People alive today vs number of people who have ever lived

February 8, 2005 10:17 PM

Several years ago, I heard, on some TV show that the total number of people who have died up until now is less than the number of people alive today. Then I saw this. (+)

Does anyone have a link to a reasoned estimate of how many people have lived and died throughout human history? I have tried Google but just can’t get the right query.

posted by arse_hat to society & culture (23 answers total)

"estimate people lived history" turned up this as the first result.
posted by vacapinta at 10:28 PM on February 8, 2005

Thats what I was looking for. I should have used lived rather than dead for my queries. Thanks vacapinta.
posted by arse_hat at 10:42 PM on February 8, 2005

I've seen this estimated elsewhere as 75 billion.
posted by ikkyu2 at 11:16 PM on February 8, 2005

I've seen this estimated elsewhere as 75 billion.

In an order-of-magnitude calculation, thats the same as 100 billion.

This was brought home to me by an Astro. professor. We were trying to infer the radius of a galaxy. He did some calculation and some astute student pointed out that he hadn't estimated the radius, he had estimated the diameter! He then chided the student, pointing out that in an estimate of that magnitude, 10^11 and 2X10^11 were the same estimate!
posted by vacapinta at 12:50 AM on February 9, 2005

This was brought home to me by an Astro. professor. We were trying to infer the radius of a galaxy. He did some calculation and some astute student pointed out that he hadn't estimated the radius, he had estimated the diameter! He then chided the student, pointing out that in an estimate of that magnitude, 10^11 and 2X10^11 were the same estimate!

Uggg.. similarly with my Solid State Component's class. I swear to god that everything involved dropping terms with a lower order of magnitude. You expand everything out to this equation with a dozen terms, and then in one swoop the professor would declare half of them zero because "they're an order of magnitude less".
posted by sbutler at 1:14 AM on February 9, 2005

As a slightly offtopic comment - it's always terrified me that the population of the earth has doubled in my lifetime...
posted by benzo8 at 2:21 AM on February 9, 2005

"Behind every man now alive stand 30 ghosts, for that is the ratio by which the dead outnumber the living." -- Arthur C. Clarke, 2001: A Space Odyssey, 1969

Of course, that was 35 years ago and I don't know what Clarke used as a source on that...
posted by neckro23 at 3:13 AM on February 9, 2005

dropping terms with a lower order of magnitude

Entirely off topic, and I apologize, but how does this make sense? Doesn't every piece of information you drop make your answer that much less accurate, even if infintisimily?

I used to get into arguments with my chem and math teachers about the similarly-crazy dropping of extra places. Let's say you multiply 1.5 x 1.33 -- the "correct" answer would only be to the first decimal. The argument being, "we only know one decimal on one of the digits, so that's the most accurate our answer can be." Except that's stupid: the answer is just the reflection of an equation with two numbers of varying accuracy. 1 1/2 x 1 1/3 would produce a fraction, which would somehow be just fine, but when it's changed over to decimals, all of a sudden the answer is only good to one decimal place? What gives?

Sorry, again. This would probably be better as its own AskMe thread.
posted by Civil_Disobedient at 5:59 AM on February 9, 2005

Civil, for a more concrete example: Let's say you are going for a walk and you know exactly how long the route is (down to the millimetre), but you only know roughly how fast you'll be walking. Doesn't it make sense that you can only know roughly how long the walk will take and not down to the millisecond?
posted by teg at 6:27 AM on February 9, 2005

teg -- I understand what you're saying. What bugs me is that your variable x -- the time the walk takes -- is really an equation. The equation is based on two variables, one you know extremely well, the other is only a guesstimate. So, x is best represented as an equation. But that equation has another way of being represented: as a number. If we're trying to get the most accurate response that's possible (given the imprecise conditions of one of the variables), we would just give a distance / speed formula with whatever variables we're given.

Of course, people just want a number, not an equation. So you give them an answer, and (now this is the part that bugs me) you truncate the number to the accuracy of the most inaccurate number. Thing is, your answer's accuracy is independent of the variable's accuracy, because your answer is just a decimal representation of an equation.

Perhaps it's just an easier shorthand for telling someone, "one of these variables wasn't very accurate". But it would be more correct to invent some kind of notation that could show the accuracy of the original variables without interfering with the accuracy of the equation result. For example, I walk at about 3 mph., but the distance that I'm walking is 30.03 miles. The walk will take 30.03/3 = 10.01 hours [&e^0]. The &e^(power) indicates the accuracy of the least-accurate variable.
posted by Civil_Disobedient at 7:07 AM on February 9, 2005

That was an interesting article. I didn't realize how low the average birth rate was, though - that 50 per 1000 is considered high?! That's 5%... or am I misunderstanding something?
posted by mdn at 7:24 AM on February 9, 2005

Thomas Browne addresses this in his Dedictaion to Urn-Burial (emphasis mine):

'We present not these as any strange sight or spectacle unknown to your eye, who have beheld the best of Urnes, and noblest variety of Ashes; Who are your self no slender master of Antiquities, and can daily command the view of so many Imperiall faces; Which raiseth your thoughts unto old things, and consideration of times before you, when even living men were Antiquities; when the living might exceed the dead, and to depart this world, could not be properly said, to go unto the greater number.
posted by driveler at 7:25 AM on February 9, 2005

On the original question, see also Snopes' take on it.

I used to get into arguments with my chem and math teachers about the similarly-crazy dropping of extra places. Let's say you multiply 1.5 x 1.33

You were right to argue with your math teachers--in math, 1.5*1.33=1.995. In chemistry (and other sciences), though, 1.5 is not 1.5--1.5 is a shorthand way of writing 1.5±0.1. (If it were 1.5±0.01, you would have written "1.50," not "1.5".) So you have (1.5±0.1)*(1.33±0.01). Or a range of 1.4*1.32=1.848 on the low end, and 1.6*1.34=2.144 on the high end. If you write "1.995" as the answer to this, you imply that the answer is 1.995±0.001, and you do not have that level of accuracy.
posted by DevilsAdvocate at 7:32 AM on February 9, 2005

Obviously there should be a lot more dead people on Earth. It wouldn't hurt the environment any.

Number who have ever been born: 106,456,367,669

But, geez, reincarnation could drastically lower that number, assuming we're counting the number of individuals who have ever been born. Just sayin'.
posted by Shane at 7:36 AM on February 9, 2005

What DevilsAdvocate said, except it's +/- 0.05 and 0.005 respectively. Here's a decent enough explanation of significant digits I found googling.
posted by drpynchon at 8:14 AM on February 9, 2005

Fascinating question. Wouldn't the answer depend, however, on where one draws the line between apes and humans?

'Cause there's a lot of dead monkeys.
posted by goodnewsfortheinsane at 8:28 AM on February 9, 2005

1 1/2 = 1.50000~
1 1/3 = 1.33333~
posted by mischief at 8:45 AM on February 9, 2005

1.5 is a shorthand way of writing 1.5±0.1.

Ah, Jesus, another great idea of mine stolen!

Thanks for explaining this; now I can have some of my sanity back.
posted by Civil_Disobedient at 8:52 AM on February 9, 2005

I've seen this estimated elsewhere as 75 billion.

(vacapinta) In an order-of-magnitude calculation, thats the same as 100 billion.


Agreed! Just thought I'd toss in someone else's independent estimate. FYI, it was in a lecture by Allen Rosenfield, Dean of the Columbia Public Health school.
posted by ikkyu2 at 10:38 AM on February 9, 2005

The issue of significant digits can also be searched for under the topic spurious accuracy.

To tie together the two different topics in this this thread: if someone were to ask you for an estimate of how many people had ever been born (say, through 2002), and you said that a reasonable estimate was 106,456,367,669, this would undoubtable be a case of spurious accuracy. There should, at most, be two significant digits in this estimate, and quite possibly only one - so "around 100,000,000,000" is the appropriate answer (if you must provide a point estimate, not a range).

An answer with a lot of significant digits (in this case) is misleading because it implies that the estimate is far more precise than it really is.
posted by WestCoaster at 10:53 AM on February 9, 2005

So I was wrong, it's closer to one in twenty than one in ten.

The original ratio was one that my brother used to quote; he usually had some reputable source when he made claims like this, so the factoid stuck in my head. I would ask him where he got the number, but sadly his is unavailable nowadays, being one of the nineteen (or nine).
posted by Quinbus Flestrin at 10:55 AM on February 9, 2005

That's why they talk about joining the majority.
posted by languagehat at 12:23 PM on February 9, 2005

average birth rate was, though - that 50 per 1000 is considered high

Doesn't that mean 50/1000 per year? A population would double, simplistically, in ~ 20 years.

or do I have this wrong?
posted by PurplePorpoise at 2:52 PM on February 9, 2005