How NameAlmanac counts names
Every figure on NameAlmanac is drawn from public sources: the Social Security Administration’s baby-name files, SSA’s cohort life tables, the Census Bureau’s count of first names in the 2020 Census, and Wiktionary; the living figures are NameAlmanac’s own estimates from them. Each page names the file behind each section and its date, and links to the source so you can check it. The name records cover 105,948 first names recorded since 1880.
In short
NameAlmanac loads SSA’s national file (every first name given to at least 5 babies of one sex in a year, 1880 to 2025) and its state file (the same by state, 1910 to 2025) into one database with a scripted pipeline, and computes every rank, share, trend and estimate from those counts. The pipeline stops if a rebuild would lose rows it cannot restore, and it checks known figures against SSA’s own tables before it writes anything: for example, Emma’s 22,719 births in 2003 and Michael’s 64,934 in 1985. Pages read their figures from that database when you open them.
Sentences about a name are written from that name’s own numbers by fixed templates, so two names never share a sentence that is not true of both.
NameAlmanac is made with AI tools: they write the code, the page designs and the explanatory text on guides and research pages; the numbers come from database queries, and each analysis states its method and limits. Meanings are quoted from Wiktionary with credit, never written by us.
Sources: SSA national and state baby-name files, published May 8, 2026; SSA cohort life tables, 2024 Trustees Report; Wiktionary via the kaikki.org extract of Sep 25, 2026.
The sources at a glance
| Source | What it holds | Edition | Used for | Terms |
|---|---|---|---|---|
| National baby-name file, Social Security Administration | Every first name given to at least 5 babies of one sex in a year, with the count, 1880 to 2025 | Published May 8, 2026 | Counts, ranks, shares and trends on every page | Public domain (U.S. government work) |
| State baby-name file, Social Security Administration | The same for each state and the District of Columbia, 1910 to 2025 | Published May 8, 2026 | State pages and the state measure | Public domain (U.S. government work) |
| Cohort life tables, SSA Office of the Chief Actuary | Survival by birth year and sex, intermediate assumptions | 2024 Trustees Report | How many people with a name are alive, and their ages | Public domain (U.S. government work) |
| First names in the 2020 Census, Census Bureau | Every first name reported 100 or more times, by sex, with the Bureau’s disclosure-avoidance adjustment | Released April 14, 2026 | The Census comparison on name records and in the most-common-names study | Public domain (U.S. government work) |
| Wiktionary, via the kaikki.org extract | Meanings, origins and pronunciations of names | Extract of September 25, 2026 | Quoted meanings, credited and linked to each entry | CC BY‑SA 4.0 |
What the SSA files hold
The national file lists, for each year since 1880, every first name given to at least 5 babies of one sex, with the count. The state file does the same for each state and the District of Columbia since 1910. The counts come from applications for Social Security cards, so a name is spelled the way it was written on the application: Sofia and Sophia are two names, and so are Aiden and Aidan. SSA shortens names to 15 letters and leaves out any name given to fewer than 5 babies of a sex in a year, to protect privacy. That is why a rare name can have gaps in its history, and why its true count may be higher than shown.
The files also hold records that are not names, such as Unknown, Baby or Infant, entered when a card was issued before a name was chosen. NameAlmanac keeps them in totals, because they are births, but never lists, ranks or pages them.
SSA rebuilds its files from its records with every release, and late applications change the counts of recent years slightly, so a figure here can differ by a few dozen from a site that loaded an earlier release; every figure on NameAlmanac comes from the files SSA published on May 8, 2026.
Ranks follow SSA’s convention: by count within each sex and year, with ties broken alphabetically. A share is the name’s births per 1,000 babies of that sex SSA recorded that year, which corrects for the rise and fall in the number of births. Rare names show a share per million instead.
SSA background: limits of the national file and the state file.
How the living estimate is made
The estimate of how many Americans with a name are alive in 2026, and how old they are, carries each year’s births under the name forward with the survival rates in SSA’s cohort life tables for that birth year and sex (2024 Trustees Report, intermediate assumptions). A cohort table follows the people born in one year through their lives, so it reflects the mortality they actually meet rather than today’s rates. SSA’s files only hold people who received a Social Security number. Before 1937 most people got one as adults, so for birth years before 1937 survival is counted from the age they were in 1937, not from birth. Births before 1900 are counted as no longer living.
The result is an estimate, not a count, and every page says so. SSA’s files count only people born in the United States who received a Social Security number: anyone born abroad is left out, so names common among immigrants are undercounted. The same survival rates apply to every name, so a name more common in a group that lives longer or shorter than average is over- or under-estimated. We round the estimates to 3 significant figures.
The Census Bureau published in April 2026 how many people gave each first name in the 2020 Census (every name given 100 or more times, by sex, with the Bureau’s disclosure-avoidance adjustment). Name records and the most-common-names study set that count beside the same estimate carried forward to Census day, April 1, 2020. For names most Americans got at birth here the two agree closely (Emma: 498,221 counted, about 502,000 predicted); where the Census counts far more, the likely reasons are people born abroad, whom SSA’s files leave out, short forms people give (Zach for Zachary) and compound names SSA writes as one word; the Census file records no birthplace, so each record states the difference and gives these as possible reasons. Where the Census counts fewer than predicted (Christopher, 22% below), people who give a short form or another name, and a survival estimate that runs high, are the likely reasons.
Life tables: SSA Office of the Chief Actuary, cohort life tables, 2024 Trustees Report. 2020 Census: Census Bureau, first names in the 2020 Census, released April 14, 2026.
How the state figures work
A state’s top names are its own counts. To say which names a state gives more often than its national share predicts, we compare the births a name actually had in the state with the births its national share would predict there, year by year: the state’s listed births that year times the name’s share of the national births of the names the state lists that year, summed over 1910 to 2025. The state file leaves out names given to fewer than 5 babies in the state, so the share is taken among the listed names; otherwise every listed name would look over-represented, most in small states. Only names given at least 1.05 times as often as predicted are listed. Standardizing by year stops a state that grew late, such as Nevada, from favoring recent names. We only rank names with at least 500 births in the state (100 for a single decade). The same measure over 2016-2025, for names given to at least 200 babies in the state in those years, shows what a state names today; it leads each state page.
State file: SSA, 1910-2025, published May 8, 2026.
Similar names, trends and the typical names of a time
Names with a similar history are the names whose share of births rose and fell across the same decades, measured by the similarity of their decade-by-decade shares, among names of the same main sex; a name given at least 10% to each sex is compared only with other such names. A decade’s or a generation’s most typical names are those whose share of births then is highest relative to their share across 1880 to 2025, among names given to at least 10,000 babies in the span and at least 90% to one sex; spellings of one name are grouped. Risers and fallers compare ranks in 2024 and 2025 within the top 1,000, as SSA measures them. Generations use the Pew Research Center’s birth years; Generation Alpha (2013 to 2025) is our own band, because Pew has not defined one.
Where meanings come from
Meanings, origins and pronunciations are quoted from Wiktionary’s entries for the name, under Wiktionary’s CC BY‑SA 4.0 licence, from the kaikki.org extract of September 25, 2026. We take the English section when it has the name, otherwise the first section with an etymology, and we say which section the quote comes from. We cut the text at Wiktionary’s cross-references, because an entry’s etymology can go on to describe something else with the same name. Where Wiktionary has no entry for a name, its page says so: NameAlmanac does not write meanings.
Wiktionary content is available under CC BY‑SA 4.0; each quote links to its entry.
Which pages search engines are offered
Every page that holds a record is offered to every search engine; pages are never hidden from one engine for being thin. A name gets a full record page when it has more than a short SSA series: 1,000 or more recorded births, a sourced meaning, a 2020 Census count, a state with 100 or more births, or a place in SSA’s 2025 top 1,000, or when readers already find its page through a search engine. Every other name is a row of its letter’s A to Z directory (for example every name starting with A), with the babies given it since 1880, the years SSA recorded it, its peak year and its 2025 births, and its old address leads to that row. When a name gains a meaning, a Census count or readers, it gets its page at the next load. Search results, tool answers with parameters and error pages are not offered to any engine.
Who reviews the pages, and when
Before each release, figures on every page type (names, letters, years, states, decades, origins, rankings and the research pages) are checked against the SSA, Census and Wiktionary records they cite, and the pages are read in separate review passes against Google’s and Bing’s published quality guidelines, for meanings or origins a reader could dispute, and for any figure that reads differently on two pages. The studio runs these checks with AI tools and records each one; the last full review was for the release of October 8, 2026.
Checks, updates and limits
SSA publishes each year’s files in spring. When a new file appears we rebuild the database, and the rebuild must reproduce known totals before it is published. Automated tests compare the pages’ figures with the database, and the full record of a name is only built for names with at least 50 recorded births over at least 5 years (or 50 in the latest year); other names get a short lookup record. The data cannot tell you why parents chose a name, how a name is pronounced in a given family, or the sex of the person today: SSA records the sex marked on the application.
If you find an error, write to hello@namealmanac.com. We check it against the source file, correct it if we got it wrong, and list the correction with its date on the corrections page.