As it happens, some folks at UW are very involved in developing the machinery that UNPop uses to make these predictions. I haven't read the Post article, but if you want to dig in to the methodology behind it, this could be a good starting point: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3367999/ --Abie -----Original Message----- From: change-bounces@change.washington.edu [mailto:change-bounces@change.washington.edu] On Behalf Of Thomas Anderson Sent: Tuesday, July 23, 2013 11:06 AM To: Gaetano Borriello Cc: Change Group Subject: Re: [change] Interesting take on possible future of population growth we're not disagreeing. The data/model they used is from the UN (there's a link in the article). What they appear to have done is to use the median projection from the UN as to what will happen in the future, even though the UN says that they don't really know much beyond the next twenty years how things are going to play out. On Jul 23, 2013, at 10:50 AM, Gaetano Borriello wrote:
There is a lot wrong with the article. More than the fact that it isn't an analysis with explicit uncertainty is that there was an implicit assumption that the underlying model driving fertility/mortality will not change under the intense pressures these shifts will create. It is pure extrapolation for dramatic effect. However, the overall directions (the current derivatives) are a bit harder to argue with and are pointing us toward some dramatic changes.
On Tue, Jul 23, 2013 at 10:43 AM, Thomas Anderson <tom@cs.washington.edu> wrote:
doesn't the fact that there are no error bars on their projections tell you they need a little more nate silver, and a little less omg?
ps. the underlying data they used has error bars: eg., the UN fertility rate for Angola in 2050 has a 95% confidence interval of being between 1.8 and 5. not a tight bound.
On Jul 23, 2013, at 9:58 AM, Gaetano Borriello wrote:
http://www.washingtonpost.com/blogs/worldviews/wp/2013/07/16/the-ama zing-surprising-africa-driven-demographic-future-of-the-earth-in-9-c harts/#fb _______________________________________________ change mailing list change@change.washington.edu http://changemm.cs.washington.edu/mailman/listinfo/change
_______________________________________________ change mailing list change@change.washington.edu http://changemm.cs.washington.edu/mailman/listinfo/change