a few fun facts

[ekorber]

Some interesting facts at the Population Reference Bureau. I was talking to someone about this kind of thing recently (easwaran?) - we were rather far off on our estimate of the percentage of people ever born who are alive now (not that the estimate here is perfect, but it is at least based on something other than a random guess

Comments

[easwaran]

So that's the Everett these philosophers of physics all like!

And it's interesting to see that estimate carried out - I have severe doubts about their life expectancy of 10 years between 8000 bc and 1 ad though. Not for any particular good reason of course. But that's approximately half of all people ever on their account. Say we cut that in half (by assuming a life expectancy of 20) - then we get 6 out of 80 billion people are alive today, which is about 7%, so that's interesting.

It's also interesting to note that on their projection, the current population of China is almost the total population of the world in 1900 (more than that of 1850). Which makes the 50% estimate not terribly ridiculous.

If we assume constant exponential growth, then we can estimate the percentage of the total population that is currently alive pretty straightforwardly. If the increase is by r% per current expected lifespan, then I believe that we can calculate that r% of the total population is alive today. Here, I mean that one lifespan ago, the total population was (1-r) times the current population. Since the world population has more than doubled in the last 50 years, this projection would yield more than half of the total population being currently alive. However, a prediction of doubling every 50 years means that in the past 2000 years it's doubled 40 times, so if the current population is 6 terapeople, then the population 2000 years ago should have been 6 people. Which seems implausible, to say the least.

[ekorber]

I guess I have no reason to either doubt or believe their 10-year guess. For me at least, I don't intuitively "count" people unless they live for more than a few months, but they are probably counting everyone who survives birth as contributing to the population. I would imagine that the ubiquity of hard slave labor in early civilizations contributed significantly to the death rate of children and young adults. My totally baseless intuition says that maybe 10 is too little but I'm pretty sure 20 is too much.

And somehow I don't think the world's current religious landscape would be quite the same if only 6 people had been alive in ~0 (BCE/CE) (is the year 0 designated at one or the other?)

[ekorber]

Also, interesting graph from Wikipedia here. The immediate observation is that the world population looks like basically just the population of Asia plus a constant factor that might be slightly perturbed by the rest of the world.

[easwaran]

Well, it's a logarithmic scale, so that's the population of Asia times a constant factor.

That graph surprised me, because I was sure that Africa still had fewer people than Europe, but it suggests that Africa passed Europe a couple years ago.

[quale]

Population growth is going to be pretty different from exponential on long time scales I would think. While at any given time it *may* be exponential (supposing it hasn't maxed out) things like the development of agriculture will have huge impacts.

Also major catastrophes (like the one that supposedly reduced the human population to just a few thousand somepoint back when we were in africa) are going to cause issues.

They should be able to get a rough graph of total population that didn't get wiped out by doing genetic analysis and that would be very interesting.

[gustavolacerda]

This 5-7% estimate considerably weakens the doomsday-like argument saying that humanity will be extinguished within the next 50 years or so.