The year was 1969, and Vivien Kellems of Connecticut was once again unhappy with America’s tax collectors. Two decades earlier, she had made waves by refusing to withhold taxes from her employees’ paychecks; she wasn’t an IRS employee, so why should she do the agency’s work? This time around her complaint was slightly less Tea Party and a bit more modern liberal: America’s tax code discriminated against the unmarried.

Her congressional testimony was a tour de force of federal income-tax history. In the early years of the tax, she explained, married Americans had discovered a clever trick: in states with “community property” laws declaring that marital belongings were shared equally, couples with one earner could divide the income in half and pay taxes separately, availing themselves of the lower rates available to lower earners. Those living in common-law states, which did not allow this trick, naturally protested having to pay more, and some states changed their laws as a result.

Yet, rather than simply rewording the law to ban this practice, known as income-splitting, Congress had effectively made it universal. As of 1969, when Kellems spoke, married couples had separate tax brackets with thresholds set at double the tax brackets for singles. (The initial 14 percent rate applied up to $500 for singles and $1,000 for couples.) In making these reforms, Kellems argued, Congress had sought to be fairer to common-law states, but had paid little attention to how the unmarried were treated.

For traditional couples with one breadwinner and one homemaker, income-splitting meant a significant tax cut upon marriage, as a single worker could double his bracket thresholds by marrying a nonworker. Under the 1969 tax code, taking account of the relevant deductions and exemptions as well, a single person earning the median household income of $9,400 paid about $1,750. If this person added a nonworking spouse, that fell to $1,360, quite a substantial sum in a world where every $100 was worth about $800 today.

“All we single people ask is the same tax break,” Kellems told Congress. Lawmakers met her partway, trimming down the rates charged to singles so their burden would never exceed that of a married couple with the same income by more than 20%.

That only created new problems. In addition to breadwinner couples and single people, the system also had to work for dual-earning couples. And the very features that had penalized single people relative to sole-breadwinner households—married tax brackets that are double the brackets for singles, with the same rates charged in each bracket—also prevented marriage penalties when two people with similar incomes married.

For instance, under the new, improved system passed in 1969, money earned between $8,000 and $10,000 was taxed at 25 percent for single taxpayers. But if two workers married, their combined $16,000–20,000 bracket would make them pay a 28 percent rate. That difference alone could cost a couple $120.

As more and more married women worked, and as women’s wages rose, such disparities generated growing pushback. Some couples even tried divorcing at the end of each year so they could file as single, and then remarrying.

It’s a Catch-22 inherent in the idea that married couples should pay taxes on their joint income. If the married thresholds are simply set at double the single ones, sole-breadwinner households get a large bonus relative to singles. But reduce that bonus, and dual-earner households are punished for marrying.

 Over time, we’ve managed to bring back the system that annoyed Kellems—albeit in a new social context with additional layers of complexity.

Nowadays, the married income-tax brackets are once again double the single ones, at least until the top 37 percent rate (which kicks in at about $578,000 for singles but $694,000 for couples for 2023). There is a large breadwinner bonus—a $65,000 income translates to $6,560 in taxes for a single and $4,040 for a couple—but two single workers who marry won’t be penalized unless they’re extremely well-off.

Childbearing outside of marriage has gone from rare to common, however, making it highly consequential how single parents, and not just single individuals, fare. Single parents can use the head-of-household status instead of filing as purely single, gaining access to better tax brackets that will not fully double if they marry. As a result, breadwinners get a bonus relative to singles, and unmarried working parents can get a bonus relative to married ones.

Another major change of the past half-century has been the expansion of the safety net, some of it done through the tax code. Most prominently, the Earned Income Tax Credit phases out only a little more slowly for married couples than for single parents, which can give a moderate bonus to breadwinners and levy an astounding penalty on couples with equal incomes. An unmarried couple with two kids, each earning $30,000, can get about $5,000 in EITC between them; a married couple with two kids and $60,000 combined income gets nothing.

To dual earners’ benefit, though, there is also a child-care tax credit for those with younger kids, at least if they earn enough to have tax liability to offset. For most taxpayers, it’s worth up to $600 for one kid and $1,200 for two or more. The tax system doesn’t just coldly pick a side in the mommy wars, in other words: it does at least something for everyone.

I’ve been writing about family policy for years, and these dynamics have long fascinated and frustrated me. Hailing from the center-right of the political spectrum, I don’t want government using tax policy to push parents around. I don’t want it giving large bonuses to breadwinner couples or even, as some have suggested, outright paying homemakers, but I also don’t want it subsidizing child care, which destroys a major financial benefit that naturally follows from a parent staying home.

I do want government policy to help parents in general, but it should do so in a neutral fashion, letting parents decide what arrangement works best for them. After all, I know firsthand that different arrangements can work best in different situations: I’m the son of a breadwinner and homemaker, the husband of a working mother, and a former stay-at-home dad, albeit one who kept working part-time.

Since the current system tries to do everything at once, I’ve long struggled to wrap my head around how all these biases add up. Who’s treated the most unfairly here, and how aggressive would reforms need to be?

Taking things perhaps a little too far, I recently tried to figure it out by writing a computer program that, for a variety of broadly middle-class couples, simulates their income trajectories and tax treatment over the course of their entire adult lives. I report what I found in a new paper for the Manhattan Institute.

My imaginary couples live in a sort of Groundhog Day world, earning incomes typical for 2022 and paying taxes under that year’s code for their whole lives—thus illustrating, with admittedly copious use of simplifying assumptions, how the current rules treat existing lifestyles. Readers can explore the simulations in an included tool.

In these simulations, a childless single person at the median-income trajectory earns about $2.3 million from age 23 to 65—topping out at about $60,000 a year, around age 50—and pays a little above $200,000 in income taxes. If that person marries a non-earner, the couple’s taxes fall to about $125,000. (Kellems must be rolling in her grave.) If the earner has two kids, the tax totals fall to about $100,000 without marriage and $50,000 with it, thanks mainly to the Child Tax Credit.

But consider if both partners earn the median income: without kids, they pay a bit more than $400,000 in taxes whether they marry or not (this being the beauty of doubled thresholds); with two kids, they face a modest marriage penalty, paying about $330,000 instead of $310,000, thanks to the loss of head-of-household status.

The penalty is much more pronounced for lower-income parents, who also have the Earned Income Tax Credit to lose: two 25th-percentile earners with two kids, whose annual incomes top out around $35,000 each, will pay more than $100,000 in income taxes in total if they marry but only about $30,000 if they do not.

My simulations also touch on Social Security, where the spouse of a worker can receive a benefit worth half the worker’s own, a policy that grants another bonus to breadwinner couples, given that payroll taxes are essentially flat (with a cap on both taxes and benefits for the highest earners). In some scenarios, dual-earning couples can even get worse benefits than breadwinners whose earnings and payroll taxes were the same in total. For example, under the current formula, someone who consistently earned $60,000 a year while working (after adjusting old earnings for wage growth), or $5,000 a month, will qualify for about $2,280 a month for himself but $3,420 in total with the spousal benefit. By contrast, two people who consistently earned $30,000 would qualify for $1,480 each, for a total of $2,960.

Altogether, our system does some things well, such as reducing taxes for lower earners, for parents in general, and for seniors. But breadwinners and single parents benefit from major provisions tailored to their needs, including the head-of-household status, the Earned Income Tax Credit, and the very existence of joint tax brackets, while dual-earning parents are often treated worse than they would be if they were not married. The child-care tax credit, as annoying as my friends on the right and I may find it, simply doesn’t compare with these other provisions.

Tax scholars often speak of a “trilemma” regarding marriage. It’s nearly impossible to have a tax system that simultaneously (A) is progressive, taxing higher incomes at higher rates; (B) taxes couples based on their total earnings, regardless of how those earnings are distributed across one or both spouses; and (C) taxes couples the same regardless of whether they are married. (The one exception is a flat tax coupled with a universal, flat, per-person basic income, which would make the flat tax progressive in a marriage-neutral way.)

Our current system mostly sacrifices (C). I posit that it should sacrifice (B) instead, by returning to individual taxation, with no income-splitting this time. This shift would, thanks to the rise of women’s work and wages, benefit about half the country. The specific arrangement that seems most sensible to me is to allow the higher earner in each couple to use the head-of-household filing status if children or a spouse unable to work are present. Bluntly, this would increase taxes for breadwinners and cut them for dual earners with kids, though many variations are possible to soften these effects.

Why abandon (B), a principle known as “couples neutrality”? In many cases this principle is not even good in theory. A couple where one person works full-time making $60,000 by himself, while the other stays home, is obviously far better off than a couple where both parents must work full-time to earn $30,000 each. The current system barely recognizes the difference, and not always in the right direction: the worse-off couple can claim the child-care tax credit while their kids are in daycare but, as we have seen, may well receive lower Social Security benefits.

To be fair, other scenarios exist where the value of couples neutrality is more apparent. For example, if both spouses earn the same wage, it probably shouldn’t matter for their taxes whether one person works 60 hours, as opposed to one working 40 and the other 20, though some to my left would like to nudge couples toward a more even division of paid work or cut taxes for secondary earners. That happens under individual taxation because the secondary earner starts at the bottom of the tax structure, with their own standard deduction and lower brackets. (By contrast, under joint taxation, the primary earner “uses up” these early stages of a progressive tax code, leaving the secondary earner to face the higher rates and thus discouraging them from entering the workforce.) In any case, such an effect strikes me as much less objectionable than subsidizing breadwinner couples while actively punishing dual earners with kids for being married, which is what the current system does.

I make no pretense of having written an entirely new tax code. Keeping this concept revenue-neutral would require careful adjustments to rates and credits, and rules would be needed to prevent backdoor efforts at income-splitting. At minimum, I hope to inspire a more honest and open discussion about how the current tax code functions and who comes out ahead, even if those of us with strong sympathies for stay-at-home parents find the numbers uncomfortable.

Happy Tax Day.

Photo: sturti/E+ via Getty Images


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next