In France when a person's exact birth date is unknown, a fictitious birth date is set on the 1st of January of the estimated or actual year of birth. I know a few African-born persons in that case. I expected to see that reflected in a slight bump in observed births on the 1st of January... But no such anomaly exists in the article's data...

I think it's funny that there's a noticeable bump in births around nine months after wintertime (Nov/Dec).