Commentary on statistics for the historical
& projected population maps

11/22/04 - Rev 4

When I was first contacted about doing the historical population research for this assignment, ODT wanted population data for 50,000 BC and 20,000 BC, and 1,000 BC. I briefly explained to ODT's editor, Bob Abramms, that coming up with figures for those dates was problematic. There is so much conflicting data, and so much disagreement among demographers, that I was not comfortable providing population numbers for those time periods. But if you get population estimates from palaeontologists for 100,000BC, this is a group that is used to making heroic guesses based on tiny amounts of info, and I decided to go with that date as the beginning point. As you would imagine, there is still a lot of uncertainty about population statistics for 100,000 BC, particularly in Africa and the Americas. And back then there were more than one “human” species around. Homo sapiens was only just getting the better of Neanderthals and, judging by recent archaeological evidence from Indonesia, probably other species as well. Some semblance of order only really begins to kick in around 2000 years ago, which is the second map panel on the Population Map. The data here for the Birth of Christ (AD1) are from those assembled by Population Connection. You will find plenty of disagreement in the academic community about even these numbers. But we are on our way to a more accurate picture.

After that, things crept along relatively slowly. Civilisations came and went, but at the continent-scale, populations gradually increased, with most growth in Europe and Asia. The statistics for my next benchmark, at 1650, are from David Lucas of the Australian National University (Beginning Australian Population Studies, Chapter Three: World Population Growth) available online at As Lucas points out, this is the first date at which there is a consensus among demographers about world population. In fact the numbers are more or less the same as those published as long ago as 1936 by British demographer Alexander Carr-Saunders.

One of the anomalies of the population figures/images is the fact that there was no “Latin” America prior to the Spanish conquest of the Western Hemisphere. So the idea of Latin America is a construct of our century applied retroactively to an indigenous population who never heard a word of Spanish at the time of Christ or probably even until the 1400’s. (EDITOR’S NOTE: See Denis Wood’s critique of this issue, along with the dilemma of dividing the Continents of North and South America along the Rio Grande). When Denis Wood and I spoke in a conference call with Bob Abramms, we wrestled with the issue of how to divide the Americas. Some statistical analyses make a break at the Rio Grande, whereas others incorporate Mexico and even the whole of Central America in with the north. After discussions with Denis Wood, I decided the Rio Grand was an appropriate demarcation for the purposes of these comparison thumbnails.

From here on, the data are better and the changes greater. Both my historical data for 1900 and the projections for 2150 are supplied by the United Nations Population Division. The division holds regular meetings among demographers to update current data and future projections. So far, they have fairly successfully predicted the rate of the decline in the percentage growth rate of the world's population that began in the 1970s. Current thinking is that we will eventually get back to a "stable" population in which the average woman has around two children, and my 2150 projection could represent a new stable state for the world’s population. But with AIDS, global warming, WMD and other perils in the wings, nobody can tell for sure.

And birth rates round the world are currently falling so fast (it's down to about 1.2 children per woman in parts of southern and eastern Europe) that some predict that within a few decades there could be so few new babies being born that, even with rising life expectancies, global population might begin to decline in the 22nd century. We shall see, but the projection I have used is the current best guess at the UN. It, and some alternative scenarios, are available in more detail at:

Fred Pearce
London, England

(EDITOR’S NOTE: see also