Interregional inequality in Africa, convergence, and multiple equilibria: Evidence from nighttime light data

• Published date: 01 May 2022
• Author: Kacou Yves Thierry Kacou
• Date: 01 May 2022
• DOI: http://doi.org/10.1111/rode.12851

918 

|

     wileyonlinelibrary.com/journal/rode  Rev Dev Econ. 2022;26:918–940.

© 2022 John Wiley & Sons Ltd

Received: 27 October 2020 

|

  Revised: 3 December 2021 

|

  Accepted: 4 December 2021

DOI: 10.1111/rode.12851  

REGULAR ARTICLE

Interregional inequality in Africa, convergence,

and multiple equilibria: Evidence from

nighttime light data

Kacou Yves ThierryKacou

Department of Economics, University of 

Ulsan, Ulsan, Republic of Korea

Correspondence

Kacou Yves Thierry Kacou, Department 

of Economics, University of Ulsan, 

93 University- Daero, Namgu, Ulsan, 

Republic of Korea.

Email: kacou1988@mail.ulsan.ac.kr

Abstract

Understanding the dynamics of interregional inequality 

in Africa  is important  when formulating  development 

policies. In  this paper,  we investigate  the convergence 

pattern of  regional inequality  within 49  African coun-

tries  over  the  period  1992– 2012  using  satellite  night- 

light- based  inequality  proxies.  We  find  evidence  of 

multiple  equilibria  in  regional  income  inequality. 

Countries  belonging  to  clubs  with  high  regional  in-

equality level do not show any tendency to converge to-

ward the clubs  with low regional inequality, indicating 

that these countries are indeed stuck in regional income 

inequality traps.  The idea  of regional  income  inequal-

ity traps  is also  confirmed by a  distributional dynamic 

analysis  that  uncovered  evidence  of  regional  income 

inequality  persistence  with  a  limited  degree  of  mobil-

ity between  convergence clubs.  From the  convergence 

analysis, we found that the initial conditions in the eco-

nomic  development,  natural  resource  rent,  urbaniza-

tion,  political  institutions,  development  aid  programs, 

and infrastructure  are very  important determinants  of 

regional income inequality. Our findings provide policy 

makers some direction for future economic policies.

KEYWORDS

Africa, club, convergence, night light, regional income inequality

JEL CLASSIFICATION

C2; C3; R10; E01; O55

|

919

KACOU

1

|

INTRODUCTION

Despite the  strong economic  growth in many  African countries,  poverty and  human develop-

ment indicators  have not  progressed as expected,  and the  continent remains  characterized by 

high and  increasing inequality. Amid the challenging  issues related to the  increase in inequal-

ity, the spatial dimension of  inequality has fueled a renewed interest among academics  and the 

general public  (Beegle et al., 2016).  Recent evidence witnesses  a broad unequal  distribution of 

income and  opportunity across subnational  regions in Africa  (Boone & Simson, 2019). For in-

stance, the Central African  Republic, Sudan, and Libya experience a  high level of interregional 

inequality. The Congo  has experienced  an increase in  interregional inequality  over the period 

1992– 2012 (Hirte etal.,2020). Regional income inequality remains therefore a challenging issue 

at least  for two  reasons. First, the  interregional inequality led  to undesirable  interpersonal in-

equality (Kanbur  & Venables,2005). Second, the  regional inequality is responsible  for internal 

conflict, especially civil wars (Lessmann,2016).

Research studies on regional inequality  in Africa focus mostly on its dynamics  and how this 

dynamics  is  an  important  policy  concern  (Brown  &  Langer, 2009;  Simson,  2018;  Shimeles  & 

Nabassaga, 2017). However, much  less is known about  the convergence in regional income  in-

equality within countries in Africa. To our knowledge, Mveyange (2015) investigates the regional 

income inequality convergence by  employing luminosity data in 10  sub- Saharan African coun-

tries. The author found  a beta and sigma convergence  in regional income inequality. While the 

paper of Mveyange is helpful in understanding regional income inequality pattern, there are still 

some areas  that need to  be investigated. For example,  given the potential  heterogeneity across 

African countries in terms of interregional income inequality and the availability of new reliable 

statistical data, it is important to revisit the existence of overall convergence or to test whether the 

countries with high regional inequality become entangled in clubs of  convergence that stagnate 

or even  deteriorate over time. The  main aim of  this paper  is to reexamine the  existence of  the 

convergence pattern in regional inequality within African countries and possible factors driving 

the convergence formation.

This study  is part of  the research  that extended the  neoclassical convergence hypothesis  to 

other  fields,  such  as  income  inequality. The  fundamental  question  that  this  literature  has  at-

tempted to answer is as follows: does a country tend to converge in long- run levels of inequality? 

One of the most important theoretical predictions on income inequality convergence dates back 

to Bénabou (1996). For Bénabou, the neoclassical growth models imply convergence of the entire 

distribution of  income. The  first and the  second moments  of a  country's distribution are  both 

important  when  we study  the  convergence  process  (Chambers  &  Dhongde, 2016).  Extending 

the notion of convergence to other moments of the distribution implies that countries with sim-

ilar fundamentals  will eventually  converge toward the  same distribution of  income (Dhongde 

&  Miao, 2013).  The  idea  of   convergence in  regional  income  inequality  has  also  been  empiri-

cally  investigated. The  earlier  studies  on  convergence  in  regional  income  inequality  covered 

different samples  over the world  and employed different  measures of  convergence  (Ezcurra & 

Pascual,2005; Lessmann & Seidel,2017; Mendoza- Velázquez etal.,2020; Tian etal.,2016).

Previous studies revealed a lack of research investigating the convergence club under a variety 

of possible transition paths in  regional income inequality. This  gap of  literature  constitutes an 

additional motivation to undertake an analysis of club convergence in regional income  inequal-

ity, particularly  in Africa.  Indeed, understanding  the convergence  pattern in  regional income 

inequality is important for implementing efficient policies. For instance, a common development 

policy would be optimal if regional income inequality within  countries in Africa converges over