Korean and American Naming Data, Side by Side

Korea's figures come from the Supreme Court's birth-registration statistics. The US figures come from Social Security card applications. Different filings, different agencies, different publication depth: Korea publishes the top 20 per year and the US publishes the top 1,000.

So raw counts are not comparable and neither are ranks past position 20. What is comparable is shape: how concentrated the top of each list is, and how often it changes.

In 2026, the top Korean name 서윤 accounts for 7.0% of all births within the published top 20. In 2025, Liam accounts for 2.6% of the equivalent US top 100.

That is a difference of roughly 2.7 times in how much of the field the leading name takes. Korean parents are drawing from a visibly narrower pool of currently fashionable names.

Across 19 years of Korean records the combined #1 name changed 13 times. Across 18 years of US records it changed 4 times.

The pairing is counterintuitive: the more concentrated list is also the less stable one. A narrow pool of fashionable names means the leader is decided by small shifts in fashion, so it moves. The US list is broad and slow, with names holding the top spot for many years at a time.

Meaning. Korean names are usually written in hangul on the register, and the Chinese characters that carry the meaning are a separate choice recorded only in that family's register. One Korean name maps to many possible character combinations, so there is no dataset of "what Korean names mean" to line up against English name etymologies.

This is why Korean name pages here list the permitted characters per syllable and their glosses rather than asserting a meaning. Any site that gives you a single confident meaning for a Korean name is choosing one combination and not telling you.

The same problem does not exist on the US side, where a name's etymology is a property of the name itself rather than of an individual family's paperwork. That asymmetry is why the two sections of this site are built differently: Korean name pages show a candidate set, US name pages show a distribution.

Both datasets have a publication cap and both caps are the publisher's, not ours. Korea stops at 20. The US releases 1,000 through its forms and nearly everything in the downloadable files, subject to a 5-occurrence privacy threshold.

So a Korean name being absent from our data means it was outside the top 20 that year, which for a country of Korea's birth numbers still means it was reasonably common. The absence carries much less information than a US absence does, and it is the reason we publish 96 Korean name pages against 345 American ones despite Korea having the more concentrated list.

Not for deciding which country has better names, and not for cross-country rank lookups. The value is in the structural questions, where both datasets answer the same question with the same kind of number.

How fast does naming fashion turn over? Korea says fast, the US says slow. How much of the field does the leading name take? Korea considerably more. Does a concentrated market imply a stable leader? Apparently the opposite. Those are findings you cannot get from either dataset alone, and they are the reason it is worth publishing both under one roof rather than treating Korea as a footnote to the US data.