Splitting the bills down the middle feels like the neutral choice. It is the arrangement nobody has to justify, which is exactly why so few couples examine it.
So here is the number that should make anyone reconsider. According to Pew Research Center's analysis of Census data, only 29% of American marriages have two partners who earn roughly the same amount. In the other 71%, one person is the primary or sole earner.
An equal split is built for the 29%. We pulled the government data, ran the arithmetic, and put it into five charts.
1. The equal-earner couple has always been the exception
The male-breadwinner household has not disappeared, but it has stopped being the default. In 1972, 85% of American marriages had a husband as the primary or sole earner. By 2022 that was 55%. Couples where both partners earn about the same grew from 11% to 29%, and marriages where the wife earns more than her husband went from 5% to 16%.
Read the green bar carefully. It has nearly tripled in fifty years, and it is still under a third of all marriages. The convention held over from a single-earner era now fits a smaller share of couples than the arrangements it was never designed for.
2. Past a certain income gap, 50/50 is not unfair. It is unpayable.
This is the part that surprises people, because it stops being a matter of opinion.
Hold a household's total take-home pay steady at $8,000 a month and its shared bills at $4,000. Now change only one thing: the gap between what the two partners earn. Under a proportional split, each partner always gives up exactly 50% of their own income, no matter how wide the gap gets. Under an equal split, the lower earner's burden climbs without limit.
Follow the terracotta line:
- At a 2:1 gap, the lower earner is handing over 75% of their take-home pay.
- At 3:1, they hit exactly 100%. Their entire paycheque goes to shared bills and nothing else.
- At 4:1, the arithmetic breaks. Their share is 125% of what they earn, so they go roughly $400 a month into debt while their partner banks a surplus.
Here is the cleanest way to describe what is happening. A 50/50 split is a flat tax on a household. A fixed charge always takes a bigger bite out of a smaller income, which is the same reason flat taxes are called regressive. The dollar amount is identical and the sacrifice is not.
And the damage compounds past the monthly squeeze. The lower earner cannot build an emergency fund, cannot invest, and cannot contribute to retirement, while the higher earner accumulates a surplus every month. Ten years of an arrangement both partners described as equal produces two wildly unequal balance sheets inside one household.
So what does the fair version actually cost? Take a couple bringing home $5,600 and $3,200 a month, with $4,000 of shared bills. Split evenly, they each pay $2,000, which is 35.7% of the higher earner's pay and 62.5% of the lower earner's. Split by income, the higher earner pays $2,545 and the lower earner pays $1,455, and both give up 45.5%. The lower earner keeps $290 more each month, and the higher earner still has $3,055 left. You can run the same maths on your own two incomes in about thirty seconds.
3. A detail almost nobody gets right: gross or take-home?
Most couples who move to a proportional split use their salaries, because those are the numbers they know. That choice quietly undermines the thing they were trying to fix.
Higher earners lose a larger share of their pay to tax and payroll deductions, so the gap between two salaries is wider than the gap between two take-home figures. Split on gross salary and the higher earner ends up giving 47.6% of their actual take-home pay while the lower earner gives 41.7%. The method promises equal sacrifice and overshoots in the opposite direction.
Only a split calculated on take-home pay lands both partners on the same 45.5%. A 60/40 ratio on gross salary is closer to 58/42 once tax comes out, which is why our own calculator asks for take-home pay and says so on the page.
Worth noticing what the fourth column shows, too. Equalising the money left over rather than the share given up swings hard the other way, leaving the higher earner on 57.1% and the lower earner on 25%. Every method has a bias. The useful question is not which one sounds fairest, but which one you can both still describe as fair a year from now.
4. Couples have already voted with their bank accounts
None of this is a prediction. The behaviour has been shifting for nearly three decades, and it is documented in a government longitudinal survey rather than a marketing poll.
Between 1996 and 2023, the Census Bureau recorded three changes at once. Couples holding every account jointly fell from 53% to 40%. Couples holding both joint and separate accounts nearly doubled, from 9% to 17%. And couples with no joint account at all rose from 15% to 23%.
Among unmarried couples living together, only 16% hold a joint account. Any product or piece of advice that assumes a shared pot excludes that entire group by design.
The Census offers a plain explanation: people marry later now. Median age at first marriage moved from 24.8 for women and 27.1 for men in 1996 to 28.4 and 30.2 by 2023. Arriving at a shared life at thirty rather than twenty-four means arriving with a salary history, a credit file, debts and habits already formed. Merging becomes a decision that has to be argued for.
5. The younger the couple, the more separate the money
According to Bankrate, 51% of Gen Z couples keep their finances completely separate, against 34% of Millennials, 23% of Gen X and 15% of Boomers. That is a 3.4x gradient running in one direction.
Generational gaps sometimes close as people age and settle. This one has an economic engine behind it rather than a cultural mood: later coupling, two incomes that are rarely equal, income that moves month to month, and the largest housing costs on record. Harvard's Joint Center for Housing Studies found 22.7 million renter households spending more than 30% of income on housing in 2024, which is 49% of all renters. When the biggest shared cost in the household is that large, getting its division wrong stops being a rounding error.
The strongest argument against all of this
Any honest read has to include the counter-evidence, so here it is.
There is a respected line of consumer research, associated with Emily Garbinsky and colleagues, finding that couples who pool their money into a joint account tend to report higher relationship satisfaction than couples who keep it apart. That work is real and it deserves to be taken seriously rather than waved away.
It also answers a different question. Pooling is about where the money sits. Proportional splitting is about how much each person puts in. A couple can run a single shared account and still agree that the partner earning two thirds of the household income funds two thirds of it. The 17% of couples running a deliberate mix of joint and separate accounts are doing both at once, and they are the fastest-growing group in the Census series precisely because that arrangement needs an explicit rule.
Our own view, for what it is worth: the account structure is a preference and the contribution rule is arithmetic. Argue about the first one all you like. The second one has a right answer for any given pair of incomes.
What we deliberately left out
A report is only as strong as its weakest number, so a few notes on what did not make it in.
- A widely quoted divorce statistic. One analysis attributes a disproportionate share of divorces to female-breadwinner households. It is a correlation from a commercial source, the likely drivers are unequal domestic labour and social friction rather than income itself, and presenting it as cause would be both wrong and insulting to a great many women. Excluded.
- A hybrid-account figure we could not reproduce. A secondary source put the mixed joint-and-separate group at 37% of married couples. The Census release puts it at 17%, and the 37% appears to be a different measure entirely: the share of joint-account holders in 1996 who also had separate accounts. We used the Census figure.
- Market size estimates. Published projections for this category disagree by more than an order of magnitude. None of them are load-bearing for anything a couple actually needs to decide.
- Per-country pooling percentages. Precise to a decimal place and traceable to nothing. Left out.
The five charts above rest on four sources: the US Census Bureau, Pew Research Center, Bankrate/YouGov and Harvard's Joint Center for Housing Studies. Everything else is arithmetic you can check yourself.
The equal split was never a considered position. It was the obvious answer in a world where one person earned the money, and it survived because it sounds like fairness. For the 71% of couples who do not earn the same, it is worth about ten minutes of deliberate thought, which is roughly nine more than it usually gets.
