McGillivray, Mark (2017). 'Yet Another Empirically Redundant Composite Development Indicator? New Evidence and Insights for the Human Development Index' Paper presented at the annual conference of the HDCA, Cape Town 2017.


Twenty-six years ago I wrote a paper that was critical of the selection of variables for the three dimension achievements on which the Human Development Index is based.Using data published in the Human Development Report 1990, the paper concluded that the extent of statistical correlation between the HDI and income per capita, and between the HDI and any one of its individual dimension achievements, was such that the HDI should be seen as empirically redundant, adding little to pre-existing empirical information, including that provided by income per capita. Although not articulated in the paper, the underlying concern was whether the HDI was able to sufficiently empirically capture, to do justice to, the rich conceptualization on which the index was based. This paper revisits the analysis of that earlier paper. In begins with replicating the very same empirical analysis, but using data from the Human Development Report 2015. Very similar empirical results to those of the original paper are obtained, leading to the conclusion that the HDI remains an empirically redundant composite development indicator.  The paper then conducts an analysis that isolates and highlights the new information provided by the HDI. In so doing, it seeks both respond to and shift attention away from the redundancy issue, seeing the glass as half full as opposed to half empty. This analysis decomposes the HDI into two achievements: those predicted by income per capita and those orthogonal to income per capita. The latter is described as non-economic human development achievement. It concludes by arguing the case that data on this achievement should be reported alongside that on the HDI and its three individual achievements.

scroll to top