A Spatial Model of Internal Displacement and Forced Migration

• DOI: 10.1177/0022002720958470
• Published date: 01 February 2021
• Author: Javier Gardeazabal,Jon Echevarria-Coco
• Date: 01 February 2021
• Subject Matter: Articles

Article

A Spatial Model of

Internal Displacement

and Forced Migration

Jon Echevarria-Coco

1

, and Javier Gardeazabal

1

Abstract

This article develops a spatial model of internal and external forced migration. We

propose a model reminiscent of Hotelling’s spatial model in economics and Schel-

ling’s model of segregation. Conflict is modeled as a shock that hits a country at

certain location and generates displacement of people located near the shock’s

location. Some displaced people cross a border, thus becoming refugees, while

others remain as Internally Displaced Persons (IDPs). The model delivers predic-

tions about how the fractions of a country’s population that become refugees and

IDPs ought to be related with the intensity of the shock, country size, terrain rug-

gedness and the degree of geographical proximity of the country with respect to the

rest of the world. The predictions of the model are then tested against real data

using a panel of 161 countries covering the period 1995-2016. The empirical evi-

dence is mostly in line with the predictions of the model.

Keywords

internal displacement, forced migration, spatial model, conflict

Forced displacement occurs when a group of people are obliged to leave their home

location unwillingly. According to the internationally agreed definitions (see

UNHCR 2006, 2011; Ka¨lin 2008), displacement can be caused by armed conflict,

1

The University of the Basque Country UPV/EHU, Bilbao, Spain

Corresponding Author:

Javier Gardeazabal, The University of the Basque Country UPV/EHU, Lehendakari Aguirre 83, 48015

Bilbao, Spain.

Email: javier.gardeazabal@ehu.eus

Journal of Conflict Resolution

2021, Vol. 65(2-3) 591-618

ªThe Author(s) 2020

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0022002720958470

journals.sagepub.com/home/jcr

lack of civil liberties or political rights and natural disasters. This article analyzes

just conflict induced displacement and focuses on two gro ups of forcibly displaced

persons: refugees and Internally Displaced Persons (IDPs). The distinguishing

feature between conflict originated refugees and IDPs is whether they cross an

international border. As they have a common cause, it seems logical to study both

groups of people simultaneously. However, the literature on forced displacement is

mainly divided into refugee studies on the one hand and internal displacement

studies on the other. Simultaneous analysis of refugees and IDPs is very scant (see

Moore and Shellman 2007; Bohra-Mishra and Massey 2011, for notable excep-

tions), and therefore little is known about factors that determine whether displaced

individuals will become refugees or IDPs and the transition between the two

statuses. Our goal is to contribute to this literature a theoretical model where armed

conflict causes displacement and spatial factors determine the divide between

internal and external forced migration.

A real example using the cases of Burundi and D.R. of the Congo will help us

motivate and delineate the extent of our research. From 2000 to 2016 both countries

experienced periods of armed conflict that generated large forced displacement

movements. The yearly average stock of forced displaced people during this period

was 1,719,425 Congolese and 379,338 Burundi people, 2.61%of D.R. of the Congo

population and 4.93%of Burundi’s. The percentage of population displaced in

Burundi almost doubled the percentage in D.R. of the Congo. The split between

refugees and IDPs was also very different. While in Burundi refugees were 4.22%of

the population and IDPs account for 0.71%, IDPs in D.R. of the Congo added up to

1.87%of the population, more than doubling the 0.75%share of refugees. The goal

of our study is to understand the role conflict intensity and geography play in

determining the fractions of population forcibly displaced and their split into refu-

gees and IDPs. Intuitively, more intensive conflicts ought to generate a larger share

of the population displaced and that is the case in the example: the average conflict

Intensity Level indicator from UCDP-PRIO armed conflict dataset was 0.70 in

Burundi and 0.53 in and D.R. of the Congo, so the country with more intense conflict

has the largest share of population displaced. In our model, spatial variables have a

bearing in determining total displacement and the split into internal and external.

Take for instance territorial size, D.R. of the Congo is eighty eight times the size of

Burundi. The model suggests that, as a result of this size difference, a conflict of

given intensity would affect a larger share of the population in Burundi than in its

neighboring D.R. of the Congo. Furthermore, the model also suggests that, as a result

of the size difference, displaced people would be more likely to cross a border to a

neighboring country in Burundi than in D.R. of the Congo, so the share of refugees

would be greater in Burundi. Although territorial size is certainly not the only factor

contributing to explaining the differences in IDP and refugee shares between Bur-

undi and D.R. of the Congo, intuition suggests that it can play a significant role. In

addition to conflict intensity and territorial size, there are other determinan ts of

displacement. Arguably, differences in the type and nature of conflicts can also

592 Journal of Conflict Resolution 65(2-3)