Costs of civil war and fragile states in Africa

• Author: John Paul Dunne,Nan Tian
• DOI: http://doi.org/10.1111/rode.12612
• Published date: 01 August 2019
• Date: 01 August 2019

1220

|

wileyonlinelibrary.com/journal/rode Rev Dev Econ. 2019;23:1220–1237.

© 2019 John Wiley & Sons Ltd

1

|

INTRODUCTION

Civil war has been both prevalent and increasingly persistent during the past sixty years and yet, until

relatively recently, had received little attention from economists (Fearon, 2004). Blattman and Miguel

(2010) estimate that since 1960, 20% of all countries have experienced at least 10 years of civil con-

flict and, given the important role civil conflict plays in shaping nations, it ought to be central to the

study of economic development. It has not been, and while there has been an impressive growth in the

literature on the causes and duration of civil conflict (e.g. Collier & Hoeffler, 2004; Fearon & Laitin,

2003), research on the consequences of civil conflict in developing economies has been less evident

(Smith, 2014).

Given that civil war tends to break out in the world's poorest regions and the consequences inevi-

tably affect all parts of the economy, understanding the economic impact of civil war and its costs is

DOI: 10.1111/rode.12612

REGULAR ARTICLE

Costs of civil war and fragile states in Africa

John PaulDunne1

|

NanTian2

1School of Economics,Southern Africa

Labour and Development Research

Unit,University of Cape Town,

Rondebosch, South Africa

2Stockholm International Peace Research

Institute, Stockholm, Sweden

Correspondence

Nan Tian, Stockholm International Peace

Research Institute, Stockholm, Sweden.

Email: nan.tian@sipri.org

Abstract

This paper provides an analysis of the costs of conflict for

the countries that are most affected by it, namely low‐in-

come and fragile countries in Africa. It provides an analy-

sis of the impact of conflict on economic growth using a

panel of African countries and investigates the differences

between those categorized as fragile and others. It finds that

there are significant differences, with fragile countries most

affected by conflict. It also considers the potential spillover

effects of conflict and finds that fragile states also suffer

more from conflicts within neighboring states. Finally, it

considers whether the results are robust to endogeneity of

the conflict variable. Using instrumental variable estimation

confirms the negative effect of conflict and the larger effects

on fragile states and gives even larger coefficient estimates.

JEL CLASSIFICATION

D74; C3

|

1221

DUNNE and TIAN

an important area of research, with some arguing that the destructive forces of civil conflict are large

enough to explain the income gap between the poorest and richest nations (Collier, Elliot, Hegre,

Reynoal‐Querol, & Sambanis, 2003). Collier (1999) estimated that during a war, the growth rate of

a country is typically reduced by 2.2% per year. These losses can continue post‐war, with Cerra and

Saxena (2008) finding output to decline by about 6% after a civil war.

While the general consensus has been that civil conflict is detrimental to economic development,

this impact could be even greater for what have been termed “fragile states” (Rodrik, 1999), but there

has been no substantive attempt to investigate whether this is actually the case.

This paper provides such an analysis comparing the impact of conflict in fragile and non‐fragile

countries, with a focus on growth and also considering the potential impact of spillovers from con-

flicts in neighboring countries.

The next section offers a review of the cost‐of‐conflict literature. Section 3 considers the nature of

fragile states and presents the data. This is followed by the empirical analysis using a growth model

framework and then extending it to consider conflict spillovers. Section 4.1 considers the issue of

endogeneity. Section 5 provides some conclusions.

2

|

COST OF CONFLICT

While civil conflict is a major concern for developing countries, it is only relatively recently that econo-

mists have come to see it as an important area of research. The contribution of economists to an area that

had been the preserve of international relations led to a considerable debate over “greed versus griev-

ance” and a recognition that the causes of conflict are as varied as the nature of conflict and the roots

of war are multifaceted, with important historical contexts. There are a number of potential factors that

can be identified, including colonial legacy; military governments and militaristic cultures; ethnicity

and religion; unequal development; inequality and poverty; bad leadership and/or polity frailties and in-

adequacies; external influences; greed, opportunity, and/or feasibility; and natural resources (Collier &

Hoeffler, 2004; Fearon & Laitin, 2003; Ross, 2006; Cunningham & Lemke, 2014). Very few conflicts

are simple, they are often a combination of factors, and this fact can have important implications for the

achievement of peace and the success of post‐conflict reconstruction policies (Dunne, 2013).

Rather less research has focused on evaluating the costs of civil conflict, despite their being poten-

tially massive. They can be wide‐ranging, both economic and social. with consequences in the short,

medium and long term (Blattman & Miguel, 2010). In the short and medium term, these include the

lives lost, the permanent injuries, the refugees, military expenditure, asset losses (i.e. destroyed capi-

tal, including human), gross domestic product (GDP) or production losses (i.e. income), trade losses.

Long‐terms costs, which are less frequently considered, include intergenerational effects, transbound-

ary effects, and environmental effects. Alternatively, the costs can be categorized as destruction and

deferred accumulation and legacy costs (Smith, 2014).

Loss of life is the immediately apparent cost of conflict. Most wars are relatively small, with most

incurring 25–999 battle‐related deaths per annum. In some datasets these would not be considered,

as the definition of an active war involves 1,000 or more battle‐related deaths per annum. But battle‐

related deaths are by no means the whole story, as in many countries conflict leads to many hidden

casualties and the devastation can mean people die for years after a conflict ends – a legacy cost of the

conflict that is often ignored.1

Deaths are only part of the violence engendered by conflict, as the existence of conflict in a coun-

try creates an environment in which all forms of violence are possible and the less “ headline” forms

of violence often remain as a legacy when the conflict is supposedly over (Brauer & Dunne, 2012).