The United Nations Development Programme publishes an influential annual indicator known as the Human Development Index. This index is used to rank countries based on a composite of life expectancy, education, and standard of living. We look at recent changes in how these three dimensions are measured and combined.

In the past the component measures were first normalised onto a zero-one scale by specifying numerical end-points for each dimension in order that they could then be added together. For example the original end points for life expectancy were 25 years (zero score) and 85 years (score of one). The end points have been re-set to different values and are somewhat arbitrary and therefore unsatisfactory, especially as they affect the relative ranking of countries.

In 2010 it was decided to change the aggregation method from an arithmetic mean to a geometric mean. “It thus addresses one of the most serious criticisms of the linear aggregation formula, which allowed for perfect substitution across dimensions” (UN Development Programme, 2010, p.216).

The geometric mean has many advantages, which include not requiring the use of end-points. We shall consider whether these advantages have been lost because some features of the old additive scheme have been retained unnecessarily.

Bibliography:

Anand, S., and Sen, A. (1993) “Human Development Index: Methodology and Measurement.” Human Development Report Office Occasional Paper 19. UNDP, New York.

Desai, M. (1991). Human development: concepts and measurement. European Economic Review, 35 (2/3), 350-357.

Fukuda-Parr, Sakiko; Shiva Kumar, AK (Editors) (2003) Readings in Human Development: Concepts, measures, and policies for a development paradigm. Oxford University Press.

Sagar, AD and Najam, A. (1998). The human development index: a critical review. Ecological Economics, 25, 249-264.

United Nations Development Programme. (2010) Human Development Report 2010: The Real Wealth of Nations.

Wolff, H, Chong, H, and Aufhammer, W. (2011) Classification, Detection and Consequences of Data Error: Evidence from the Human Development Index. Econ J.

Keywords: composite indicators; multidimensional measures; aggregation; indices

Biography: Dr. Tofallis holds a first class degree in physics, a masters in Operational Research from the LSE, and a PhD in applied mathematics. His publications include work on performance measurement using data envelopment analysis, multicriteria ranking, chaos theory, and novel methods of fitting models to data which include neutral data fitting – a units-invariant approach which, unlike regression, treats all variables on the same basis. His maximum correlation modelling method has been applied to produce sharper MRI scans of brain activity.

His research is aimed at developing methods which are easy to comprehend and yet provide advantages over existing approaches. He feels that much academic research is ignored because of its complexity, and strongly believes that methods must be simple and transparent to gain acceptance.

His interests include percentage regression, robust statistics, ranking under multiple criteria, multiple input -output efficiency measurement, and neutral data fitting.