This blog is intended to go along with Population: An Introduction to Concepts and Issues, by John R. Weeks, published by Cengage Learning. The latest edition is the 12th (it came out in 2015), but this blog is meant to complement any edition of the book by showing the way in which demographic issues are regularly in the news.

If you are a user of my textbook and would like to suggest a blog post idea, please email me at:

Wednesday, December 19, 2012

Probabilistic Population Projections

Most population projections are deterministic, in that they employ certain assumptions about future demographic trends and show the results of applying those assumptions down the road. If you change the assumptions, then the outcome changes accordingly. Another way of looking at future outcomes is probabilistically, assigning a probability to a particular outcome. This is a somewhat qualitative probability, of course, akin to the odds of a horse winning a race, which is derived from current conditions (in the case of population, this is the current age structure and fertility, mortality, and migration rates), but building in past experiences as a guide to the future (a type of Bayesian perspective). People who know a country well are in a position to offer expert opinions about how likely some outcomes are compared to others. 

Wolfgang Lutz and his colleagues at IIASA and the Vienna Institute for Demography (VID) have been pursuing this idea for several years (click here for a summary of this approach), and it has now been taken up by the demographers at the UN Population Division. They have some of the highlights available online, with the option to purchase a CD-ROM with all of the data for a nominal cost of $US 15. They note, for example, that bigger increases in population in Nigeria are more likely than smaller increases; and that aging in Europe is vastly more likely than a youthening in that region of the world. And they put numbers on those likelihoods. Although the methods of projection are not necessarily different than in a deterministic model, the interpretation of the results is different, in that it is more nuanced, and thus potentially more relevant to policy decisions.

No comments:

Post a Comment