How couples’ work earnings compare

What the $5,000 comparison measures

The Census Bureau compared annual work earnings for opposite-sex couples who shared a home in its 2025 Current Population Survey. It included married and unmarried partners. In 26% of those couples, the partners' earnings were within $5,000 of each other. The number tells us how often two reported annual work-earnings amounts were close by a fixed dollar test. It does not tell us whether partners had the same hourly pay, worked the same hours, or made the same unpaid contribution at home.

A fixed $5,000 band looks different at different income levels. Two hypothetical partners earning $25,000 and $29,000 differ by $4,000, or 16% of the lower figure. Partners earning $100,000 and $104,000 differ by the same $4,000, or 4% of the lower figure. Both pairs fit the Census 'within $5,000' category, though the relative gap feels different. These examples are arithmetic illustrations, not pairs observed in the survey.

The comparison covers earnings from work in a year. It leaves out savings, property, debt, investment income, and the value of unpaid care. A couple can have similar wages but unequal assets, or different earnings but equal control over a shared budget. The source did not measure either arrangement. If your question concerns household resources, two work-earnings amounts are only part of the answer. For example, hypothetical earnings of $0 and $4,000 meet the near-equal test, while $80,000 and $86,000 do not. The threshold classifies a gap; it does not rank how secure either household is.

Couples with and without children

Among opposite-sex co-resident couples without children in the home, 36% had partners whose annual work earnings were within $5,000. Among those with at least one child in the household, fewer than 10% did. These are shares within two separate groups. They do not say that 36% of all couples had no children, or that fewer than 10% of all couples both had children and earned similarly.

Nor can we average 36% and a value below 10% to recover the 26% overall share. We would need to know how many couples were in each group and the exact share hidden by 'fewer than 10%.' An unweighted midpoint would treat a small subgroup and a large subgroup as if they contained the same number of households. Keeping the three published results beside their own denominators is more accurate than manufacturing an intermediate rate.

A reader might see that difference and conclude that having a child caused an earnings gap. The cross-sectional comparison cannot settle that. It does not isolate work hours, childcare arrangements, jobs, ages, prior earnings, or when the partners formed the household. Some couples had unequal earnings before a child arrived. Others may divide paid and unpaid work differently after. The survey tells us the pattern at one point in time, not a before-and-after effect for the same couple.

A child can also live in the home without telling us which adult's paid hours changed, who performs care, or how the family assigns expenses. The Census result says 'at least one child in the household' for this comparison. It should stay attached to that household condition rather than become a claim about all parents' wages. The figure is a prompt to ask a better question about work and care, not a measure of the value of care itself.

Read the 52% share correctly

Census also reported that in 52% of the observed opposite-sex co-resident couples, the male partner earned at least $5,000 more annually. That share includes married and unmarried partners in the same home. It does not mean 52% of men earn more than 52% of women in the whole workforce, or that a male partner earns more in every category of couple. A comparison within an existing household has a different denominator from an individual wage survey.

Consider two hypothetical households. In one, the man earns $62,000 from work and the woman $54,000 during the same year. The $8,000 difference fits the 'man at least $5,000 higher' condition. In another, the woman earns $62,000 and the man $54,000; that pair does not fit the reported 52% condition. The study's short public summary does not give a separate headline share for the second direction. These examples explain the rule and are not new survey observations.

The 52% and 26% groups use the same $5,000 threshold but answer different questions. One identifies couples with a male partner at least $5,000 ahead; the other identifies couples whose work earnings lie within $5,000. Do not add them and label the result 'men who earn as much or more.' The categories are about couples, and the near-equal band has no older-partner or higher-earner direction. The published summary does not present a complete table of every earnings interval.

Neither share measures equal pay for equal work. An hourly pay comparison would need similar jobs and hours; an annual-earnings comparison combines pay rates with the amount of work during the year. It also omits unpaid household work. A statement about fairness, contribution, or economic dependence needs more information than these two category shares provide.

How to use the figures in a real budget

If you are comparing your own household with the study, first check the unit. Add up each person's work earnings over the same year, then calculate the absolute dollar difference. If the difference is $5,000 or less, your pair fits the published near-equal band. Comparing one person's full-year total with the other's last three months would give a false gap. A partner who began work midway through the year may also look different on an annual total from their current monthly pattern. That classification says nothing about whether the income covers your housing, care, or savings needs. It only places two annual work-earnings numbers on one side of a fixed threshold.

A useful budget conversation can then go further than the survey. Identify predictable costs, unpaid care, and the income each person can rely on now. A change in hours or childcare can make the same couple look different next year without changing either person's worth or commitment. The Census percentages can put your dollar gap in population context; they cannot decide how a couple should split bills or care.