Methodology

Sources

United States. US Social Security Administration, national baby-name data, 1880–2025. Public domain. SSA lists a name only if it was given to at least 5 babies of that sex that year.

New Zealand. NZ Department of Internal Affairs, births by first name, 1900–2025. CC BY 4.0. DIA lists a name only if it was given to at least 10 babies of that sex that year.

How the files were obtained

The New Zealand file is downloaded directly from data.govt.nz, where the Department of Internal Affairs publishes it under CC BY 4.0.

The United States file is a mirrored copy, and that matters enough to say plainly. ssa.gov answers automated requests from this network with a 403, whatever the request looks like. The SSA name archives (names.zip and namesbystate.zip) were therefore taken from the Internet Archive's 2026 snapshot of ssa.gov rather than from the agency directly. They are the agency's own files, unaltered — but they are a copy of a particular moment, and the honest statement of provenance is "SSA, via a 2026 archived copy", not "SSA".

The practical consequence: the US series runs to birth year 2025, because that is what the snapshot contained. If the SSA revises an earlier year — it occasionally does, as late registrations arrive — this site will not know until the file is re-fetched. Every US figure on the site carries that retrieval route in its source line for exactly that reason.

A source that is not stated as archived, when it is archived, is a false provenance claim. It costs one sentence to be accurate.

Why shares, not counts

Roughly four times as many births were registered in the United States in 1957 as in 1890. A name holding a flat raw count across that span was in fact collapsing as a share of babies born. Every curve on this site is therefore drawn on births per 100,000 of that sex in that year, which is the only comparison across eras that means anything. Raw counts are still printed in the tables, because they are what the registrar published.

Why ranks are computed, not copied

A rank is only true relative to the whole field. Ranks here are computed against every name in that year's government file — all 117,820 US name records — not against the 12,250 names this site publishes pages for. Ranking within our own subset would quietly promote every name on the site.

Why a gap is not a zero

Both registrars suppress small numbers: the US below 5 births of that sex in a year, New Zealand below 10. When a name is missing from a year, this site says "below the reporting floor" and never "0" or "none". The distinction matters most for New Zealand, where the floor is higher and the country is small — plenty of real New Zealand names never clear ten in a single year.

Which names get a page

A name earns its own page at 1,000 lifetime US births, or 200 in New Zealand. Below that the "curve" is single digits wobbling around the reporting floor — a chart of noise. Those names remain in the underlying data and in the year tables; they simply do not get a standalone page built on a line that cannot be read.

On meanings and origins

NameCurve publishes what the birth registers contain: who was named what, when, and where. It does not print a meaning or etymology it cannot cite, because the popular ones are frequently invented and endlessly copied between name sites. Where origins are added they will carry a named reference work, not a vibe.

SOURCE·US Social Security Administration, national baby-name data·Public domain. retrieved from the Internet Archive's 2026 snapshot of ssa.gov (ssa.gov refuses automated requests); files unmodified. SSA lists a name only if it was given to at least 5 babies of that sex that year.·REVIEWED JUL 2026

SOURCE·NZ Department of Internal Affairs, births by first name·CC BY 4.0. DIA lists a name only if it was given to at least 10 babies of that sex that year.·REVIEWED JUL 2026

SOURCE·US SSA (1880–2025) + NZ DIA (1900–2025) birth registrations·ranks and shares computed from the source files, see method·REVIEWED JUL 2026