Development Data Constraints and the Human Development Index

Until the late 1970s, social indicators' pre-eminent function lay in their application as alternatives to income as the measure of development. Since then they have increasingly been pressed into more active service, especially as guides for formulating and assessing national policy. In recent years, a great deal of effort has been devoted to theoretical and empirical research on social indicators. The publication of the UNDP Human Development Report 1990 (HDR) has aroused widespread interest among researchers, policy makers, the international development community and the general public on the use of social indicators to measure national performance with regard to human welfare and development. The HDR provides a useful compilation of existing social statistics and a helpful analysis of the factors determining differential performance concerning social progress. The centrepiece of the HDR, the human development index, has aroused a great deal of comment. Like all such indices, it suffers from some weaknesses. The present paper by Dr. Christopher Murray discusses these and suggests ways in which the index may be improved.

While applauding the HDR for attracting widespread attention to the need to monitor and assess attempts to foster human development, Murray says it "risks to be counterproductive". The human development index, which combines indicators of income, life expectancy and education into a single dimension, has been challenged by many on statistical grounds. Dr. Murray leaves these questions aside and addresses himself to the methodological and data constraints encountered in computing the three components of the index. He also proposes alternative methods of measuring progress in longevity, education and income.

Adjustments in the methods of computation will not, however, redress the inadequacies of available data. In the short term, for reasons cited in previous work by UNRISD and others and noted by the author, most of the 160 countries covered in the report cannot yet provide reliable and comparable data required to compute the index on an annual basis. Nonetheless, the index is to be recomputed and published annually. The author argues that this "gives the false impression that we know the levels of these important activities in all developing countries... [and] devalues a host of efforts that are needed to get real and timely information on mortality and education in developing countries".

Dr. Murray concludes with a plea that future versions of the HDR contain only estimates based on real data without the assumed model changes or other short-cut estimates. He also recommends that the dates of the most recent empirical estimates for each country and for each variable be clearly marked. By so doing, the HDR will highlight to the international community "the true extent of our ignorance on the current levels of important social indicators such as mortality and educational attainment and even income in real purchasing power terms".