People love talking about baby names, and so the time of year when Retraite Québec announces the most popular names of the year is always a tempting fruit for a journalist looking for a quick story.
Unfortunately, the top five names doesn’t tell us that much. Liam and Olivia have been popular for a while now, and the top five doesn’t change that much year to year. Some journalists go a bit further and look at the end of the list for more unusual names (misspellings, mashed-up names, or words you wouldn’t think of as names), which can be amusing if you can avoid making fun of a name in a language you don’t understand.
Fortunately, Quebec’s open data website has full datasets of first names used since 1980 for boys and girls, so how about we do some more interesting number crunching?
Rather than just look at the most and least popular, I decided to see if I can suss out some trends. So I took the full data set, with almost 400,000 first names, and added a column calculating how many times they were used in the past 10 years versus the past 40 years. Normally this would be about 25%, but many names have gotten much more or much less popular.
A few caveats about these lists:
- The database only goes back to 1980, so names that had already fallen out of favour by then won’t show up.
- I’m assuming the database is correct. There may be errors here throwing off the results.
- The lists are separated by gender because it’s two different data sets. In several cases names have gotten less popular because they’re traditionally associated with the other gender. Combining them would take a while and probably crash my computer.
- For data entry cleanup reasons, I’ve excluded compound names from these lists (whether they have a hyphen or a space separating them) as well as names that are in the database as just one letter (which I assume are initials incorrectly coded as names). I’ve also set minimums for the number of times a name has to be used to weed out outliers.
- Since the database does not include accents, I have not included them here.
Good? OK, here we go: