In the Census Bureau’s 2024 American Community Survey estimate, 31.3% of married same-sex couple households were interracial, compared with 19.5% of married opposite-sex couple households. The difference is 11.8 percentage points across two household groups.
On this page
The comparison the Census published
The Census Bureau's analysis of the 2024 American Community Survey found that 31.3% of married same-sex couple households were interracial. The corresponding share for married opposite-sex couple households was 19.5%. Both numbers come from the same survey year and concern married couples identified within households. Their difference is 11.8 percentage points, by subtraction of the published rounded shares. That is the useful comparison: a larger share of one measured married-household group was interracial.
The two percentages have different denominators. For the first, imagine all married same-sex couple households identified by the survey and ask what share had spouses of different races. For the second, do the same among identified married opposite-sex couple households. There is no single pool of "all couples" underneath both percentages. A difference in shares also does not mean that 11.8% more same-sex couples existed, or that 11.8% of the entire population belonged to an interracial couple.
The existing table converts the comparison into a simple picture: in 100 households of each type, the rounded shares would correspond to about 31 interracial married same-sex households and about 20 interracial married opposite-sex households. Those are illustrations of rates, not counts of surveyed households. The table's third row uses our arithmetic, not a separate Census estimate. Keep the "percentage points" unit whenever you repeat the 11.8 difference.
A larger percentage does not imply a larger absolute number of interracial households. The two married-household groups differ in size, and the comparison presents shares within each group. To make a claim about national counts, a source would need to supply or support both the relevant group total and the interracial count, with the same definitions and rounding. Here the published percentages are enough for the share comparison. They are not a reason to present a new precise count.
Read the difference in percentage points
The table may be wider than the screen. Focus it and use the arrow keys to reach every column.
| Married household group | Interracial share | Population |
|---|---|---|
| Same-sex couples | 31.3% | Married partners in the same home |
| Opposite-sex couples | 19.5% | Married partners in the same home |
| Difference, by our subtraction | 11.8 percentage points | 31.3 minus 19.5 |
Our illustration: Imagine 100 households of each type. The published shares correspond to about 31 and 20 interracial households, respectively; these are illustrations from rounded shares, not new counts.
How to read it: The shares use separate married-household denominators. They do not measure dating preferences or unmarried couples.
Download the table (CSV) ↓Why these are household figures
The American Community Survey identifies a spouse or unmarried partner through that person's relationship to the householder. This makes the published estimate about couples the survey can identify in a shared home. It is not a register of every marriage. A married couple living apart is outside a shared-home couple measure. Another couple may live in a household where their relationship is not identifiable through the householder relationship question. The Census Bureau notes this limit in its account of the 2024 figures.
This distinction matters when you encounter a larger-sounding total for same-sex couples. The same Census article estimates about 836,000 married same-sex couple households and about 551,000 unmarried same-sex couple households in 2024, rounded to about 1.4 million same-sex couple households together. The 31.3% interracial share belongs to the married same-sex group. Applying it to all 1.4 million would silently add unmarried couples to the denominator and produce a count the source did not report.
The phrase "couple households" also describes a living arrangement rather than every person in the home. Children, relatives, or roommates can be present. The rate compares the races of the married partners, not all household members. It cannot tell whether an interracial couple has children or what anyone else in the home identifies as. If a chart switches from households to people, it has changed its unit even if the topic sounds similar.
Race and Hispanic origin need separate labels
The Census collects race and Hispanic origin as separate concepts. This article's 31.3% and 19.5% comparison uses the Census Bureau's "interracial" label for the reported married-household measure. A study that counts partners of different Hispanic origins, combines racial and ethnic differences, or uses another race classification can report a different share. That difference does not, by itself, show that either result is wrong. It may show that the studies asked different questions or sorted partners into different categories.
When citing the result, use the source's terms and observation year: "interracial share of married couple households in the 2024 ACS." Avoid replacing "interracial" with "interracial or interethnic" unless the source actually supplies that combined measure. Avoid calling the figure a measure of mixed-ethnicity couples. The published comparison has a defined classification, and changing the label changes what readers think was counted.
If you are comparing two reports, look for the exact question or table note defining race, Hispanic origin, and partnership status. Then check whether each uses a shared home, a legal marriage record, or a survey of individuals. A result for newlyweds, for example, has a different base from a result for all currently married couple households. A result for unmarried couples has a different relationship category. Keep those differences visible before trying to explain a change in the percentages.
What the gap does not explain
The survey describes existing households at one point. It does not ask each person whom they considered dating, which partner they would have preferred, or why a couple formed. The difference between 31.3% and 19.5% cannot establish that people in one group are more open to interracial dating. It also cannot show whether geography, age, marriage timing, or any other factor caused the gap. Those are separate research questions requiring data that measure the proposed factors.
Nor does an interracial share tell us how an individual marriage is experienced. Couples differ in family background, culture, language, and the way they describe their own identities. A category built for a survey comparison should not be used as a shortcut for the quality or difficulty of a relationship. The finding is useful because it accurately describes a measurable difference between two groups. It does not need a story about the couples' motives to matter.
Be careful with the word "likely." Saying "31.3% of measured married same-sex couple households were interracial" keeps the denominator explicit. Saying a same-sex person is "more likely to date interracially" changes the outcome from a household state to a person's dating behavior, which this source did not measure. Even "more likely to be interracial" can confuse a percentage-point difference with a relative percentage. The table gives a clearer sentence: the interracial share was 11.8 percentage points higher in the measured married same-sex household group.
How to use this statistic in context
Use the number when the question is about the racial composition of married couple households identified in the 2024 ACS. Keep all four parts together: the year, the household unit, married partnership, and the two separate denominators. If you only have room for a short line, use the answer at the top of this page. If you need a visual comparison, use the existing table and its note rather than translating the shares into invented national counts.
For a question about all same-sex couple households, first separate married and unmarried groups. The Census article gives counts for both, but the interracial percentage quoted here belongs to married households. For a question about relationships between people who do not live together, this ACS household analysis does not cover that population. For a question about new marriages in a particular year, seek marriage-record or newlywed data with that exact scope.
The comparison is also a poor way to estimate a local dating pool. A national household share cannot say how many available people of any race live in a city, how many want to date, or how they describe their identity. The broader interracial dating guide discusses other measures, but each figure must keep its own unit and source. That discipline lets the 2024 ACS finding stand for what it does well: a clear, bounded comparison of two married-household groups.
