Are more babies born under a full moon?
What do you think? Vote to see the verdict
The key figure
+5
more babies per 1,000 under a full moon. Sounds like something, but randomly picked days match it 26 times out of 100.
We looked at 85.7 million US births. Under a full moon about 5 more babies are born per 1,000, a gap so small it also shows up when you pick days at random.
The data
Ask in any hospital and someone will tell you: full-moon nights bring more births. It sounds logical. If the moon moves the tides, why not a baby?
Testing it takes a huge number of births, counted day by day. In Spain that isn’t possible: the statistics office records the day of every birth, but its open data only go down to the month. So we used data from the United States.
- 85.7 millionbabies born
- 21 yearsof data, 1994-2014
- 259full-moon days
What changes births most isn’t the moon
Far fewer babies are born at weekends
Babies born each day in the US in 2014. The coloured dots are the year’s 12 full moons.
The line dips every weekend and on holidays such as Christmas. The year’s 12 full moons land some high, some low, without standing out.
Source: CDC/NCHS and SSA, via FiveThirtyEight
Show the data as a table
| Month | Daily mean | Min | Max | Full moon (mean) |
|---|---|---|---|---|
| January 2014 | 10,644 | 6,924 | 12,598 | 11,822 |
| February 2014 | 10,746 | 7,060 | 12,519 | 12,494 |
| March 2014 | 10,525 | 6,789 | 12,342 | 6,973 |
| April 2014 | 10,744 | 6,877 | 12,528 | 11,912 |
| May 2014 | 10,855 | 7,145 | 13,411 | 12,234 |
| June 2014 | 10,923 | 7,359 | 12,704 | 12,442 |
| July 2014 | 11,531 | 7,567 | 13,575 | 8,650 |
| August 2014 | 11,486 | 7,630 | 13,418 | 12,589 |
| September 2014 | 11,683 | 7,639 | 13,863 | 13,629 |
| October 2014 | 11,112 | 7,279 | 13,082 | 12,130 |
| November 2014 | 10,622 | 7,219 | 13,349 | 12,422 |
| December 2014 | 10,959 | 6,749 | 13,634 | 7,196 |
Every dip in the line is a weekend. Many births are scheduled, like caesareans, and hardly ever on a Saturday or Sunday.
Full moons, on the other hand, land some high, some low. At a glance they don’t stand out.
So, to look for the moon’s effect, we compare each day with similar days: a Saturday with the other Saturdays of that month. That tells us how many babies were “due” each day.
Day by day, through the moon’s cycle
Are more babies born under a full moon?
Babies above or below what was due, per 1,000, by day of the lunar cycle
The bars barely move and all of them stay inside the zone chance produces, full-moon days included.
Source: CDC/NCHS and SSA, via FiveThirtyEight
Show the data as a table
| Lunar day | Extra babies per 1,000 | Chance zone (low) | Chance zone (high) |
|---|---|---|---|
| 0 | −0.1 | −7.4 | +8.3 |
| 1 | −1.4 | −7.4 | +8.9 |
| 2 | +2.7 | −7.1 | +9.0 |
| 3 | +3.3 | −6.9 | +8.7 |
| 4 | +2.0 | −7.9 | +9.6 |
| 5 | +0.9 | −7.6 | +8.5 |
| 6 | +0.3 | −7.3 | +8.5 |
| 7 | +1.7 | −7.4 | +8.6 |
| 8 | +1.3 | −7.6 | +8.5 |
| 9 | +2.9 | −6.9 | +8.9 |
| 10 | +0.4 | −7.1 | +8.8 |
| 11 | +2.1 | −7.3 | +8.7 |
| 12 | +1.1 | −7.7 | +8.5 |
| 13 | +4.7 | −7.5 | +9.2 |
| 14 | +7.0 | −7.3 | +7.9 |
| 15 | +1.1 | −7.0 | +8.6 |
| 16 | +1.5 | −7.5 | +8.5 |
| 17 | −1.5 | −8.0 | +8.7 |
| 18 | −1.7 | −7.5 | +8.3 |
| 19 | +0.4 | −7.0 | +8.5 |
| 20 | +1.8 | −7.5 | +8.6 |
| 21 | −2.5 | −7.0 | +7.9 |
| 22 | −1.0 | −7.5 | +8.4 |
| 23 | +0.3 | −7.2 | +8.4 |
| 24 | +4.2 | −7.5 | +8.1 |
| 25 | +2.7 | −7.5 | +8.3 |
| 26 | −0.5 | −7.4 | +8.9 |
| 27 | −1.4 | −7.6 | +8.8 |
| 28 | −1.8 | −7.7 | +8.5 |
| 29 | +2.8 | −11.4 | +10.9 |
Each bar is a day in the moon’s cycle, from one new moon to the next.
If the bar goes up, more babies were born that day than were due. If it goes down, fewer.
The full moon falls on days 14 and 15. And day 14 is the tallest bar of all: +7 babies per 1,000.
Looks like the myth is right…
…but the bars always wobble a little, even when nothing special happens.
The grey zone shows how far they go by pure chance: about 8 babies up or down per 1,000.
Every bar stays inside the grey zone, full-moon days included.
One of them being the tallest means nothing: some bar had to be.
What if it’s just chance?
On full-moon days about 5 more babies per 1,000 were born. To tell whether that’s a lot or a little, we ran a draw.
In each lunar month we picked an ordinary day, as a fake full moon, and checked how many extra babies it had. That way the fake moons land once a month, just like the real ones. We did it 2,000 times.
In 26 out of 100 draws, random days match the full moon
Each square is a draw of randomly picked days
- 26draws where random days stray from normal as much as the full moon, or more
- 74draws where they stray less
In 26 out of 100 draws, randomly picked days strayed from normal as much as full-moon days did, or more.
Source: CDC/NCHS and SSA, via FiveThirtyEight
Show the data as a table
| Squares | |
|---|---|
| draws where random days stray from normal as much as the full moon, or more | 26 |
| draws where they stray less | 74 |
If the moon really pushed births, random days would hardly ever catch up: statisticians want that to happen fewer than 5 times in 100. Here it happens 26 times in 100.
What if there were a tiny effect?
Not seeing an effect isn’t the same as proving there isn’t one. So we looked at something else: how big it could be, at most.
At most, the full moon would add 13 babies per 1,000
What the real effect could be, according to the data, versus what the myth would need
The full moon’s real effect would be between -3 and 13 babies per 1,000. For the myth to be right it would take at least 20.
Source: CDC/NCHS and SSA, via FiveThirtyEight
Show the data as a table
| babies per 1,000 | |
|---|---|
| What we measured | +5 |
| What it could be | −3 … +13 |
| What it would take | +20 |
For a midwife to notice busier full-moon nights, the moon would have to add quite a few births. We set the bar at 20 per 1,000, and the data say it’s 13 at most. Even the best case falls short.
In short
- Weekends and holidays change the number of births a lot. The moon doesn’t.
- Under a full moon about 5 more babies are born per 1,000, and you get the same picking days at random.
- Even if there were some effect, it would be 13 per 1,000 at most: far from what it would take to notice.
- Verdict: myth. It probably survives thanks to memory: a hectic night under a full moon sticks in your mind; an equally hard one without it doesn’t.
What people thought
How we measured it
We counted the babies born each day in the United States over 21 years: 85.7 million in all.
Each day is compared with similar days: same weekday, same month, same year. That tells us how many babies were “due” that day, so we don’t mix up the moon with weekends or holidays.
Then we set aside the 259 full-moon days. On those days about 5 more babies per 1,000 were born than were due.
To see whether that could be chance, we ran a draw: take one random day from each lunar month, as fake full moons, and redo the sum 2,000 times. In 26 out of 100 draws the gap was as big or bigger.
The draw also shows how far the real effect could go: between -3 and 13 babies per 1,000. For the myth to be right, the full moon would need to add at least 20. Since we can rule that out, the verdict is myth.
What this analysis doesn’t say
- The data are from the United States. Spain’s statistics office (INE) records the day of every birth, but its open data only go down to the month, and months can’t show the full moon.
- We work out the full moon with a simple formula that can be off by half a day. With 259 full moons, the odd slip doesn’t change the result.
- We know which day each baby was born, but not the time.
- The bar of 20 babies per 1,000 is our call, not a law. If an effect below 13 per 1,000 matters to you, these data can’t rule it out.
- Not seeing an effect doesn’t mean it is exactly zero. It means that, if it exists, it is too small to notice even with 21 years of data.
Sources
- CDC / NCHS — nacimientos diarios en EE. UU., 1994-2003 Compiled by FiveThirtyEight. We use 1994-1999.
- Social Security Administration — nacimientos diarios en EE. UU., 2000-2014 Compiled by FiveThirtyEight. We use 2000-2014.