FIRST GLOBAL FORUM ON HUMAN DEVELOPMENT
29-31 July 1999 . United Nations Headquarters . New York
Comment on "Innovation in Measuring Human Development"
Jean-Christophe Dumont
Université Laval, Canada
These findings along with the numerous that one can find in Human Development Reports and National Human Development Reports demonstrate the obviousness: growth in per capita GNP, because it can be consistent with rising levels of poverty, sustained inequity and violation of human rights, may not imply increases of well being for the majority of the population.
Some of the papers presented also advance toward a better comprehension and quantification of Human Development. We must thank the authors for their contributions in that direction.
As the topic of the session relates to innovation in measuring Human Development let me spend some time to recall the process that has led to the Human Development Index (HDI) construction. As Professor A. Sen brilliantly explained in his introductory speech, the HDI concept arises from the conjunction of two factors : a need and a vision.
The need has been recognized for a long time. For instance, Nordhauss and Tobin (1972) tried to correct GNP to be a measurement of economic welfare. With this approach GNP is adjusted by subtracting military expenditures and by taking into account the cost of pollution, crimes and so forth. It also adds an estimated value of leisure time and durable consumption services.
The use of social indicators to evaluate development progress emerged even before the late 1960’s. At this time the analysis often relies on single variables and usually refers to input rather than output measurement. Hicks and Streeten (1979) clearly demonstrate the limits of this approach. Since then, a great amount of work has been devoted to computing composite indexes based on achievements. The interest for such calculations relies on factors like ease of use or knowing all facts by one crucial word.
The United Nations Research Institute for Social Development undertook deciding work in this area. The Physical Quality of Life Index developed by Morris (1979) and the General Index of Socioeconomic Development by Mc Granaham et al. (1985) are the most well known examples of this pioneer work. The Human Development Indicators concepts are grounded on these experiments. Its strength and weakness have already been widely discussed so it is not necessary to come back on it.
If HDI stands the test of time and finally gains the acceptance of all, it is much because in a progressive process the research program on Human Development has succeeded to adapt and to resolve most of the problems or at least has recognized and internalized the criticism. On the contrary to a degenerating research program most of the work on Human Development is still undertaken within the hard core of the theory. The constant ongoing sophistication of the HDI measurement, disaggregation of HDI and the inclusion of new dimensions such as gender are typical examples. The Gender Development Index (GDI) is undoubtedly one of the most remarkable steps toward the quantification of gender inequalities at the macro level. As far as I know this work has been widely acknowledged. Partly this is due to the fact that GDI is a direct extension of HDI and thus it relies on its legitimacy. In that sense, innovation in measuring human development is very stimulating and should be pursued.
But there is also another kind of innovation in measuring human development that relates to issues addressed. The Human Poverty Index (HPI) and security index, as the one presented here by M. Rodrigo Martinez, are such examples. The problem that occurs in this case is that the construction of the indicator does not and cannot rely directly on HDI concept. It has to innovate not with respect to HDI but to the current theory and practice of the field.
Let me concentrate on HPI. First of all it is important to realize that for poverty the question is neither as simple nor as natural as it was for welfare measurement. To address poverty issue, the Human Development Report proposes an index that combines different dimensions of deprivation. However it is not clear to me that such a simple construction would be sufficient to produce a proper poverty index and that it does improve previous calculations.
Still I am convinced that it is worth emphasizing the importance of being able to gauge poverty through measurements other than those based on income. Nevertheless by its construction, the HPI presents limitations that make quantification erroneous and limits its operational ability capacity. There are three main problems: one of definition, one of measurement and one of the sensibility of statistical evaluation.
1- The individual aspect of human poverty is not defined. In fact the method of calculating the aggregate index is applied with no concrete definition of what is poverty at the individual level. Although it is only at the individual level that A. Sen’s concept of capability and functionning applies.
2- As it stands the HPI can not be interpreted as a percentage applicable to the whole population. In other words it does not tell us how many poor there are. There, we must recognize the great effort that has been made in the recent reports to avoid misleading interpretation, but still this property is certainly undesirable.
3- The availability of data in numerous fields is problematic. This remark also applies to HDI but in a much more loose way because of the nature of the statistical needs.
Thus I think that if the construction of a relevant HPI would be a great innovation in measuring Human Development, most of the work is still ahead. It covers methodological reflections as well as statistical collection.
To conclude I want to say once
again that there are two kinds of innovation in measuring Human Development.
One is to improve HDI and another is to develop new tools to address new
issues. Both are desirable but they imply different strategies of research.
In other words I am not certain that the recipe that has worked so well
in the case of HDI, that is simplicity, could be automatically reproduced
to handle all the issues that Human Development has to face.