The Geo-Politics of Forced Migration in Africa, 1992–2001

Abstract

Of the consequences of war, the movement of large numbers of displaced peoples across international borders is among the most significant. I examine the determinants of international refugee flows in Africa during the period from 1992 to 2001. I begin with a gravity model of refugee flows, which posits that levels of such flows will vary as a function of the population of the two states in question and the distance between the two. To this, I add consideration of two key factors theorized to generate refugee populations: the presence of armed conflict, and the regime type of both the source and target nations. Most important, distance between the source and the target countries is the single most important determinant of dyadic refugee flows; in addition to its direct effect, distance also mediates the effects of conflict and regime type on volumes of refugees. My findings confirm the strong influence of distance on the levels of refugee flows. Although I find clear effects for the presence of conflict, with war in the source country substantially increasing refugee flows and war in the host country decreasing them, I also find strong evidence that this effect is mediated by distance.

Keywords:

• Africa
• contiguity
• forced migration
• refugees
• war

A previous version of this paper was presented at the Annual Meeting of the Peace Science Society, Houston, TX, November 12–14, 2004 and at the Tinbergen Peace Science Annual Conference, Amsterdam, The Netherlands, June 9–11, 2004. Thanks to Will Moore, Harvey Starr, and Christopher Zorn for helpful comments.

Notes

¹ The people fleeing their homes due to a fear of persecution may flee to another country or become internally displaced within the borders of their own state. While internally displaced persons comprise a large portion of forced migrants in the world, this study focuses solely on international refugee flows and examines the characteristics of both the source and host states involved in forced migration.

² Davenport and Armstrong (2004) find no strong evidence for a negative effect of democracy on repression at lower levels of democracy. However, they find that significantly high levels of democracy are associated with lower levels of human rights abuses.

³ This measure of distance does not account for terrain and physical barriers between states, such as mountains, rivers, and deserts. These barriers can have an influence on the ability of refugees to migrate from one state to another, regardless of geographical proximity, but have not been included here for reasons of tractability. Such barriers, in addition to linear distance, would be an important consideration for future research on refugee migration.