The Analysis of Well‐Being Using the Income‐Adjusted Multidimensional Synthesis of Indicators: The Case of China

• Published date: 01 September 2021
• Author: Mario Biggeri,Luca Bortolotti,Vincenzo Mauro
• Date: 01 September 2021
• DOI: http://doi.org/10.1111/roiw.12488

© 2020 International Association for Research in Income and Wealth

684

THE ANALYSIS OF WELL-BEING USING THE INCOME-ADJUSTED

MULTIDIMENSIONAL SYNTHESIS OF INDICATORS: THE CASE OF

CHINA

by Mario biggeri* and Luca bortoLotti

University of Florence

AND

Vincenzo Mauro

University of Macerata

In multidimensional indexes employed to measure well-being and deprivation, income is sometimes

included and sometimes excluded. The aim of this paper is to reconsider the role of income in the meas-

urement of multidimensional well-being by recognizing that it can indirectly contribute to individual

well-being, even if it is not regarded as a goal in itself. This involves introducing a new composite index:

the Income-adjusted Multidimensional Synthesis of Indicators (I-MSI). To illustrate this index, indi-

vidual-level data from the 2015 China Household and Nutrition Survey (CHNS) are analyzed. Results

confirm the soundness of I-MSI approach as a multidimensional aggregation method and show that it

can capture disparities across Chinese macro-regions and variations among different segments of society.

JEL Codes: I31, O53

Keywords: multidimensional well-being, composite indexes, Income-adjusted Multidimensional

Synthesis of Indicators, China

1. introduction

In recent decades, academic research has increasingly looked beyond tradi-

tional income-based measures of well-being and deprivation to embrace a multidi-

mensional perspective (Maasoumi, 1986; Sen, 1999; Chakravarty and D’Ambrosio,

2006; Alkire, and Foster, 2011; Bossert et al., 2013; Mazziotta and Pareto, 2013; see

Burchi et al., 2018 for a review).1 Three main issues emerge in the debate on how to

1A similar debate took place in several international institutions and forums from the end of the

1980s including the UNDP, FAO and WFP. More recently, the Commission on the Measurement of

Economic Performance and Social Progress (Stiglitz et al., 2009), the 2030 Agenda for Sustainable

Development (SDG 1.3), the OECD Better Life Index. Moreover, some National Statistical Offices

such as CONEVAL in Mexico, ISTAT in Italy and Eurostat in the EU, have developed multidimen-

sional measures of well-being that include the income dimension.

Note: The authors want to thank the China Institute for Income Distribution (CIID), and The

National University Centre for Applied Economic Studies (c.MET05). We are grateful to David A.

Clark and Camilla Guasti for their suggestions and comments. Moreover, we are grateful to three anon-

ymous referees and editor for constructive comments. This research did not receive any funding from

public agencies, commercial organizations or the not-for-profit sector.

*Correspondence to: Mario Biggeri, Department of Economics and Management, University of

Florence, Via delle Pandette, 9 50127 – Firenze, Italy (mario.biggeri@unifi.it).

Review of Income and Wealth

Series 67, Number 3, September 2021

DOI: 10.1111/roiw.12488

bs_bs_banner

Review of Income and Wealth, Series 67, Number 3, September 2021

685

© 2020 International Association for Research in Income and Wealth

build appropriate composite indexes: firstly the selection of dimensions, secondly

the choice of weights and, thirdly, the choice of the aggregation method, which is

the main focus of this paper. The selection of dimensions is dealt with in Section3;

while the choice of weights, traditionally based on normative assumptions or

empirical approximations, is simplified by adopting equal weights. However, it is

important to acknowledge that different weighting systems can be applied. For

instance, following Maasoumi (1986) it is possible to use data to weigh factors of

well-being (empirical weights). For a review of this approach, see also Decancq and

Lugo (2013). For Maasoumi and Racine (2016) heterogeneous weights are found

ex-post, based on the distribution of outcomes across different populations in their

sample. In contrast, Alkire and Foster (2011) defend the use of equal weighting,

thus adopting a normative approach.2

Another strand of literature focuses on the role of income in measuring mul-

tidimensional well-being (Maasoumi, 1986; Anand and Sen, 1997; Kovacevic,

2010). Income is sometimes included and in other cases excluded from the selec-

tion of relevant multidimensional indicators depending on whether it is viewed as

a means or as a goal in its own right (Sen, 1999). For instance, income has always

been a component in the Human Development Index (HDI) as a proxy for living

standards. Conversely, in the Multidimensional Poverty Index (MPI) and Multiple

Overlapping Deprivation Analysis (MODA), assets, rather than income, are taken

into account. Nevertheless, the role of income with respect to well-being is still

unresolved and open to debate (Laderchi, 2003; Pogge, 2010).

In terms of the aggregation method, two distinct approaches are usually

considered. One approach provides a continuous measure of multidimensional

well-being, eventually resorting to arithmetic means across dimensions or more

sophisticated operations that penalize the heterogeneity of outcomes across dimen-

sions. A well-known example is the post-2010 HDI, which now adopts the geomet-

ric mean (Klugman et al., 2011). The rationale is to make an allowance for greater

homogeneity in outcomes across dimensions as a valuable achievement in itself. To

improve these measures, Mauro et al. (2018) suggest a new aggregation method to

address problems associated with the geometric mean (Klugman et al., 2011). The

second approach utilises a binary distinction between multidimensionally poor

and non-poor individuals. This is obtained by considering a set of deprivation

indicators that can be aggregated through the union approach, the intersection

approach, or more often, using a dual cut-off procedure. This procedure has been

used for measuring social exclusion (Chakravarty and D’Ambrosio, 2006), estimat-

ing the overlap in deprivations among children (Gordon et al., 2003 and UNICEF

MODA, de Neubourg et al., 2013) and other multidimensional poverty indexes

(e.g. MPI, Alkire and Foster, 2011).

The objective of this paper is to try to reconcile the role of income in the mea-

surement of multidimensional well-being by incorporating it as a means to other

ends and not as an end in itself, and by introducing a new composite index: The

Income-adjusted Multidimensional Synthesis of Indicators (

I−MSI

).

2It is beyond the scope of this paper to consider empirical weights within our approach. Future

investigations may consider combining our technique with empirical weights so that preferences for

different dimensions may vary across different groups of individuals.