What the Census Means by a 'Household'

Read almost any demographic figure and you’ll find the word household doing silent work underneath it. Median household income. Households below the poverty line. Owner-occupied households. The household is the basic unit of American demographic data — there are about 127 million of them — and most people read right past the word without noticing that it means something specific, and something different from “family.” Once you see the distinction, a lot of otherwise-confusing numbers snap into focus.

A household is a housing unit, not a family

To the Census Bureau, a household is simply everyone who lives in one housing unit — a house, an apartment, a mobile home. That’s the whole definition. The people in it may be a married couple and their kids, or three roommates, or a single person living alone, or a widow and her adult son. If they share the housing unit, they’re one household.

A family, by contrast, is a narrower thing: two or more people living together who are related by birth, marriage, or adoption. So every family is a household, but a huge share of households aren’t families at all.

That gap is bigger than most people expect. More than a quarter of all American households are a single person living alone, and a growing share are unrelated people sharing space. The Bureau sorts every household into one of two buckets:

• Family households — a householder living with at least one relative.
• Nonfamily households — a person living alone, or a householder sharing only with non-relatives (roommates, an unmarried partner counted under older definitions, and so on).

When you see “household income,” it’s counting the combined income of everyone in that unit, related or not — the roommates’ paychecks get added together just as a married couple’s would.

Why the family-versus-household gap bends the numbers

This isn’t a definitional footnote. Which unit you use changes the statistics in predictable directions, and knowing that keeps you from misreading a place.

Income. Median family income always runs higher than median household income, for a simple reason: families tend to have more earners and exclude the people-living-alone who pull the household figure down. If one source quotes a place’s family income and another quotes household income, you’ll see a gap of many thousands of dollars that is pure definition, not a real difference. On this site we use household income throughout, because it’s the most inclusive and most widely reported.

Household size. Divide a place’s population by its number of households and you get average household size — and it varies enormously. A college town or a downtown full of studio apartments has tiny households; a suburb full of young families, or a community where multiple generations live together, has large ones. Two places with identical household incomes can have very different per-person living standards if one is packing four people into the household earning that income and the other just one or two.

Poverty. Poverty status is determined at the family level and applied to the people in it, using thresholds that scale with the number of people and children. So household composition feeds directly into who counts as poor — the same income supports one person comfortably and a family of five barely at all, and the poverty measure is built around exactly that.

Who the “householder” is

One more piece of vocabulary shapes the data: the householder. This is the person (or one of the people) in whose name the home is owned or rented — historically called the “head of household.” A lot of demographic detail is organized around this reference person: household type, and in older tabulations the race and age of the household, were all defined relative to the householder.

It’s worth knowing this is a reference point, not a ranking. In a household of equals sharing rent, one person is still designated the householder for tabulation purposes. When you see a statistic “by householder,” it’s using that one person as the anchor for classifying the whole unit, not making a claim about who’s in charge.

Reading household data well

A few habits keep household statistics honest:

• Confirm the unit before comparing. Household income and family income aren’t interchangeable. A “$95,000 median” means very different things depending on which one it is — check the label before you conclude one place out-earns another.
• Pair income with household size. A high household income in a place with large households is spread across more people than the same income in a place of singletons. Population divided by households tells you which you’re looking at.
• Remember the solo-dweller effect. Places with lots of people living alone — dense cities, retirement areas, university towns — post lower household incomes partly because so many households are a single earner. It’s a composition effect, not necessarily a sign of a poorer place.

The household is a deceptively simple idea that carries most of American demographics on its back. Keep in mind that it’s a housing unit and not a family, that its size varies wildly from place to place, and that the same dollar figure stretches differently depending on how many people share it — and the numbers stop tricking you. Every place profile on this site reports household-based figures; the precise definitions are in our glossary, and how we handle the data is on the methodology page.