International inequality in subjective well‐being: An exploration with the Gallup World Poll

• DOI: http://doi.org/10.1111/rode.12356
• Date: 01 May 2018
• Author: Leonardo Gasparini,Pablo Gluzmann
• Published date: 01 May 2018

REGULAR ARTICLE

International inequality in subjective well-being: An

exploration with the Gallup World Poll

Pablo Gluzmann

1,2

|

Leonardo Gasparini

1,2

1

Universidad Nacional de La Plata, La

Plata, Argentina

2

National Scientific and Technical

Research Council (CONICET), Buenos

Aires, Argentina

Correspondence

Leonardo Gasparini, Center for

Distributional, Labor and Social Studies

(CEDLAS), Facultad de Ciencias

Econ

omicas, Universidad Nacional de La

Plata, Calle 6, entre 47–48, 3er piso,

Oficina 312, (1900) La Plata, Argentina.

Email: gasparinilc@gmail.com

Funding information

Universidad Nacional de La Plata

Abstract

In this paper we compute inequality measures over the

distribution of a subjective well-being variable con-

structed from a life satisfaction question included in the

Gallup World Poll in almost all countries in the world.

We argue that inequality in subjective well-being may be

a better proxy for the degree of unfairness in a society

than income inequality. We find evidence that inequality

in subjective well-being has an inverse-U relationship

with per capita GDP, but it is monotonically decreasing

with respect to mean subjective well-being. We argue that

this difference might be associated to inequality aversion

in the space of utility.

1

|

INTRODUCTION

Inequality has been typically studied in the space of income, or other objective dimensions of

well-being (Atkinson & Bourguignon, 2000). However, a recent literature stresses the relevance of

analyzing individual well-being with subjective measures obtained from answers to life satisfaction

questions (Stiglitz, Sen, & Fitoussi, 2009; Deaton, 2012). If these answers were a meaningful

approximation to real levels of individual welfare, inequality in subjective well-being could be

computed. In fact, analyzing inequality in that space could have some theoretical advantages over

the typical income distribution analysis. If some income differences are the outcome of free

choices subject to similar constraints, then they should not be considered unfair. In fact, this is one

of the main insights of the growing literature on equality of opportunity (Roemer, 1998; Roemer

& Trannoy, 2015). In contrast to income, individual utility is less prone to be “contaminated”by

these issues. Two individuals with different preferences facing similar circumstances may end up

with very different incomes as a consequence of their choices, but individual utilities may not be

very different. In that framework, inequality in perceived happiness could be a better approxima-

tion to social unfairness than income inequality.

DOI: 10.1111/rode.12356

610

|

©2017 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/rode Rev Dev Econ. 2018;22:610–631.

In this paper we discuss these issues and measure inequality in the distribution of subjective

well-being, exploiting life satisfaction questions in the Gallup World Poll 2006, a survey that

includes identical questionnaires in almost all countries in the world. We also present indicators

computed over the answers to perception questions in the World Values Survey (2009). These sur-

veys allow us to have an international perspective of inequality in subjective measures of well-

being, and compare the results with those drawn from income variables. Although we are aware of

the multiple difficulties in measuring individual well-being with the few simple questions included

in general surveys, as well as the concerns about adaptation, comparability, and cardinality (Sen,

1987; Graham, 2009), we still believe that these questions include valuable information on peo-

ple’s well-being that is worth exploring. In fact, the literature on subjective well-being has been

growing at a quick pace, raising a number of interesting issues relevant to the economic develop-

ment debate (Decanq, Fleurbaey, & Schokkaert, 2015; Nikolova, 2016).

This paper makes three main contributions. First, it highlights the role of measuring inequality

in subjective well-being (SW) as a relevant tool in the analysis of social unfairness, in comparison

with the traditional income inequality framework and the increasingly influential equality-of-oppor-

tunity paradigm. Second, the paper provides estimates of inequality in SW in almost all countries

in the world based on the same question. Although the literature on SW has been booming, the

contributions on inequality in this space have been scarce, and limited to specific countries.

1

Third,

the paper reports some interesting results regarding international inequality in SW. In particula r,

we find evidence that inequality in SW has an inverse-U relationship with per capita GDP, but it

turns out to be monotonically decreasing with respect to mean SW. We argue that this difference

might be associated with inequality aversion in the space of utility.

The rest of the paper is organized as follows. In Section 2 we discuss the role of inequality

measures in SW in the analysis of social unfairness. In Section 3 we review some issues regarding

the measurement of SW, and then specify our approach and present the data. The results of mea-

suring inequality in the space of SW are shown in Section 4, where we also explore the links with

some indicators of inequality of opportunity. In Section 5 we explore the evidence of a Kuznets

curve for inequality in SW, and put forward an argument that may account for an inverse-U rela-

tionship between inequality and per capita GDP, but a monotonically decreasing association with

mean SW. Section 6 closes with some remarks.

2

|

THE ROLE OF SUBJECTIVE WELL-BEING IN FAIRNESS

ANALYSIS

There is a large literature in philosophy, political science and economics about the space in which

inequality should be measured to approximate unfairness in a society (Sen, 1973; Atkinson &

Bourguignon, 2000). Despite the richness of the theoretical debate, in practice most empirical anal-

yses estimate the degree of social unfairness by some measure of inequality of outcomes, typicall y

income or consumption. The main drawback of this simple approach is that outcomes are in part

the result of choices, and hence some differences in outcomes could be socially acceptable, raising

no equity concerns. The approach of equity as equality of opportunity is better grounded in the

philosophy literature (Arneson, 1989; Cohen, 1989), and generally more accepted by people in

opinion surveys than the equality of outcomes alternative. There is a growing literature on the

measurement of inequality of opportunity (IO) that, although promising, faces formidabl e method-

ological challenges.

2

A typical IO analysis requires identifying all the factors that affect a relevant

outcome and considers as unfair only those outcome differences that are driven by circumst ances

GLUZMANN AND GASPARINI

|

611