Schuyler Null has a post up on The New Security Beat on the 2010 revision of the United Nations (UNDESA) World Population Prospects, noting that this new revision suggests that by 2100 roughly 1 in every 3 people in the world will live in sub-Saharan Africa – a total of 3.36 billion people.  It is far too early to pick apart these projections, especially as the underlying assumptions used to guide their construction are not yet available to the public. Null is quite right to note:

the UN’s numbers are based on projections that can and do change. The range of uncertainty for the sub-Saharan African region, in particular, is quite large. The medium-variant projection for the region’s total population in 2100 is 3.36 billion people, but the high variant projection is 4.85 billion and the low variant is 2.25 billion.

A few preliminary thoughts, though.  I pulled up the data for a country I know reasonably well – Ghana.  Under this new revised projection, Ghana’s population is expected to reach more than 67 million by 2100.  Peak population growth is supposed to take place between 2035 and 2040, with steady declines in population growth after that.  With life expectancies projected to rise to 79 years by 2100, certainly a lot more Ghanaians will be around for a lot longer than they are today (current life expectancy is just shy of 60 years).  That said, these numbers trouble me.  First, I don’t quite see how Ghana will be able to sustain a population of this size at any point in the future – the number is just too massive.  Second, it seems to me that the life expectancy estimates and the population size estimates contradict one another – as Charles Kenny quite ably demonstrates in Getting Better, as life expectancies rise and more children reach adulthood, the general trend is to lower total fertility.  The only way Ghana’s projection can be made to work is to assume massive demographic momentum that I am not sure will play out in the face of expected declines in infant mortality and the increased cost burden for prospective parents supporting older family members for much longer than they do today. In other words, this seems to me to be a rather dire overestimation of where Ghana is going to be in the future.
Now, this is just a quick cut at what appear to be the assumptions for one country, but I worry that this potential overestimation has a certain political utility.  The Malthusian specter, however inaccurate it may be, remains a great motivator for aid and development spending.  Further, presuming massive demographic momentum requires we assume that adequate reproductive health options are not in place in places like Ghana.  Given that the monitoring of reproductive health, presumably to better direct development interventions, seems to be a large focus of UNDESA’s and other UN organizations’ mandate, they might have a bit of a built-in bias against a lower population number because such a number would presume significant progress on the reproductive health front, thus challenging the need for this particular service.  In a wider sense, it seems to revive fears of a population bomb, albeit in this case limited to Africa.  While I have no doubt that demography will be an important challenge to address in the future, I think the current numbers, even the low estimates, seem overstated.
Besides, any projection of any social process 90 years into the future probably has gigantic error bars that could encompass anything from negative growth to massive overgrowth . . . the problem here is that policy makers often fail to grasp this uncertainty, see the 100-year projection, freak out entirely and reorient the next 5 years worth of aid programming to address a problem that may not exist.