorneys).

25 We all know that payors of alimony can include both males and females. See, e.g., Liesl Schillinger, Divorce Him, Support Him?, THE INDEPENDENT (LONDON), Dec. 14, 1997, at 5 (characterizing "the phenomenon of men seeking alimony" as "becoming rife in the United States" and, more objectively, reporting that “[t]he number of male clients in the US claiming financial support from high-earning exes has doubled in the last five years"). While I know that use of the word "he" as the payor of alimony can perpetuate stereotypes, I think that repeated usage of the terms "he or she" makes for distracting prose, and thus I have chosen to risk the alienation of some readers in the quest for less stilted-sounding sentences.

26

Revenue Act of 1942, Pub. L. No. 753, § 120, 56 Stat. 816 (1942) (codified as amended at I.R.C. §§ 22(k) and 23(u) (1942)).

The bill would correct this situation by taxing alimony and separate maintenance
payments to the wife receiving them, and by relieving the husband from tax upon
that portion of such payments which constitutes income to him under present law.
This treatment is provided only in cases of divorce or legal separation and applies
only where the alimony or separate maintenance obligation is discharged in
periodic payments. Moreover, the portion of such payments going to the support
of minor children of the husband does not constitute income to the wife nor a
deduction to the husband. The same is true with regard to payments in discharge
of lump sum obligations, even though made in installments.

27

Therefore, as described briefly above, the Revenue Act of 1942 switched the answer to the question, "who should be taxed on the amounts paid out as alimony?," from the payor to the recipient by requiring inclusion of alimony by the recipient and allowing a deduction to the payor. But the Act and its legislative history were quite clear that this inclusion/deduction system should not apply to child support or amounts paid to compensate for the transfer of a property interest. Child support and property settlements, in other words, remained excludable by the recipient and nondeductible by the payor, as under pre-1942 law. I'll discuss child support first and then property settlements.

The legislative history is silent with respect to why Congress chose to shift the tax incidence to the recipient only with respect to alimony and not child support. Yet, we can perhaps guess the predominant thinking of the time. In this more traditional era--when it was, indeed, only husbands who paid alimony and child support28--it might have been thought that, while shifting the tax obligation with respect to alimony would help to alleviate the immediate and quite practical problem of a husband being unable to satisfy both his alimony and tax liability if alimony were not deductible, shifting the tax obligation with respect to the man's children would be going "too far." A man's financial obligation to his children--including the obligation to pay the income tax on amounts spent to support them--might have been considered to be stronger than that to his ex-wife.

In any event, this new distinction between alimony and child support necessarily required line-drawing for the first time. The statute did so by providing that the recipient's gross income inclusion shall not apply to that part of any such periodic payment which the terms of the decree or written instrument fix, in terms of an amount of money or a portion of the payment, as a sum which is payable for the support of minor children of such husband. In case any such periodic payment is less than the amount specified in the decree or written instrument, for the purpose of applying the preceding sentence, such payment, to the extent of such sum payable for such support, shall be considered a payment for such support.

27

29

H.R. REP. NO. 77-2333, reprinted in 1942-2 C.B. 372, 409.

In other words, an amount "fixed" in the governing documents for child support would not fall under the inclusion/deduction system, and if a payor made only a partial payment, the child support payment would be considered paid first. For example, assume that a husband was ordered to pay $10,000 per year to his ex-wife for support of herself and their minor children, $6,000 of which was explicitly fixed as child support. If the husband paid only $8,000 in a year, $6,000 would be considered nondeductible child support and $2,000 would be considered deductible alimony (so long as the remaining qualifications for deductible "alimony," described shortly below, were met).

What does it mean for an amount to be "fixed" for child support within the meaning of the statute? Suppose, for example, that John and Mary had three minor children when they divorced, that their mutually negotiated divorce agreement (which the relevant court approved) provided for a "family support payment" of, say $12,000 per year, but that the payment would be reduced by one-sixth each time a minor child married, became emancipated, or died. In other words, by the time all three minor children married, became emancipated, or died, the payments would have been reduced to $6,000. Was any amount "fixed" for child support during the years of the children's minority in this agreement?

On the one hand, the agreement did not overtly fix any amount for child support in the literal sense, though the "substance" of the document seems to indicate that, prior to the time any of the three children married, became emancipated, or died, $6,000 of the payments was really disguised child support. The reduction in amounts payable upon conditions relating to the children's emancipation, etc., was, after all, captured in language in the agreement itself, and thus the agreement could be said to have adequately "fixed" $6,000 of the payments as child support.

Much litigation ensued in order to determine what the term "fix" meant in this provision--
an example of how the different labels and their different tax treatments inevitably cause
litigation.30 The split among the circuit courts finally had to go to the Supreme Court for
resolution in Commissioner v. Lester," the facts of which are essentially identical to those
recited above. The Court quoted both the Senate Finance Report as well as a report of the
Office of the Legislative Counsel to the Senate Finance Committee in concluding that no
part of an unallocated "family support payment" should be considered "fixed" as child
support, even if a portion is scheduled to be reduced on the happening of events related to
the children.

"If, however, the periodic payments ... are received by the wife for
the support and maintenance of herself and minor children of the
husband without such specific designation of the portion for the
support of such children, then the whole of such amounts is

30 See, e.g., Metcalf v. Comm'r, 271 F.2d 288 (1st Cir. 1959); Eisinger v. Comm'r, 250 F.2d 303 (9th Cir. 1957); Lester v. Comm'r, 32 T.C. 1156 (1959), rev'd, 279 F.2d 354 (2d Cir. 1960).

31 366 U.S. 299 (1961).

includable in the income of the wife as provided in section 22(k)

." S. Rep. No. 1631, 77th Cong., 2d Sess. 86.

As finally enacted in 1942, the Congress used the word "fix" instead of the term "specifically designated," but the change was explained in the Senate hearings as "a little more streamlined language." Hearings before the Senate Committee on Finance on H.R. 7387, 77th Cong., 2d Sess. 43. As the Office of the Legislative Counsel reported to the Senate Committee:

"If an amount is specified in the decree of divorce attributable to
the support of minor children, that amount is not income to the
wife .... If, however, that amount paid the wife includes the
support of children, but no amount is specified for the support of
minor children, the entire amount goes into the income of the wife
" Ibid.

This language leaves no room for doubt. The agreement must expressly "fix" a sum certain or percentage of the payment for child support before any of the payment is excluded from the wife's income. The statutory requirement is strict and carefully worded. It does not say that "a sufficiently clear purpose" on the part of the parties is sufficient to shift the tax. It says that the "written instrument" must "fix" that "portion of the payment" which is to go to the support of the children. Otherwise, the wife must pay tax on the whole payment. We are obliged to enforce this mandate of the Congress.

32

Therefore, the entire $12,000 annual payment during the children's years of minority was includable by the wife and deductible by the husband as alimony.

Thus, a divorce decree that required a husband to pay $100 weekly for the support of his ex-wife and $50 for the support of their two children would produce only a $100 weekly income inclusion for the wife and a corresponding deduction for the husband, since the remaining $50 would be considered "fixed" for child support and not eligible for the inclusion/deduction system. If the decree provided instead for a payment of $150 per week for the support of the exwife and the children, no part of it would be considered excludable/nondeductible child support. Similarly, none of the $150 payment would be considered child support if the decree also provided that the payments would be reduced by $25 per week upon the death, marriage, or majority of either of the children.

32 Id. at 302-03. I've always wondered whether Justice Scalia would have written a dissent in Lester, since the majority relies so explicitly on legislative history, a tool that Justice Scalia finds illegitimate. See generally Deborah A. Geier, Textualism and Tax Cases, 66 TEMP. L. REV. 445, 447-55 (1993); Wisconsin Pub. Intervenor v. Mortier, 111 S. Ct. 2476, 2487 (1991) (Scalia, J., concurring in the judgment) (ours is “a Government of laws not of committee reports"). If Justice Scalia would interpret the word "fix" broadly enough to cover terms explicitly requiring reduction of payments on the happening of an event linked to the child's leaving the payee's home (such as marriage, death, or emancipation), then the legislative history to the contrary would have been meaningless to him.