Weights in Multidimensional Indices of Wellbeing: An Overview

This paper explores some difficulties encountered in multi-dimensional measures of wellbeing. In particular, it highlights the need for such indices to (i) incorporate information on individual preferences, (ii) be robust with respect to estimation assumptions given the 'curse of dimensionality' problem, and (iii) reflect complementarities and substitutabilities between dimensions. The paper proposes a procedure that enhances the Alkire and Foster (2011) Multi-dimensional Deprivation Index (AFMDI), drawing on recent work in Anderson et al. (2011), which addresses these difficulties. The approach provides the requisite flexibility in the representation of wellbeing component deprivations, whilst admitting the possibility of sub-component substitutability/complementarity in the index, and retains the ability to measure the impact of improvements/worsenings of sub-components within each category. It then provides an application to the measurement and valuation of opportunity in different domains using a unique dataset for working age adults in the U.S.. Empirical findings suggest that freedoms are substitutable, that their values depend on an individual's needs, and that complementarities if they exist are weak. Policy implications are briefly discussed. The paper then concludes that such indices are feasible to implement, and holds promise in economic applications ranging from measurement of progress in wellbeing, to the multi-dimensional assessment of poverty.

2011

Abstract: There is a widespread consensus that well-being is a multidimensional notion. To quantify multidimensional well-being, information on the relative weights of the different dimensions is essential. There is, however, considerable disagreement in the literature on the most appropriate weighting scheme to be used. Making use of a recent data set for Flanders, we compare various methods to select a weighting scheme.

Cultura, educación, sociedad, 2021

Existing well-being measures differ in terms of number and format of items, factors being measured, aggregation methods, and are not comparable. A well-being measure involves combining n-number of indicators and quality of the measure depends on properties of combining procedures adopted. The paper proposes two assumption-free aggregation methods to satisfy the desired properties of an index The paper proposes two indices of well-being in terms of cosine similarity and Geometric Mean (GM) avoiding problems associated with scaling of raw data and choosing of weights. Empirical illustration is provided on application of the proposed measures. The proposed indices give better admissibility of operations and satisfy properties like time-reversal test, formation of chain indices, computation of group mean and statistical tests for comparison across time and space. The preferred index can be constructed even for skewed longitudinal data and helps to reflect path of improvement registered by a country/region over time. The index based on GM is preferred due to wider application areas. The index can further be used for classification of countries, subgroups and even individuals with morbidity in terms of overall wellbeing values. Future studies suggested.

meetings.sis-statistica.org

Nowadays, a relevant challenge regards the assessment of a global measure of well-being by using composite indicators of different features such as level of wealth, comfort, material goods, quality and availability of education, living standard, etc. The ...

This paper questions the validity of the statistical methods currently used in computing the composite indicators of well-being from their main subcomponents. The facts that most of the weights of the principal sub-components of the composite indicators are equal, that the determinants of well-being are correlated, and that the results are interpreted primarily in terms of country ranks, point out to the appropriateness of using a rank-based method for computing the composite indicators form their sub-indexes. A comparison of the actual ranks with ranks computed as averages of the ranks of sub-indexes for three well-known indicators of well-being, Human Development Index, Legatum Prosperity Index, and Social Progress Index, shows that results are almost the same. This calls into question the use of weighted averages of actual values of sub-components, as very high values for a sub-component increases a country’s relative rank, despite much lower performance on other sub-components, ...

SSRN Electronic Journal, 2000

The multidimensional view of well-being is receiving growing attention, both in academic research and policy-oriented analysis. This paper examines empirical strategies to measure poverty and inequality in multiple domains, concentrating on two problems in the use of synthetic multidimensional indices: the weighting structure of different functionings and the functional form of the index. These problems are illustrated by comparing inequality and deprivation in income and health in the four largest countries of the EU: France, Germany, Italy and the United Kingdom.

2014

This paper explores the relevance of the variables that define well-being and human progress and makes a quantitative inquiry into the validity of three of the well-known and well-documented composite indicators of well-being: the Human Development Index (HDI), the Legatum Prosperity Index (LPI) and the Happy Planet Index (HPI). After choosing the key variables that describe most of the objective and subjective dimensions of well-being, we perform cluster analysis to come up with an optimal grouping of countries based on their multidimensional performance on well-being. A comparison of the classifications obtained with the three indexes invalidates the HPI, confirms results obtained for the HDI, and validates for the first time the LPI as a reliable measure of well-being. The optimal cluster structure yields robust results, which correct the rank discrepancies between the HDI and LPI for a large number of countries. It also proves that a robust ranking of countries based on multidimensional well-being can be achieved with a relatively small number of variables, which mitigates the risk of including variables that are not reliable and/or not available for a significant number of countries. The fact that cluster analysis generates results based on similarities between observations and not on computed values based on the aggregation of variables helps overcome problems that may occur due to the distribution of variables and increases its value as a validation method. Therefore, validation results achieved through cluster analysis are more robust and help to achieve a good check of the validity and relevance of the composite indexes, provide an objective perspective that can guide policy-makers and the public in making a fair assessment of actual levels of well-being, and avoid unfounded claims that may overstate it and delay or postpone measures to increase it.

Econometric Reviews, 2013

Sociological Methods & Research, 2007

2012

An important aspect of multidimensional wellbeing distributions is the correlation between different dimensions. We propose two indices for measuring multidimensional inequality, derived from two underlying social evaluation functions. These functions aggregate both across dimensions and across individuals. The social evaluation functions differ only with respect to the sequencing of aggregation.