Weighting Socio Economic indicators of Human Development

• Completion date जनवरी., 2000
• Sponsor
• Project leader A.L. Nagar
• Consultants/Other authors Sudip Ranjan Basu
• Focus

  In this paper, human development is interpreted as an 'abstract conceptual variable' which cannot be directly measured in a straight forward manner, but is linearly determined by the interaction of a large number of causal variables (socio economic indicators). An estimator of the human development index is proposed as the weighted average of principal components of the standardised causal variables, where weights are variances of successive principal com­ponents.

  natural logarithm of real gross domestic product per capita) and data for 174 countries, as in HDR 1999 of UNDP. Later the set of socio-economic indicators, which determine the human development, is enlarged. Eleven socio-economic indica­tors are used and human development indices for 51 countries, data for which are available from Human Development Report (HDR) 1999 of UNDP, and World Development Indicators (WDI) 1999 of the World Bank, are determined.

  Ranks of countries obtained by the principal component method (the only difference be­tween the two is in standardising the causal variables), and Borda ranks are almost perfectly correlated, where as HDI and Borda ranks have smaller correlation.

  In the computation of the indicators as many principal components as the number of causal variables are used. By a little rearrangement of terms, coefficients of the causal vari­ables are compiled. These coefficients, in turn, indicate the importance of different social indicators in determining the HDI.

  In the first exercise, using the same social indicators as used for computing HDI in HDR 1999, income (loge Y) is found to be the most dominant factor in determining human develop­ment. Next, in order of importance, are LE, CGER and ALR. Further, in the exercise with eleven social indicators, income is the most dominant factor. Then, in order of importance, are available health services, environment, LE, ASW (access to safe water), CGER, etc. Thus the four variables used in the UNDP exercise maintain their order of importance as loge Y, LE, CGER and ALR. It is noted that HDI in HDR 1999 assigned higher weight (two third) to ALR and lower weight (one third) to CGER to compute the index of educational attainment. The result is at variance with the UNDP assumption.