The Desire to Return. UNHCR explained that repatriation was justified because the Rohingyas wanted to go home.46 A central factor driving this conclusion was the previously discussed camp surveys. To explain repatriation’s newly found popularity among the refugees, officials referred to a delayed bandwagon effect, claiming that the refugees were now more knowledgeable about the improved situation in Arakan, and asserting that they had finally soured on life in the camps. Yet the surveys provided, at best, contradictory conclusions regarding the preferences of the refugees. Not only did UNHCR refuse to do a follow-up survey, but it ceased to aggressively seek out the refugees or inform them of their right to stay, and reduced the number of one-on-one, confidential, interviews.
The refusal to entertain potentially disconfirming evidence is exemplified by the following story. UNHCR’s report of a dramatic increase in the percentage of refugees who wanted to return to Myanmar sounded odd to several local NGOs. Fearful that this repatriation was involuntary and that the refugees had not been properly informed of their rights and the conditions in Burma, these NGOs suggested to UNHCR that it replicate its survey. UNHCR refused. After intensive discussion over how to proceed, in February 1995 MSF/Holland conducted its own survey in the camps. The results raised serious questions regarding the claim that the refugees were voluntarily repatriating. The vast majority said that UNHCR had not informed them that they had the right to say no to repatriation, a key feature of any definition of voluntariness. And only 9% of the refugees who said that they wanted to return indicated their decision was based on their expectation that the situation in Burma was safe.47 The summary conclusion was that the refugees wanted to return home but only after the situation had improved, and while they did not think that the situation had improved they agreed to return home now because they did not believe that they could stay in the camps. MSF/H released its findings and joined with other NGOs in writing to UNHCR to protest its activities. UNHCR dismissed the validity of the survey and depicted the protests as unwarranted and unprofessional, though it did say that it would renew its efforts to ensure that refugees understood that they could refuse to be repatriated (confidential NGO memo; interviews with personnel from NGOs).48
As UNHCR officials concede, its determination to promote repatriation was based not only on the refugees’ preference but more fundamentally on UNHCR’s assessment of whether life was better in Burma relative to life in the camps. As previously noted, they used a definition of voluntariness privileged their “objective” assessment over the refugees’ desire to return or assessment of the situation at home. This represented an important departure from the factors that typically are used to determine whether repatriation is warranted, and whose voice is supposed to count. Moreover, the refugees could not remain permanently in exile and therefore would have to return at some time in the near future, and they were better off returning while UNHCR was present in Burma. This point is subtly but importantly made in UNHCR’s 1995 State of the World’s Refugees, where it referred to the Rohingyas when discussing these less ideal conditions: “While the situation in Rakhine State may not be an easy one, the refugees appear to have recognized that it is better to go home now and to benefit from UNHCR’s presence and program, rather than to remain in the refugee camps which offer them no future.”49
The shift away from absolute standards regarding the desire by refugees to repatriate given their assessment of the situation in the home country toward a comparative evaluation by agency officials regarding whether refugees would be more secure at home or in the camps had the direct implication of privileging the agency’s knowledge claims over those offered by refugees. Refugees, in this view, would have a difficult time objectively assessing the situation, that is, taking into account, first, the short and long-term prospects of the situation at home relative to the situation in the camps, and, second, for how long UNHCR officials would be present and able to maintain some necessary safety features. This case was not unique. UNHCR’s employment of “objective” criteria to determine whether it was safe to return and not the refugees’ “subjective” perception of the situation has been observed as an emergent feature of other operations (Chimni, 1999). UNHCR might well be correct that refugees should go home now because life only will get more difficult if they remained in the camps. But the issue at hand is whose voice counts and what calculations are used to determine the efficacy of repatriation, and the measures deployed by UNHCR officials.
To support their claim that the refugees wanted to go home, UNHCR officials note that 200,000 refugees did return absent physical duress. That is, no one put a gun to their heads and forced them over the border.50 Their behavior, in this view, reveals their preferences. But even some UNHCR protection officers doubt the claim that refugees voluntarily express a free choice to return. Instead, they suggest that the refugees decision was probably influenced by UNHCR’s manipulation of information and bribery. UNHCR officials were claiming that life had improved in Burma and that UNHCR now had a presence in Burma and the ability to protect the refugees. But, as we will see below, these were exaggerated claims at best, and if these claims did influence the refugees’ decision then their decision was based on misleading information. Moreover, the refugees were offered various kinds of incentives to return, including the standard assistance packages and a highly uncharacteristic offer of $20. When asked to speculate why so many refugees who once refused to go back to Burma now “suddenly” changed their minds, one protection officer said that it was because UNHCR “oversold” its presence and the human rights situation in Burma and offered each family $20 to return, which is more money than they had ever seen in their lives.51
Burma Is Better Than It Was. UNHCR claimed that repatriation could proceed because Burma was “safe” and had improved from the conditions that triggered the refugee flight. This observation ran counter to the claims of NGOs and even other UN agencies. NGOs continued to write scathing reports of the government’s treatment of its population and the Rohingyan people, and the UN’s rapporteur on human rights had been issuing damning statements concerning the military regime’s repressive character and human rights violations. In fact, at the outset of the repatriation process UNHCR failed to certify that conditions had improved for the Rohingyas (Abrar, 1995: 19), a startling omission that it later corrected by saying that the situation was likely to get better because of UNHCR’s presence in Burma.
UNHCR’s determination that the Rohingyas could and should now return was based on the following factors. First, life was improving for the Rohingyas in Burma. Most importantly, even though Myanmar did not recognize the Rohingyas as citizens, it was restoring some rights to them and UNHCR hoped that this unfortunate situation might be resolved in the near future (Abrar, 1995: 17-18). Still, UNHCR did not consider this a human rights issue or a legitimate grounds for fear of persecution (Petrasek, 1999: 8). Second, UNHCR claimed that the Burmese government had fully complied with the MoU and responded positively to all of the agency’s queries. For instance, UNHCR asked the government to respond to the allegations that it was engaging in forced sterilization, forced labor, and forced education for Muslim women in military-run schools, and accepted the denial by the government and thereafter proclaimed the information as “mere rumors” (HRW, 1996: 15, 19-21). Third, it could provide protection through its presence; that is, rather than bringing people to safety UNHCR hoped to bring safety to people (Landgren, 1998: 427).
Finally, UNHCR officials claimed that while life was hard for the Rohingyas, it was no harder for them than for any other Burmese citizen or population group (HRW, 1996: 15; interviews with UNHCR officials, Geneva, January 2000). Simply put, because the Rohingyas were treated as horridly as other Burmese, they could not claim to be a persecuted people according to the 1951 convention definition (HRW, 1996: 15; USCR, 1995: 12-13). UNHCR had a point. It is a matter of debate whether these people were, in fact, a persecuted minority under the 1951 definition. But it was very odd point for UNHCR to be making. After all, UNHCR had long argued against a narrow interpretation of “persecution” and in favor of a more flexible definition. Now it found itself arguing the case in favor of the narrow definition it had historically opposed (MSF/H 1997: 11).
The claim that life had improved in the Burma was challenged not only by NGOs but more importantly by Rohingyas who were continuing to flee to Burma, often times including those who had recently repatriated. How to explain the migration? UNHCR officials elevated economic circumstances and traditional migratory patterns, noting that historically this area has seen economically-induced migratory patterns, particularly prominent during the dry season, and many Rohingyas were attracted to the economic safeguards provided by the refugee camps (Petrasek, 1999; confidential NGO memo; interviews with UNHCR officials). Although reluctant to classify these asylum seekers as “economic migrants,” they argued that because they left due to economic circumstances they should not be classified as “refugees” and should be denied access to the camps and other rights granted to refugees (USCR, 1996: 3, 7; Amnesty International, March 5, 1999; Petrasek, 1999: 1; MSF/H 1997: 5-6); UNHCR officials were especially fearful that giving refugee status might only encourage more flight. . While NGOs officials and local populations concede that economic factors played a role in their decision to flee, they explicitly linked their deteriorating economic conditions to their standing as a persecuted minority and to human rights violations (USCR, 1996: 7; MSF/H, 1997: 10; HRW/RI, 1997: 9). UNHCR officials apparently were not persuaded.
Once UNHCR is committed to a repatriation program and the claim that life is preferable in Burma, then disconfirming evidence is to be shunned. Again, UNHCR appears to have gone out of its way to avoid asking the tough questions that might have undermined its desired policy outcome. Rather than conduct systematic or individual interviews with those who approached field officers, UNHCR made blanket designations. And when UNHCR did interview new arrivees, it claimed that the purpose was to gather information and not to determine status (Petrasek, 1999: 10; MSF/Holland, 1997: 8). “Moreover, the information UNHCR provides on `protection issues’ in Burma does not include any reference to the fact that hundreds, perhaps thousands, of Rohingyan would-be asylum-seekers have been stopped by Burmese border guards as they attempted to flee Burma - a fact that Bangladeshi border security officials on the other side of the border freely admit” (Petrasek (1999: 11). As MSF/Holland (1997: 5) succinctly puts it: “To acknowledge their legitimate rights as asylum-seekers would call into question the whole logic of the repatriation program.”
By late 1995 the rumors seemed so credible that the UNHCR office in Cox Bazaar temporarily suspended its repatriation promotion activities. Confronted by consistent evidence that the human rights situation had taken a turn for the worse in Arakan and that refugees did not want to return to a situation that they believed might place them in danger, the Cox Bazaar office unilaterally halted its repatriation activities. This decision brought it into conflict with the Geneva and the UNHCR-Rangoon office (HRW, 1996: 15, 19-21; confidential NGO memos). Geneva claimed that Cox Bazaar officials had been premature in their rush to judgement and should continue the exercise, while the Rangoon office contended that the change in the situation was temporary and a consequence of a new, less tolerant, military commander and that UNHCR still had access to the region (interview with UNHCR protection officer, Geneva, January 28, 2000). This incident exposed not only the differences of opinion within UNHCR pertaining to the conditions under which voluntary repatriation might occur, but also the determination of Geneva to proceed with the repatriation program.
A central issue is that UNHCR elevated its own first-hand accounts of the situation in Burma to justify and authenticate its assessment of the human rights situation in Burma.52 As already discussed, UNHCR had increasingly played a role in monitoring the human rights situation in those countries in which it had a repatriation exercise. The mere presence of UNHCR, in short, put it in a “very powerful position” to judge whether the information regarding abuses was accurate or not (HRW, 1996: 15). There were two disturbing features associated with this new role, however. The first concerned its ability to judge the situation given its limited knowledge. When in June 1994 it was asserting that the human rights situation had improved, UNHCR did not have unhindered or easy access to Arakan. It had few staff in the field, a difficult time monitoring all of the remote villages and towns, and when it rounded it was nearly always accompanied by SLORC. In addition, UNHCR offered its appraisal of the situation in Arakan without a more comprehensive assessment of the situation in Burma (Petrasek, 1999: 8; MSF/H, 1997: 21; confidential NGO memos). The result was that UNHCR potentially held a distorted picture of the human rights situation, a distortion that made the Burmese regime seem less brutal than it was because it based its evaluation on the limited observations of its field officers. Although its declarations that the Rohingyas could return to Burma in safety and with dignity were based on suspect information, UNHCR’s voice carried tremendous authority and weight.
Second, UNHCR might not be an objective evaluator but rather have a vested interest in “seeing” an improving Burma. Recognize that UNHCR was highly committed to repatriating the refugees as soon as possible, but that possibility was dependent on an improvement in Burma’s human rights situation. UNHCR, therefore, had a potential conflict of interest, leading some observers to suggest that UNHCR’s repatriation imperative colored its assessment of the human rights situation in Burma (Petrasek, 1999). Specifically, an agency that advocates repatriation under less favorable circumstances might not be the best judge of the human rights situation. Moreover, while UNHCR’s role permitted it to challenge Burma’s treatment of the Rohingyas, it arguably shied away from exploiting that position because of the fear that an assertive campaign might jeopardize its ability to stay in Burma and undermine its own claim that a safe and dignified return was possible. “In Burma, UNHCR has monitored through the lens of repatriation, which has a distinctly rosy hue, `seeing’ those factors that are positive to continuing the repatriation and playing down or ignoring (at least in public) those factors which call into doubt the repatriation exercise” (Petrasek, 1999: 12; also see MSF/H, 1997: 23.
Capacity to Monitor Return. The agency claimed that the situation improved, and most critical, that it could monitor the safe return and reintegration of the refugees. Indeed, for UNHCR officials its mere presence in Burma meant that while the situation might not have already improved it now had the leverage to make like better for the Rohingyas in the near future. A UNHCR cable concluded that it had a “meaningful presence enabling [it] to undertake a reasonable pro-active and re-active monitoring function of the well being of the returnees and the progress in the implementation of the movement and reintegration phases.” But UNHCR’s capacity to monitor that return is dubious. UNHCR faced significant manpower, geographical, and logistical constraints that made it nearly impossible to play this role effectively (USCR, 1995: 1-2, 14-15).53 At the outset of the repatriation exercise there was only one, very fresh, protection officer in the region. Soon there were a few others but hardly enough to cover the entire area. Moreover, Arakan is a highly inaccessible region that has few paved roads, making it difficult to get around. And whenever UNHCR officials did interview returnees they were almost always accompanied by SLORC officials and forced to work closely with the Burmese Immigration and Manpower Department (IMPD). Accordingly, UNHCR could hardly expect that the returnees would give a candid assessment of their life after return (Petrasek, 1999: 8; MSF/H, 1997: 21; HRW, 1996: 21-22; interview with UNHCR official, Geneva, January 2000).
UNHCR was caught in a bind. It was committed to a repatriation exercise that depended on monitoring the return of the refugees - a difficult task under the circumstances. But if UNHCR was to openly question its ability to monitor the refugees’ return, then the entire repatriation exercise might be jeopardized.54 “UNHCR was in effect forced to choose between either sticking to principles and pulling out of any involvement in the repatriation...or accepting that a UNHCR presence must be conditioned on a pragmatic approach, the logic of which was that some presence was better than none at all.” (Petrasek, 1999: 5).55
UNHCR’s repatriation culture generated a definition of voluntariness, and shaped its assessment of the human rights situation in Burma and its capacity to assess what happened to those who returned, that made repatriation desirable and proper under less demanding conditions and in ways that violated the principle of voluntary repatriation. While this development might be a bow to “reality,” it has the additional implication that UNHCR is more likely to violate refugee rights even when it is has choices and is facing tolerable pressure from states. UNHCR introduced a definition of voluntariness that was no longer as dependent on the subjective assessment of refugees concerning the conditions at home, glossed over the human rights situation in Arakan as it introduced new standards and measures concerning what constituted a “safe return” and the need to repatriate even though conditions were not “ideal,” and was overconfident in its ability to monitor the return and reintegration of the refugees. UNHCR officials were now using different standards to evaluate the correctness of a repatriation program. This did not sit well with many others in the agency, who were quite upset that repatriation under these conditions did violence to the sacrosanct notion of informed consent. As one high-ranking UNHCR official observed, “The definition of voluntariness has been stretched to the point that it violated refugee rights and informed consent....That is a statement of fact.”56
UNHCR has organized itself to facilitate repatriation. There is little doubt that states wanted this outcome. But there also is strong evidence that UNHCR was not simply cowered by states but also was persuaded by new ideas, knowledge, and understandings regarding how to best help and protect refugees. Determined to overcome its “exilic” bias, help an increasing number of refugees who wanted to go home and were “spontaneously repatriating,” and “protect” refugees by getting them home to less than ideal conditions, UNHCR began to develop new terminology, concepts, and standards that would make this possible. As evidenced by its healthy and thoroughgoing debate over how far it could venture toward repatriation without violating refugee rights, UNHCR was no mere plaything in the hands of states but rather had the capacity for reasoned reflection and exhibited some relative autonomy.
But soon there developed a repatriation culture that left refugees at greater risk. I am not arguing that this culture necessarily and always will run roughshod over refugee rights in all or most cases. Instead, a repatriation culture means that UNHCR is oriented around concepts, symbols, and discourse that elevates the desirability of repatriation, coats it in ethical luster, and makes it more likely that repatriation will occur under more permissive conditions. Evidence of this repatriation culture is the mere fact that UNHCR can claim that repatriation is in and of itself a “success,” and that repatriation is presumed to be the best way to protect refugees even though there have been few studies that have evaluated what has happened to those who have repatriated.57 In any event, because this culture orients UNHCR toward repatriation, creates more permissive conditions for any single exercise, and gives voluntary repatriation an ethical and proper quality, it is more likely that a repatriation exercise will violate longstanding refugee rights. This did not develop instantaneously but rather over time and as a consequence of minor deviations that became institutionalized and accepted.
What occurred in the UNHCR and the case of the Rohingyan refugees is reminiscent of what Diana Vaughan (1996) called the “normalization of deviance.” In her impressive study of the Challenger disaster, she chronicles how exceptions to rules (deviance) over time become routinized and normal parts of procedures. Organizations establish rules to provide a predictable response to environmental stimuli in ways that safeguard against decisions that might lead to accidents and faulty decisions. At times, however, organizations make small, calculated deviations from established rules because of new environmental or institutional developments, explicitly calculating that bending the rules in this instance does not create excessive risk of policy failure. Over time, these exceptions can become the rule - they become normal, not exceptions at all: they can become institutionalized to the point where deviance is "normalized." The result of this process is that what at time t1 might be weighed seriously and debated as a potentially unacceptable risk or dangerous procedure comes to be treated as normal at time tn. Indeed, because of staff turnover, those making decisions at a later point in time might be unaware that the now-routine behavior was ever viewed as risky or dangerous. Although the organization might make countless decisions that do not lead to disaster, the safety procedures that were once in place no longer are and therefore leave the organization (and those it assists) at greater risk.
This argument applies reasonably well to the case of UNHCR repatriation Before 1980 the UNHCR viewed repatriation as only one of three durable solutions to refugee crises, and discussions of repatriation emphasized that the principles of safety and voluntariness must be safeguarded at all costs. UNHCR, however, has incrementally and steadily lowered the barriers to repatriation over the years, leading to a situation where initial deviations from organizational norms accumulated over time and led to a normalization of deviance. The result was a lowering of the barriers to repatriation and an increase in its frequency. Although most repatriation exercises continue to safeguard refugee rights, the cushion once provided by UNHCR rules have now been flattened and left refugees more vulnerable than ever. As a consequence, the probability that any single repatriation exercise might violate refugee rights has increased, although in the main UNHCR maintains its traditional protection function.
This repatriation culture contains not only categories that shape how UNHCR officials “see” the world, it also contains a discourse that makes ethical and legitimate the occasional violation of refugee rights if sone for the “greater good.” The repatriation culture generates a view that repatriation can proceed even if it departs from traditional refugee rights. Indeed, voluntary repatriation is now an obstacle to “protection.” The moral benchmark is no longer whether the totality of rights available to refugees are defended and honored but rather whether one course of action is more likely to provide better protection to refugees - according to UNHCR’s assessment. Because camp life is almost always unstable and insecure and contains no hope for the future, repatriation is almost by definition a more desirable outcome assuming that the situation at home has marginally and steadily improved. While getting refugees home might be the best way to protect them, it also is the case that this interpretation of the protection mandate might very well undermine refugee rights.