The order of common birthdays differed between 2000 to 2004 and 2010 to 2014 in historical U.S. Social Security Administration (SSA) records. Among the 365 dates that occur every year, only 9 of the 20 highest-ranked dates appeared in both periods. The full lists had a rank correlation of 0.617 after dividing each date's count by its year's total births.
Using each year's weekday average changes the picture. The correlation rises to 0.923, and 16 dates appear in both top-20 lists. The measured stability depends on the reference used for weekdays. This comparison cannot explain why births happened on particular dates.
How much did ranks move?
The median date moved 35 places between the two rankings based on each year's birth share. Its median move was 20 places under the weekday comparison. A correlation nearer 1 means more similar full lists; the top-20 overlap looks only at the highest-ranked dates.
September 9 shows the difference. It moved from rank 69 in 2000 to 2004 to rank 4 in 2010 to 2014. Against its weekday average, it moved from 7 to 4. December 25 stayed last under both measures in both windows. These are ranks of recorded births on dates, not counts of living people who share a birthday.
The table also includes 2005 to 2009. The raw lists agree less in the two comparisons involving that middle period than in the comparison between the first and last periods. The weekday-relative lists agree strongly in all three comparisons. That pattern does not describe a steady shift in one direction.
| Five-year windows | Birth-share correlation | Weekday-relative correlation | Shared birth-share top 20 | Shared weekday-relative top 20 |
|---|---|---|---|---|
| 2000–2004 vs. 2010–2014 | 0.617 | 0.923 | 9 | 16 |
| 2000–2004 vs. 2005–2009 | 0.314 | 0.901 | 5 | 16 |
| 2005–2009 vs. 2010–2014 | 0.142 | 0.894 | 2 | 17 |
The year-by-year study shows each year's winner and September 9's rank in each year. Here, the question is whether the whole list stayed similar.
How we compared the windows
We used the pinned SSA daily U.S. birth file for every period. In each year, we divided the count for a date by that year's total births. We averaged the five yearly shares for each date and ranked the 365 dates that occur every year. Ties share a rank. Ranking five-year birth totals instead gives nearly the same endpoint result.
For the weekday check, we divided each date's births by that year's average for its weekday, averaged the five ratios, and ranked them. This changes the reference for each count; it does not remove holidays, seasonal patterns, or other influences. Our weekday and weekend study shows the daily pattern.
We leave February 29 out of the 365-date comparison because it occurred twice in the early window and once in the late one. The records end in 2014 and cannot establish today's rankings. See the methodology for source limits, or check your birthday against the site's longer 1994 to 2014 baseline.