150 C. FAULKNER AND S. SCHIFFER The involuntary nature of repatriation and the Rohingya’s status as stateless all but ensured its failure. Alternative reasons for the repatriation’s failure can be found in both the logistical difficulties for Bangladesh – accommodating 200,000 refugees – and the bilateral issues between Bangladesh and Myanmar. Following the partition of India, both East Pakistan and Bangladesh faced difficulties in accommodating large populations of refugees. For Bangladesh, limited resources and high population density were contributing factors to stunted socio-economic development, only complicated by refugees. Thus, while the Bangladeshi government arguably sympathised with the Rohingya, it was largely in the state’s interest to keep the presence of refugees limited and temporary (Grundy-Warr and Wong, 1997). Tensions between Bangladesh and Myanmar also necessitated a speedy and comprehensive repatriation. With Burmese security forces in pursuit of fleeing Rohingya, skirmishes were reported between Burmese soldiers and Bangladeshi border posts (Kamaluddin et al., 1992). With 50,000 Burmese soldiers mobilised on the border, the Bangladeshi government concluded that the situation ‘clearly constituted a threat to peace and security’ (Kamaluddin et al., 1992, p. 26). Further, recognising that it could not afford to continue to assist the vast number of refugees, Bangladesh agreed to repatriation. Although these points suggest that host states’ inability to accommodate can contribute to failed repatriation efforts, the Rohingya’s condition of statelessness still assists in explaining why the cycle persists. By April 1992, with an estimated 250,000 Rohingya in Bangladesh, Myanmar and Bangladesh reached a bilateral agreement, signing a memorandum of understanding (MOU) which established the guidelines for repatriation. The terms included a requirement that returnees have identification documents proving their residence in Myanmar, this despite the fact that many Rohingya had any documentation confiscated prior to entering Bangladesh (Abrar, 1995, p. 11). Further, although playing a role in the latter stages of the repatriation, the UNHCR was never given any specific part on the ground early in the process. In October 1992 an agreement was reached with Bangladesh to allow UN verification of the repatriation process (Abrar, 1995, p. 12). UN monitoring began in early October 1992, but the UNHCR withdrew its support by December 1992, citing the coercive nature of the repatriation. Roughly 5000 Rohingya had been involuntarily repatriated at this point and it would not be until May 1993 that the UNCHR would sign another agreement with Bangladesh, this time allowing the agency free access to transit camps, the ability to interview Rohingya refugees, and the commitment from the Bangladeshi government that it would refrain from coercive tactics aimed to oust refugees (Abrar, 1995; Human Rights Watch, 2000). While their status as stateless alone did not motivate a lack of UN involvement, it illustrates how statelessness can convolute refugee protections. Six months would pass before the UNHCR would reach a similar agreement with Myanmar by which point approximately 50,000 Rohingya had been involuntarily repatriated (Human Rights Watch, 1996). Threats, beatings, and denial of ration cards were all tactics employed to coerce the Rohingya to comply with repatriation (Human Rights Watch, 1996, p. 18). A 1995 Médecins Sans Frontières survey found that 63% of the Rohingya did not want to return to Myanmar and 65% were unaware of the right to refuse repatriation (Médecins Sans Frontières, 2002). Even with UN