How we estimate the age of a name

The short version

Social Security has counted how many babies got each first name, every year, since 1880. Its actuaries also publish cohort life tables: for people born in a given year, the share still alive at each age. For every name, we multiply the babies born in each year by the share of that year's babies still alive in 2026, separately for girls and boys, and add it all up. That gives an estimate of how many people with the name are alive, and in which years they were born, which gives their ages.

The data

• Baby names. SSA's national baby-name file (names.zip): for every year from 1880 to 2025, every first name given to at least five babies of the same sex, with the count. It covers 105,966 different names and about 375 million births.
• Baby names by state. SSA's state file (namesbystate.zip): the same counts by state of birth for births from 1910 on, for the 50 states and DC. It powers the state pages, which describe people born in each state, wherever they live now.
• Life tables. SSA's cohort life tables from the 2026 Trustees Report (alternative 2 (intermediate) assumptions), by sex, for every birth year from 1900 to 2100. For each birth year they give l(x): how many of 100,000 babies are still alive at each age. Babies born before 1900 would be over 119 today, so we count none as living.

A worked example

In 1947, SSA recorded 99,908 babies named Linda (almost all girls). The 2026 Trustees Report cohort table says that of girls born in 1947, about 64% are still alive at 79. So about 64,000 of 1947's Lindas are alive today, aged 79. Repeat for every year from 1880 to 2025 and both sexes, and you get 1.1 million living Lindas, with an average age of 70. The Linda page charts every year.

What each number means

• Average age is the mean age of all the living people with the name, estimated as above. It's the number in each page's headline.
• Median age: half of living people with the name are older, half younger.
• Middle half: the range from the 25th to the 75th percentile of ages. “Half of all living Lindas are between 66 and 78.”
• Estimated alive is rounded on the page (“about 1.1 million”), because it's an estimate, not a count.
• Peak year is the year the most babies got the name (both sexes together).
• Generations follow the Pew Research Center's birth years: Silent 1928–1945, Baby Boomers 1946–1964, Gen X 1965–1980, Millennials 1981–1996, Gen Z 1997–2012. We call 2013 on Gen Alpha, and 1901–1927 the Greatest Generation.
• Older than N in 100 names ranks the name's average age against every name with a page.
• Curve shape. Each name is classed by the shape of its popularity over time (its share of all births, smoothed over five years): a spike got half or more of its babies in one 15-year stretch; a steady name never had a third of its babies in 15 years and is still given; a comeback peaked before 1950, fell below 30% of that peak, and has climbed back; a fading name is well below its peak; a rising name peaked in 2005 or later and is still near the top.
• Same-age names are the names of the same main sex whose average ages are closest. A name twin is the name whose whole age curve (the share of its living people born in each year since 1900) is the closest match.
• Other spellings are names that sound alike under a set of English spelling rules (Caitlin, Kaitlyn, Katelyn) and belong mostly to the same sex. The rules are a heuristic, so we call them spellings, not the same name.

State pages

For each state we apply the same national life tables to the babies born there (SSA publishes no state life tables), so a state's numbers describe people born in it, not people living in it today. “More common than nationally” compares a name's share of living people born in the state with its share of all living Americans in the national file, among names with enough people in the state to be meaningful. State counts leave out a name in any year and state where fewer than five babies of one sex got it, so they add up to less than the national counts.

The 2056 projections

For “Will this baby name sound old?” and the forward-looking lists, we assume each name keeps getting the same number of babies per year as its average over 2023–2025, and age everyone to 2056 with the same cohort life tables. Real fashions change, so these are directions, not forecasts.

What the estimates miss

• Rare names. SSA leaves out any name given to fewer than five babies of one sex in a year, to protect privacy. Names that are rare every year are missing entirely, and rare years of common names are undercounted.
• Older records are thin. SSA's counts come from Social Security card applications, which started in 1936. Many people born before the late 1930s never applied, so early births are undercounted, and the oldest names look a little younger and less common than they were.
• Immigration. People born outside the US aren't in SSA's birth counts, and people born here who moved abroad still are. Names common among immigrants (María, Mohammed, Wei) have more living people in the US than we estimate.
• Survival isn't name-specific. The life tables give the share alive by birth year and sex for the whole country. If people with a name live longer or shorter lives on average, we can't see it.
• Spelling and changes. Each spelling is its own name in SSA's records, and nicknames (Bob, Peggy) count only when they were the name given at birth. People who change their first name later still count under their birth name.
• Projections, not counts. Life tables beyond the latest observed year are SSA's projections. The newest years of the tables are estimates too.

Which names get a page

A name gets its own page when at least 5,000 people with it are estimated alive: 3,713 names today. The search box covers 100,000 names, with a quick estimate for those without a page; for very rare names that estimate rests on few people and can swing a lot.

Where the idea comes from

The method was popularized by Nate Silver and Allison McCann at FiveThirtyEight in a 2014 article, “How to Tell Someone's Age When All You Know Is Her Name” (archived; FiveThirtyEight has since closed), using the same two SSA sources. AgeByName runs it on the newest data for every name with enough people, and refreshes every spring when SSA adds a year of births. We compared their 2014 numbers with ours in FiveThirtyEight's name ages, 12 years later.

Updates

These numbers use births through 2025 and the 2026 Trustees Report life tables, with ages as of 2026. SSA publishes a new year of baby names each May; we rerun every page then. Spotted something off? Tell us.