A careful look at "A re-look at recent statistics on mortality in the context of HIV/AIDS with particular reference to South Africa"

An article recently published in this journal argues that the life expectancies (and other mortality statistics) produced by models of the HIV/AIDS epidemic in Southern Africa are inconsistent, and questions their reliability. To demonstrate the argument, the author of that paper derived empirical estimates of several mortality statistics from three different sources of data and, on the grounds that the estimates of life expectancy for 2001 and 2006 are somewhat higher than is typically estimated by projection models, concludes that the empirical evidence supports the theoretical view outlined in that paper. If correct, the reasoning (and its empirical demonstration) could be construed as a strong challenge to a dominant orthodoxy surrounding the estimation of mortality statistics in an era of HIV/AIDS and offering some comfort to governments with low Human Development Indices because of the index's dependence, inter alia, on estimates of life expectancy at birth derived from such models. This paper shows how, on theoretical, methodological and empirical grounds, the reasoning and estimates in the paper are severely flawed, and thus that the conclusions drawn in that paper are unjustified.