Unfair inequality and growth*

• Published date: 01 October 2023
• Author: Gustavo A. Marrero,Juan G. Rodríguez
• Date: 01 October 2023
• DOI: http://doi.org/10.1111/sjoe.12531

Scand. J. of Economics 125(4), 1056–1092, 2023

DOI: 10.1111/sjoe.12531

Unfair inequality and growth*

Gustavo A. Marrero†

Universidad de La Laguna, 38200 San Crist´

obal de La Laguna, Santa Cruz de Tenerife, Spain

gmarrero@ull.edu.es

Juan G. Rodr´

ıguez‡

Universidad Complutense de Madrid, 28040 Madrid, Spain

juangabr@ucm.es

Abstract

Fighting against economic inequality is one fundamental social goal in the agendas of most

governments. However, recent studies highlight that people actually prefer unequal societies, as

they accept inequality generated by an individual’s effortand wish to reduce only unfair inequality

(generated by factors beyond an individual’s control). This distinction might help to explain

the fundamental unsolved question about whether inequality is good or bad for growth: unfair

inequality (UI) could be growth-deterring, while fair inequality (FI) might be growth-enhancing.

We derive a reduced-form growth equation from a stylized overlapping-generations model

with human capital that includes FI, UI, and poverty. Then, using an instrumental variable

approach, we show for alternative samples and inequality measures at the worldwide level that

the estimated coeff‌icient associated with UI is always negative, while the coeff‌icient of total

inequality increases when UI is included in the regression. Moreover, we f‌ind that poverty

mediates this relationship because the higher the poverty rate, the smaller the impact of either

type of inequality on growth.

Keywords: Unfair inequality; growth; poverty; human capital

JEL classif‌ication:D63; O40; E24

1. Introduction

A fundamental social goal at the center of the political agenda of most

countries is to f‌ight against inequality. However, despite this concern, recent

*We acknowledge comments and suggestions from the participants at the INET Oxford Research

Seminar, the International Conference on Equal Chances: equality of opportunity and social

mobility around the world (Bari, Italy), Meetings of the Society for the Study of Economic

Inequality (ECINEQ; New York and Paris), and seminars at the Universidad de Gerona and

Universidad Pablo de Olavide. This paper has received f‌inancial support from the Ministerio

de Econom´

ıa y Competitividad of Spain (Marrero through project PID2019-107161GB-C33

and Rodr´

ıguez through project PID2019-104619RB-C42), and from Comunidad de Madrid

(Spain) under project H2019/HUM-5793-OPINBI-CM. All views, and any remaining errors or

omissions are our sole responsibility.

†Also aff‌iliated with IUDR, CEDESOG, and EQUALITAS.

‡Also aff‌iliated with ICAE, EQUALITAS, and CEDESOG.

c

2023 The Authors. The Scandinavian Journal of Economics published by John Wiley & Sons Ltd on behalf of F¨

oreningen

f¨

or utgivande av the SJE.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License,

which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial

and no modif‌ications or adaptations are made.

G. A. Marrero and J. G. Rodr´

ıguez 1057

studies highlight that when people are asked about the ideal distribution, they

actually prefer unequal societies (Starmans et al., 2017). This preference is

observed in a wide range of countries and across people with opposite political

views (Norton and Ariely, 2011; Kiatpongsan and Norton, 2014). These two

sets of f‌indings seem to be contradictory, but they can be reconciled through an

empirical fact: people are not concerned with economic inequality, but with

economic unfairness. When studies distinguish carefully between fairness

and equality, they f‌ind that people choose fairness over equality (Baumard

et al., 2012; Sloane et al., 2012). This distinction is critical because these

concepts may have different effects on growth: fair inequality could increase

growth while unfair inequality could harm growth.

The recognition that fairness and equality are different concepts is

important for two reasons. First, public policy should focus on the reduction

of unfairness – not inequality – in society. Because the problem for people

is economic unfairness, political action should frame the problem of relating

the existing distribution of outcomes to factors such as merit and initial

circumstances (Sugden and Wang, 2020). Second, inequality in the distribution

of any economic outcome can be thought of as resulting from a combination

of fair and unfair processes, which may have different effects on growth.

Indeed, an important challenging question in economics – over which there

is still no consensus – is whether inequality is good or bad for growth

(Panizza, 2002; Banerjee and Duf‌lo, 2011; Voitchovsky, 2011; Berg and

Ostry, 2017; Milanovic and Van der Weide, 2018).1

In this paper, our central hypothesis is that the reason for this ambiguous

result is that unfair inequality (UI) and fair inequality (FI) have opposite

impacts on growth: UI is growth-deterring, while FI is growth-enhancing

(World Bank, 2006; Marrero and Rodr´

ıguez, 2013; Ferreira et al., 2018; Aiyar

and Ebeke, 2019).2Making this distinction helps to explain the ambiguous

inequality–growth relationship (see footnote 1), as growth-enhancing channels

can be associated with FI, while growth-deterring channels can be directly

linked to UI. In addition, because inequality and poverty are different

but related aspects of the income distribution, we must consider whether

1The lack of consensus is attributed to the coexistence of a variety of channels through which

inequality is affecting growth. The positive channels are related to the incentives for saving

and investing (Kaldor, 1956; Barro, 2000), asymmetric information (Mirrlees, 1971), and

productivity premiums (Goldin and Katz, 2008; Mankiw, 2013). The negative channels are

related to imperfect capital markets (Banerjee and Newman, 1993; Galor and Zeira, 1993),

political economy issues (Alesina and Rodrik, 1994; Stiglitz, 2012), and the development

process (Dasgupta and Ray, 1986). An additional complication is that estimated channels may

change over time (Blotevogel et al., 2020).

2In the literature on inequality of opportunity, a similar hypothesis (in terms of inequality

of opportunity and inequality of effort) has been named the “cholesterol hypothesis”

(Ferreira, 2007).

c

2023 The Authors. The Scandinavian Journal of Economics published by John Wiley & Sons Ltd on behalf of F¨

oreningen

f¨

or utgivande av the SJE.

1058 Unfair inequality and growth

poverty mediates in this relationship between FI, UI, and growth. Hence, our

second main hypothesis is to analyze whether poverty affects the impact that

both sources of inequality have on growth, and to test whether poverty is

growth-deterring (Ravallion, 2012; Marrero and Serv´

en, 2022).3We analyze

these aspects from both a theoretical and an empirical perspective.

To reach our goals, we f‌irst need to conceptualize UI in an objective way.

For this task, following Brock (2020), we use the concept of inequality of

opportunity (Roemer, 1998), which says that inequality due to responsibility

factors, such as effort, is fair, while inequality due to circumstances over

which one has no control is not. Consequently, we use a measure of inequality

that isolates the unfair portion of overall inequality by focusing on the

inequality that can be attributed to factors beyond a person’s control, such

as race, place of birth, health endowments, and macroeconomic conditions

(Rodr´

ıguez, 2008; Marrero and Rodr´

ıguez, 2012). In a fair society, factors

at birth should not affect an individual’s outcomes, although they have been

shown to heavily inf‌luence outcomes for disadvantaged groups (Altonji and

Blank, 1999; Bertrand and Mullainathan, 2004).

We f‌irst build a model that combines wage determination and human capital

accumulation (Glomm and Ravikumar, 1992) with unfairness (Roemer, 1993;

Fleurbaey, 2008) and poverty traps (Azariadis and Stachurski, 2005). The

economy is populated by a continuum of dynasties, where effort is a

non-monetary factor that generates disutility but is needed to accumulate

human capital. A dynasty is def‌ined as a common individual who lives for two

periods (childhood and adulthood) and gives birth to another individual during

adulthood, so the population remains constant over time. For tractability,

parents care about the total amount of resources that they leave to their

offspring (warm-glow preferences) and not about the offspring’s utility. Each

dynasty is characterized by an initial level of human capital, exogenous

factors (beyond the individual’s control such as race, place of birth, or

macroeconomic conditions) that affect human capital productivity, and an

idiosyncratic parameter of preference for effort, which is assumed to be

uncorrelated with any other characteristic of the dynasty. We associate

inequality of exogenous factors to UI, and inequality of preferences to exert

effort to FI.

To consider the existence of a poverty trap, which is of especial relevance

when modeling less-developed countries, the human capital accumulation

process is assumed to be non-convex. In this setting, each dynasty faces two

3The literature has studied a variety of mechanisms through which poverty may harm growth,

most of them based on the existence of poverty traps (for a survey, see Azariadis and

Stachurski, 2005). The idea is that below a certain level of income or wealth, individuals are

too poor to afford the investments in human or physical capital, or the technologies necessary to

raise their future levels of human capital and income.

c

2023 The Authors. The Scandinavian Journal of Economics published by John Wiley & Sons Ltd on behalf of F¨

oreningen

f¨

or utgivande av the SJE.