How we estimate

Sources

Dataset	Publisher	Coverage
National baby names	Social Security Administration	Births 1880–2025, by name and sex
State baby names	Social Security Administration	Births by name, sex, year and state
Surname file	US Census Bureau	162,253 surnames from the 2010 census

All three are public-domain US government datasets. We download them unchanged and process them into the files this site loads; nothing is hand-edited and no figures are invented to fill gaps.

Step 1 — from births to living people

Birth records count everyone ever registered with a name, including those who have since died. For a name that peaked in 1950 this is a very large difference, so using raw birth totals — as many name sites do — substantially overstates how many people you could actually meet.

For each name we take its births in every year and multiply by the probability that someone born that year is alive today:

living(name) = Σ_year births(name, year) × survival(year)

The survival term combines a standard US period life table with a cohort adjustment, because people born in 1920 faced far higher childhood mortality than today’s life table implies. Summed across all names, this produces an estimated living population which we then express as shares.

Step 2 — putting both names on one base

The two datasets are not directly comparable: one counts births since 1880, the other counts people in 2010. So we convert each to a share of its own universe and restate both on the current US resident population of about 342 million. A first name’s displayed count is its share of the living name population applied to that base; a surname’s is its share of the census surname universe applied to the same base.

This is why our surname figures differ slightly from the raw 2010 census counts, which we also show. It keeps the two halves of the calculation consistent with each other.

Step 3 — combining first and last

people(first, last) = P(first) × P(last) × population

This treats the two names as statistically independent. It is the standard approach for this kind of estimate and it is also the single largest source of error on the site. See the limitations below.

Geography

State results are reported as a concentration index rather than a raw count, because ranking by count simply lists the largest states every time. The index is the name’s share of births in that state divided by its share nationally, so 3.0× means the name is given three times as often there as you would expect. States with too small a sample for the name are excluded to stop tiny counts producing wild ratios.

Known limitations

Independence is wrong, and we know it

First names and surnames correlate strongly with ancestry. Pairings that are common within a community are therefore underestimated by the independence assumption, sometimes by a wide margin, while culturally mismatched pairings are overestimated. Correcting this properly would need joint first-name and surname data, which is not published. This is the limitation to keep in mind above all others.

Immigration is not captured

Social Security first-name data records births registered in the United States. People who arrived as adults are not in it, so names common among recent immigrant communities are understated, and their age profiles skew younger than reality.

The surname data is from 2010

It is the most recent file the Census Bureau has published. Rescaling to today’s population assumes every surname grew at the same rate, which is not true: fast-growing surnames are understated and long-established ones slightly overstated.

Small counts are suppressed at source

The SSA publishes a name for a given year only once at least five babies received it, and the Census Bureau only lists surnames held by 100 or more people. Roughly 29 million Americans carry a surname below that line. A name missing from this site is usually genuinely rare rather than absent in error.

Errors compound at the rare end

Multiplying two small, uncertain probabilities gives a result that is proportionally far less reliable than either input. When the tool says eleven people, read it as “a handful”. The rarity tier is meaningful; the exact digits are not.

What this site is not

It is not a people search. We hold no records about individuals, we cannot tell you who shares your name, and nothing here can be used to find or identify anyone. Every figure describes a population in aggregate.

Reproducing this

The processing is deterministic: the same source files produce the same output every time. The datasets are linked above and are free to download from the SSA and Census Bureau directly, so any figure here can be checked against the original.

Corrections

If you believe something here is wrong, please say so via the contact page. Methodology notes and data vintages are updated whenever the underlying files change.