The U.S. tax code does not treat individuals. It treats households, and the household it has in mind is a married couple. This is not a neutral choice. It is a political decision made in 1948 and elaborated for seventy-five years afterward, and it produces winners and losers depending on which kind of couple you are and how you earn.
Until 1948, federal income tax was levied on individuals. Each filer reported their own income, claimed their own deductions, and paid at their own rate. Then Congress, responding to pressure from states with community property regimes (where high-earning husbands in Texas and California could already split their income with non-earning wives and pay at lower rates), introduced joint filing nationally. The deal was simple: a married couple could combine their incomes and apply a rate schedule that was, in effect, twice as wide at each bracket. A couple earning $100,000 paid the same as a single person earning $50,000.
This was a gift to single-earner households. A husband earning $100,000 with a non-earning wife could "split" his income on paper and pay roughly half what an unmarried man earning $100,000 paid. It was simultaneously a punishment for dual-earner households. Two people each earning $50,000 paid the same as a couple where one earned $100,000 — but they had to combine and stack their incomes, pushing them into higher brackets than they would face filing singly. This is the marriage penalty. Its mirror image is the marriage bonus, which favors couples with unequal earnings.
Edward McCaffery's Taxing Women dissected this structure in detail. The joint return is not gender-neutral; it taxes secondary earners (historically wives) at the marginal rate set by the primary earner (historically husbands), which means the secondary earner's first dollar is taxed at a high rate. For a working wife considering whether to remain in the labor force, the joint return makes work more expensive. McCaffery argued this is one of the largest hidden disincentives to female labor force participation in the U.S. tax code, larger than most overt policy levers.
Beyond rate structure, marriage triggers a cascade of other tax consequences. The marital deduction in estate and gift tax law allows unlimited tax-free transfers between spouses, both during life and at death. A wealthy spouse can give the other spouse a billion dollars with no gift tax, and leave a billion dollars at death with no estate tax. The marital deduction is the single largest preference in the estate tax code, and it is available only to married couples. Unmarried partners pay tax on transfers above the annual exclusion and the lifetime exemption.
Marriage also affects the Earned Income Tax Credit. The EITC is structured around household income, so when two low-earning singles marry, their combined household income often pushes them out of the credit. Couples in the EITC range face the largest percentage marriage penalty in the code — sometimes losing thousands of dollars in credit by marrying. Anne Alstott and others have documented this as a regressive feature: the marriage penalty falls hardest on the poor, while the marriage bonus accrues mainly to the rich.
Then there are the secondary effects: filing status determines eligibility for IRA contributions, premium tax credits under the ACA, student loan repayment plans, capital gains treatment on the sale of a home, and dozens of smaller provisions. Each of these uses a different definition of household, and marriage changes how you are counted in each.
Lily Batchelder has argued that the tax code's treatment of marriage is overdue for redesign. The joint return made sense when most households had one earner; it makes less sense when most households have two. The marital deduction in estate tax favors dynastic wealth transmission; it is one of the largest sources of tax-advantaged intergenerational transfer in the developed world. The EITC's marriage penalty undermines the credit's anti-poverty purpose. Each of these features is defended by some constituency, and each survives despite repeated reform proposals.
The honest summary: the tax code marries you to a particular economic identity. If you are a high-earning single-earner couple, marriage saves you tens of thousands of dollars annually and millions over a lifetime through estate planning. If you are a dual-earner couple with similar incomes, marriage costs you money. If you are poor and on the EITC, marriage often costs you money. The romantic decision and the tax decision are entangled, and the entanglement is not symmetric across class.