b babynamesuggestion
Menu

Your Shortlist

0 names saved

Original research

The End of the Shared Name: A Century and a Half of Baby-Name Diversity

In 1880 the ten most popular boys’ names covered 44% of all American newborn boys. Today they cover 8%. The same collapse appears in every country whose birth register is deep enough to measure it.

The finding. In 1880 the ten most popular US boys’ names accounted for 44.22% of all recorded newborn boys. In 2025 they accounted for 7.98%. The top hundred names fell from 80.14% to 38.03%, while the number of distinct boys’ names recorded in a single year rose from 1,058 to 13,930. This is not a US peculiarity: every national register in our data with a deep enough published list shows the same collapse.

Ask someone to picture a mid-century American classroom and they will imagine several boys called Michael and several girls called Mary. That intuition is correct, and it is measurable. What is less widely appreciated is how completely that world has ended — and that it ended everywhere at once, in countries with entirely unrelated naming cultures and unrelated birth-registration systems.

The American century, in one chart

BoysGirls
0%17%34%51%68%85%18801900192019401960198020002020BoysGirlsBoysGirls

Scroll the chart sideways to see all of it.

Share of all recorded US births in each year taken by the 100 most popular names of that sex. Source: Social Security Administration national name file, 1880–2025.
YearBoys: top 10Boys: top 100Boys: distinct namesGirls: top 10Girls: top 100Girls: distinct names
188044.22%80.14%1,05824.65%76.55%942
190034.22%74.65%1,50619.48%68.03%2,223
192031.57%72.12%4,98923.25%65.76%5,767
194033.46%75.26%3,93723.08%66.54%5,028
196029.08%74.34%4,59816.96%61.75%7,332
198023.58%69.73%7,29317.1%54.27%12,161
200013.37%55.4%12,11910.22%42.02%17,660
20207.86%39.65%14,0607.89%32.55%17,527
20257.98%38.03%13,9306.99%31.5%17,297

Two things in that chart are worth separating. The first is the decline itself. The second is that girls’ names were always less concentrated than boys’ names — the girls’ line sits below the boys’ line in almost every year on record, and the gap has never closed. Whatever cultural machinery makes parents willing to pick an unusual name has been acting on girls’ names for longer and more strongly, and it is one of the most stable regularities in the entire dataset.

The same thing happened in six other countries

If this were an artefact of how the SSA collects data, it would not replicate. It replicates. Below is the share of births taken by the ten most popular names of each sex, drawn separately for every national register in our data that publishes a list deep enough to make the calculation meaningful. Each panel is a different country, a different statistics agency, and a different set of names.

United StatesUnited States1880: 35.4%2025: 7.5%ScotlandScotland1974: 28.2%2025: 12.6%Canada (Ontario)Canada (Ontario)1917: 35.7%2024: 8.8%FranceFrance1900: 43.5%2022: 10%IrelandIreland1964: 44.1%2025: 11.4%NorwayNorway1945: 28.4%2025: 15.1%New ZealandNew Zealand1900: 35.8%2025: 11.3%

Scroll the chart sideways to see all of it.

Share of births taken by the ten most popular names of each sex. One panel per national register; the vertical scale is shared so panels are directly comparable. Countries publishing only a short top-N list are excluded — see the method note on truncation.
CountrySourceYears coveredTop-10 share, first yearTop-10 share, latest yearNames listed, latest year
ScotlandNational Records of Scotland1974–202528.2%12.59%1,982
Canada (Ontario)Service Ontario / StatCan1917–202435.72%8.78%3,515
FranceINSEE1900–202243.48%10.01%12,377
IrelandCentral Statistics Office1964–202544.09%11.36%2,004
NorwayStatistisk sentralbyra1945–202528.43%15.12%1,052
New ZealandDepartment of Internal Affairs1900–202535.8%11.3%929

The starting levels differ, the slopes differ, and the years covered differ — but the direction does not. No register in our data shows naming becoming more concentrated over any sustained period.

Why this matters for anyone reading a “most popular names” list

A number-one name today is not comparable to a number-one name in 1950, and treating the two as equivalent is the single most common error in popular naming coverage. When the top ten covered a third of all boys, being at number one meant a name a child would share with several classmates. At an eight-per-cent top ten, the number-one name is closer to a mild preference than a norm. The rank is the same; the thing the rank describes is not. That is also why a name can climb the chart while the actual number of babies given it falls.

The same arithmetic is why our own site treats a rank as one signal among several rather than a headline, and why runs at number one have collapsed from decades to single-digit years.

Method, and what this cannot tell you

  • What is computed. For every year and sex we sum recorded births, count distinct names, and sum births for the names ranked 1–10 and 1–100. The reported figure is that sum as a percentage of all recorded births for that year and sex. No smoothing, no modelling, no interpolation.
  • US records before about 1937 are not a birth register. The Social Security Administration’s name data comes from card applications, so the early years count only people who later applied for a card — which understates births and skews toward people who lived long enough to need one. Figures before 1937 should be read as indicative of relative popularity, not as complete counts.
  • The SSA suppresses any name/year cell with fewer than five births, so the true long tail of rare names is larger than any count here can show.
  • National registers differ in how deep a list they publish. Where a source reports only its top names, its unlisted tail is missing from the denominator, which inflates a measured concentration figure. Every concentration number here is therefore a conservative floor: the real decline in concentration is at least as steep as shown, not steeper than reality.
  • Excluded on purpose. Germany is a shifting panel of city registries rather than one national file, so its yearly totals move for reasons unrelated to naming and it cannot support an over-time series. Finland and Iceland publish living-population counts rather than births. Sources publishing only a top-20 list carry too little tail to measure concentration at all.
  • The registry behind these studies covers 28 countries and territories, 26 of them with official statistics-agency or civil-registry data. India and Japan are the two exceptions — neither publishes a national given-name ranking, so their figures come from documented survey and aggregated datasets and are never presented as official. Finland and Iceland publish a living-population register rather than annual births and are excluded from every study in this series.
  • Meanings and etymologies elsewhere on this site are partly AI-assisted and labelled as such. Nothing in this study depends on that content — every figure here is a count, a rank or a ratio over registry data.
  • Reproducing this. The per-name, per-year, per-country figures behind every number here are available through our free public Baby Name API. If you want the aggregate series as a file, ask.
How to cite this
Ganesh G Kamble, “The End of the Shared Name: A Century and a Half of Baby-Name Diversity,” BabyNameSuggestion.com, 2026. https://babynamesuggestion.com/baby-name-research/baby-name-diversity-since-1880/ — figures computed 2026-08-28 07:56 UTC. Journalists and researchers are welcome to reproduce any figure here with attribution; get in touch for the full per-source provenance or a specific cut of the data.

Other studies in this series

  • How fast names travelOn a fixed panel of nine national registers with an equal observation window, the median time for a name to spread from its first country to its fourth fell from 113 years to 7.
  • Names that changed genderTwenty-eight American names crossed decisively from boys to girls over 146 years of birth records. Under the same test, none crossed the other way.
  • When a spelling overtook its rivalSteven passed Stephen in 1949 and never gave it back. US birth records let us date the exact year one spelling of a name overtook another — something a single name page structurally cannot show.
  • Why boys’ names end in NThe share of American boys given a name ending in the letter N rose from 13% in the 1940s to 35% in the 2010s. In the 2020s it fell for the first time in eighty years.
  • The biggest one-year jumpsSome names do not climb. They detonate. These are the largest single-year multiplications in 146 years of US birth records — reported without any claim about what caused them.
  • Names that came back after a centuryThe folk claim is that a name needs roughly a century out of fashion before it can return. It can be tested: a set of names were common before 1930, effectively disappeared for forty years, and are common again now.
  • How long number one lastsIn 146 years of US birth records there have been only 24 runs at number one. The runs are getting shorter, for the same reason behind almost every other finding in this series.

All of these draw on the same normalised registry described on the research index, and the underlying figures are queryable through the free Baby Name API.

Help us make this better for every parent

Found a mistake · Know a name we’re missing · Have a different or better meaning for a name · Any question or recommendation — your insight helps families worldwide choose with confidence.

Share your feedback →
Scroll to Top
Follow & Connect