Wednesday, September 26, 2007

Do living people outnumber the dead?

I can't resist commenting on studies like this. Tom Kemp in a Genealogy Librarian post links to this article in the March 2007 issue of Scientific American by Clara Curtin.

In the article, the author provides the following information:

"In 2002 Carl Haub, a demographer at the Population Reference Bureau, a nongovernmental organization in Washington, D.C., updated his earlier estimate of the number of people that have ever existed. To calculate this, he studied the available population data to determine the human population growth rates during different historical periods, and used them to determine the number of people who have ever been born."

<----- snip ----->

"To calculate how many people have ever lived, Haub followed a minimalist approach, beginning with two people in 50000 B.C.—his Adam and Eve. Then, using his historical growth rates and population benchmarks, he estimated that slightly over 106 billion people had ever been born. Of those, people alive today comprise only 6 percent... "

The population numbers at different time points given are:

* 50,000 BCE, homo sapiens appears (an estimate)
* 9,000 BCE, population is 5 million
* 1 CE, population is 300 million
* 1650 CE, population is 500 million
* 1800 CE, population is 1 billion
* 2002 CE, population is 6.2 billion

Based on historical growth rates that the researcher "knows" (e.g., less than 0.1% from 9,000 BCE to 1 CE), he comes up with a total number of people who have ever lived on Earth - 106 billion. He doesn't present the actual growth rates other than the one listed above.

Assuming that he has some sort of birth rate, child mortality rate, proportion of adults who bore children, average number of children birthed, and average age of motherhood, it would be theoretically possible to come up with an equation or a spreadsheet that would calculate the total number of people who ever lived (at least since a mythical Adam and Eve in 50,000 BCE). But how accurate could the result be if the input data were only estimates? Is there demographic data for those rates and other data?

Let's consider the only growth rate that is quoted - less than 0.1% average population growth between 9,000 BCE and 1 AD. Assuming his numbers of 5 million (9,000 BCE) and 300 million (1 CE) are correct, a 0.1% average population growth rate yields 40.373 billion people, not 300 million. The actual average population growth rate is about 0.0455%, not 0.100% (calculated by multiplying (5 million) times [{1.0+growth rate}^9001] - for this calculation 1.000455^9001 = 60.0, times 5 million = 300 million).

For the years between 1 CE and 1650 CE, the average population growth rate calculates to 0.031% (this may be highly variable, as the author points out the effects of plague and other population controls). Between 1650 and 1800, the average population growth rate is 0.46%. Between 1800 and 2002, the average population growth rate is 0.91%. The current average population growth rate is 1.2%.

The real challenge is making estimates for the rates of birth, childhood mortality, the percentage of adults who bore children, the average lifespan of adults who survive childhood, etc. Each of these variables may change significantly during each generation and century. It would be interesting to see graphs of these variables and of the world population based on them. The "error bands" for the rates would be useful too. I would love to see and understand the spreadsheet!

Of all the numbers given above by the article, the one that I find most curious is 300 million people in 1 CE. If the numbers are correct, the average population growth rate from 9,000 BCE to 1 CE is 50% greater than the average population growth rate from 1 CE to 1,650 CE. I would have thought that they had plagues and disasters in the pre-Roman times also.

Obviously, I can't authoritatively challenge the population estimates given, or the estimated result of 106 billion people since 50,000 BCE. I'm just an old aerospace engineer with a calculator, not a demographer like Carl Haub with a large spreadsheet and the considered scientific opinion of many demographers at his fingertips. Frankly, it all seems pretty reasonable to me, now that I think more about it.

You know, 50,000 BCE is about 2,000 generations. Each of us has over 1.1 quadrillion potential ancestors in the 50th generation. The number of potential ancestors 2,000 generations ago is very high: 2^2000 = about 10^600 (that's 10 with 600 zeroes - at least a zillion!). The actual number of ancestors in 50,000 BCE is much closer to 2, I think!

I wrote this post to get my thoughts down on paper - math is something I was good at in school. This was FUN to think about! Thanks to Tom Kemp for posting the link to the article.

UPDATE 6:30 PM: Now if we could only get the Mother of All Genealogy Databases (MOAGD) to find the names, dates and places for all 106 billion of these folks, we would all be out of a job. Can FTM 2008 handle 106 billion people? How long would that take to input? Let's see, if everyone did 60 people an hour, that's 16.67 million hours...or over 200,000 full-time person-years. Now that would be a big project. Scratch the idea!


Janice said...


I boggle at not only the numbers, but the rationale used to determine how many people have (lived and then) died. If we truly had a way to know the exact number of people who have lived and died, what amazing insight would it give us? How could we use those numbers to improve our world?

And so, while reading these articles, I was strangely reminded of the old joke told at a cemetery:

“How many dead people are in there?”

The answer: “All of them.”


naci erdem said...

If anybody believes in systematic re-incarnation, a quick calculation for the end of our today's world" could be made with three variables:
i)number of systematic reincarnations (i.e. 12 or 13 for many);
ii)number of people that will exist when the world population stabilizes (expected to happen around A.D. 2200 at 10 Billion inhabitants);
iii) number of people that died until now (i.e. say around 105 Billion people).

You can easily say that the number of stabilization is the total number of existing "karma"s.

Say, if there are 12 reincarnations (worst case as far as i am informed - and i can be wrong on this information), then the system has to finish when the total number of 120 Billion dead people are reached.

If today, we have approx
l05 million total dead people by 2007, it means 15 Billion more have to die for "system" finalization.

My worst case calculation is:
at least 125 years from 2007 --> 2132 is the approx. date of "dooms day".

If number of reincarnations are systematically 13 for all karmas (seems more logical to me as the "dooms day" date would pass the expected population stabilization date of 2200); then the approx "game over" date is in the range of 3.5 generations
from 2007 arriving at approx year of 2240 (worst case).

What good is it to know "the date" ? Perhaps, your great grand children might plan their pension funds more carefully. Moreover, with this info, perhaps your "Evangelist neo-cons" would leave the world in peace until 2200s, giving at least "our generation" time to develop a more rational, sustainable, humane and "up-lifting" global system. Who knows ?