INTRODUCTION: RECONSIDERING THE POLITICS OF ‘REPATRIATION’

In April 1994 the United Nations High Commissioner for Refugees (UNHCR) formally implemented a programme to facilitate the ‘voluntary’ repatriation of an estimated 250,000 Mozambican ‘refugees’ from South Africa. By March 1995, when this phase of the programme was terminated, a total of only 31,985 persons had returned to Mozambique (UNHCR, Johannesburg, Personal Communication, August 1995). Through tracing the planning, implementation and initial impact of this ‘Voluntary Repatriation Programme’ (hereafter VRP), I examine some of the implications of this limited response for the majority of Mozambicans who remained in South Africa with fewer resources and no longer formally recognised as ‘refugees’. My assessment of the VRP, therefore, examines its relevance for social identities within the emerging state discourses of post-apartheid South Africa and postwar Mozambique.

By focusing on the relationship between local-level social processes and their broader implications, this chapter challenges established ways of understanding the UNHCR and its practice of promoting ‘repatriation’ as the most desirable ‘durable solution’ to any refugee crisis (Cunliffe 1995: 286; Harrell-Bond 1989: 41). As the ‘principal international actor for the assistance and protection of refugees’ (Cunliffe 1995: 278), the activities of the UNHCR are determined largely by international conventions that define its mandate, such as the 1951 ‘Convention Relating to the Status of Refugees’, or regional initiatives that have shaped it, such as the Organisation of African Unity (OAU) ‘Convention Governing the Specific Aspects of Refugee Problems in Africa’ (UNHCR 1994: 4-5). Recent evaluations of UNHCR operations have led to accusations that the primary role of the organisation is to enforce the will of the powerful states that constitute the principal source of funding for the organisation (Cunliffe 1995: 288-9; Hancock 1989; Harrell-Bond 1986: 188; 1989: 46).

Consequently, this ambiguous position of the UNHCR in the global context of assistance to ‘refugees’ has led to analyses of ‘repatriation’ from two broad perspectives (Allen and Morsink 1994: 2). Less critical approaches tend to provide descriptions of specific programmes as socially reconstitutive, measuring their ‘success’ in terms of the numbers of refugees who ‘benefited’ from the operations (e.g. Compher and Morgan 1991; Wood 1989).