Marriage Bonus

A marriage bonus is the amount by which the tax on a couple's joint return falls short of what the two spouses would have paid had they not been married. Treasury's Office of Tax Analysis states it as the mirror of the penalty: "a couple has a marriage bonus if they owe less income tax filing a joint return than the spouses would pay if they were unmarried and each were taxable as a single or head of household filer."

Treasury's explanation of the cause has aged well, unlike the one it gives for penalties: "Marriage bonuses generally arise because joint filing effectively allows couples to average their incomes, which can reduce the tax rate, and therefore the tax, on the income of the higher-earning spouse." That is still exactly how it works, and it works because the joint rate bands are double the unmarried bands through the top of the 32 percent band. The same doubling that removed the rate-table marriage penalty for most couples is what creates the bonus for couples with unequal incomes: one structure, two outcomes, separated only by how the income is split between the spouses.

Like "marriage penalty," the phrase is not a defined term in the Internal Revenue Code. Section 1(f)'s heading names only the penalty. The bodies that use "marriage bonus" analytically are Treasury's Office of Tax Analysis, the Congressional Budget Office and the Joint Committee on Taxation.

The bonus was designed in, and it predates the penalty by three years. Treasury's paper sets out the history, and it explains why the two effects exist at all. Before 1948 there was a single rate schedule for everyone, so in most cases there was neither a penalty nor a bonus: spouses could file separately and pay what they would have paid unmarried. The exceptions were the ones that created the pressure for reform. In community property states all earned income and income from property accumulated during the marriage was treated as split evenly between the spouses, and elsewhere property income could be shifted between spouses by gift, trust, family partnership and other means, so two couples with identical combined income could owe very different amounts depending on where they lived and how their income was arranged.

Congress's answer, effective in 1948, was a separate rate schedule for joint filers "with rate brackets that were double the width of the rate brackets for single filers." Treasury describes the result precisely: "There were marriage bonuses, but no marriage penalties, under this rate structure since a couple would pay no more tax (and in many circumstances less tax) filing jointly than the spouses would pay if they each filed a separate return." Penalties arrived afterwards, and from changes made for other reasons: the head of household status created by the Revenue Act of 1951, and the broadening of the single brackets legislated in 1969 and effective in 1971, each of which narrowed the gap in favor of unmarried filers and in doing so imposed penalties on some couples. So the bonus is the original feature and the penalty is the side effect.

Beyond the rate tables, four further bonuses are worth naming, because none of them depends on the spouses having unequal incomes.

The standard deduction. Section 63(c)(2)(A) sets the joint basic standard deduction at "200 percent of the dollar amount in effect under subparagraph (C)," which is the amount for an unmarried filer. So a spouse with no income whatsoever brings a full unmarried-sized deduction to the return. For 2026 that is $16,100 of shelter arriving with a person who earned nothing, and because the doubling is a formula rather than a pair of numbers it cannot drift apart.

The unlimited marital deduction on transfers. Property passing between spouses is generally free of gift and estate tax without limit, which is a benefit no unmarried couple can replicate at any income level. It has one important exception: it does not apply to a spouse who is not a United States citizen, which is the reason a separate and much larger annual gift exclusion exists for gifts to a non-citizen spouse, $194,000 for 2026, under section 2523(i)(2). The mechanics of transfer tax between spouses belong to the estate tax page.

Portability of the unused estate and gift exclusion. Sections 2010(c)(4) and (c)(5) let a surviving spouse claim the portion of a deceased spouse's basic exclusion amount that went unused, effectively giving a married couple one combined allowance rather than requiring each to use their own. It is election-dependent rather than automatic, which is its main practical trap. Note that there is no equivalent for the generation-skipping transfer tax: unused GST exemption dies with the transferor.

The spousal IRA. A spouse with little or no earned income can still make an IRA contribution based on the working spouse's compensation, which is not available to an unmarried person with no earnings at all.

The principal-residence exclusion has an asymmetry inside it that is a bonus in its own right. Section 121(b)(2)(A) doubles the exclusion to $500,000 for a joint return, and its clause (i) requires only that either spouse meet the ownership requirement, while both must meet the use requirement. So a spouse who moves into a home the other already owned, and lives there for the required period, can be counted toward the doubled exclusion without ever having been on the title.

What cannot honestly be said about size. There is no current, reachable measurement of how many couples get a bonus and how large it typically is. Treasury's distributional tables describe 1999 law and the structure has changed materially since, so quoting a share or an average from them would be quoting a different tax system. What can be said structurally, and is enough to be useful, is that unequal incomes produce a bonus and similar high incomes produce a penalty, and that since most couples do not have closely matched high incomes the bonus is the more common outcome.

Joint filing applies one rate table to the couple's combined taxable income. Where the incomes are unequal, part of the higher earner's income is taxed in bands the lower earner did not use, at rates below the ones it would have faced on a return of its own. That is the entire mechanism, and it scales with the gap between the two incomes rather than with their total.

Three cases, in order of how large the bonus is.

One earner, one non-earner: the largest bonus available. The whole of the non-earner's share of the doubled bands, plus a full unmarried-sized standard deduction, is applied to the earner's income.

Very unequal earners: a large bonus. The lower earner uses part of their bands and the higher earner's income takes up the remainder.

Two similar incomes: no bonus, and possibly a penalty. Each spouse fills their own half of the doubled bands, so nothing is absorbed, and the provisions that were never doubled start to bind. That case is the subject of the marriage penalty page.

A hypothetical built to isolate what the bonus actually tracks. Hold the household's income constant and vary only the split. Suppose a couple has $160,000 of combined taxable income, and consider three ways it could be divided. In all three, their joint liability is identical, because the joint table is applied to the combined figure and does not know or care who earned it. What changes is the comparison.

Split evenly, $80,000 each. On the unmarried table each spouse's income runs past the top of the 12 percent band, $50,400, and stops inside the 22 percent band, whose ceiling is $105,700. Neither reaches 24 percent, and neither has unused bands for the other to borrow. Marrying changes almost nothing from the rate tables.

Split $130,000 and $30,000. Now the higher earner alone runs past $105,700 and into the 24 percent band, while the lower earner leaves most of their own 12 and 22 percent bands unused. Filed jointly, the combined $160,000 sits below the joint 22 percent ceiling, $211,400, so none of it reaches 24 percent at all. The income that would have been taxed at 24 percent on one return is taxed at 22 percent on the joint one. That difference is the bonus.

Split $160,000 and nothing. The same effect, larger, because there are no bands used by the second spouse at all.

So the household's income never changed and its joint tax never changed. The bonus appeared, grew and vanished purely as a function of the gap between the two incomes. Two further points make it larger still than the rate tables alone suggest: the non-earning or lower-earning spouse brings a full unmarried-sized standard deduction to the return, sheltering a tranche of the other's income outright rather than merely taxing it lower, and a joint return raises the point at which most income-based phase-outs begin, so credits and deductions a single filer would have lost may survive.

The other side of it, which is not a tax point but belongs beside the tax point. The joint return that produces the bonus also creates joint and several liability: under section 6013(d)(3) either spouse can be pursued for the entire tax on the return, including tax on income they did not earn. The bonus and the liability arrive on the same piece of paper, and the filing mechanics are covered on the page for married filing jointly.

A marriage bonus is an arithmetic outcome rather than a product, so what follows is what is genuinely good and bad about the way current law produces it.

What works well

• It tracks the gap between the two incomes rather than their total, so a single-earner household gets a larger bonus than two matched earners on the same combined income.
• It requires no election, no form and no planning. It arrives automatically with the joint return.
• It is stable: the doubled bands and the 200 percent standard deduction rule are written into the statute rather than being a coincidence of this year's figures.
• Several of the non-income-tax bonuses, particularly the unlimited marital deduction and portability of the estate exclusion, are worth far more than the income tax effect for the households they reach.

The costs and criticisms

• It is a subsidy to a household with one earner, which is a real policy choice rather than a neutral outcome, and it raises the effective tax rate on a second earner's first dollar of income.
• It is unavailable to unmarried couples in identical economic circumstances, and no contract can replicate it.
• It comes attached to joint and several liability under section 6013(d)(3), which is the price nobody quotes alongside the saving.
• Its size cannot be predicted from income alone, only from how income is split, so it is easy to assume and hard to estimate without running both computations.
• The marital deduction does not reach a non-citizen spouse, so the bonus most people assume is unconditional is not.
• Portability requires an election on a timely estate tax return, so the largest transfer-tax bonus is the one most often lost by inaction.