Chapter 7. Deaths, Destruction and Displacement: Establishing the June 2012 Record
Section 1. Counting the Dead and Missing: Official Figures and the Rohingya Record
Establishing how many people were killed during the June 2012 violence is one of the most difficult problems in reconstructing the Rohingya historical record. The surviving figures differ not merely because organisations reached different conclusions, but because they were counting different populations, over different periods, with radically different access to affected areas. Government statistics, Rohingya organisations, Rakhine organisations, humanitarian agencies and international human-rights investigators consequently produced numbers that cannot responsibly be collapsed into a single unquestioned total.
For Rohingya, this problem has particular historical importance. Community organisations argued from the beginning that official figures substantially understated Rohingya deaths. Families reported people killed, missing, arrested or removed without knowing what ultimately happened to them. Access to affected Rohingya areas was restricted, independent investigators were not present during many incidents, and bodies were not always available to relatives for identification. A count based only on deaths acknowledged by state authorities could therefore exclude precisely those cases that Rohingya witnesses considered among the most serious.
The Myanmar government reported on 30 June 2012 that the violence had resulted in 78 deaths and 87 injuries, while 4,822 houses had been destroyed and approximately 52,000 people displaced. These figures covered victims from the affected communities rather than representing a Rohingya death toll. They became widely reproduced in international reporting because they were among the few consolidated figures released by the authorities.
Yet the official total was contested almost immediately. Human Rights Watch, after conducting interviews in Arakan State and Bangladesh, concluded that the government's figure of 78 deaths was likely a gross underestimate. Its investigators emphasized that the actual number killed was unknown. This conclusion is important because it demonstrates that questioning the official figure was not limited to Rohingya political organisations. Independent investigators who gathered testimony after the violence also found substantial reason to believe that the official count did not capture the full number of deaths.
Rohingya organisations had already been reporting a much larger loss of life while the violence was underway. On 10 June, eight Rohingya organisations—including the Arakan Rohingya National Organisation, Burmese Rohingya Organisation UK and Burmese Rohingya Community in Norway—issued a joint statement reporting that at least 100 Rohingya were believed to have been killed, with many more injured. This was an early emergency estimate rather than a final count, produced when communications were extremely difficult. Its significance lies in the fact that Rohingya organisations were already reporting more Rohingya deaths alone than would later fit comfortably within the government's combined total for all communities.
As additional information reached Rohingya organisations, their estimates increased substantially. On 28 June, BROUK President Tun Khin told a meeting at the British Parliament that information collected by his organisation indicated that as many as 650 Rohingya had been killed, approximately 1,200 were missing and more than 80,000 had been displaced. These numbers represented the Rohingya community's attempt to construct a wider picture from reports arriving from affected areas rather than a count based on unrestricted forensic access.
Other Rohingya organisations and networks circulated still higher estimates. A joint Rohingya situation report covering the period from 8 to 22 June alleged that the number killed could exceed 10,000 in Akyab/Sittwe alone, with a similarly enormous toll alleged across Maungdaw, Rathedaung and Kyauktaw. Such estimates represent an important part of the contemporary Rohingya documentary record because they show the scale at which community organisations believed people were disappearing or being killed. At the same time, the available independent evidence does not permit these very high figures to be presented as verified casualty totals. Rohipedia should preserve them explicitly as Rohingya organisational estimates, recording who made the claim and when, rather than either suppressing them or presenting them as conclusively established numbers.
The category of the missing is central to understanding why Rohingya estimates could rise so dramatically. A missing person was not necessarily dead. Some people had fled without being able to contact relatives; others were detained; families were separated during displacement; and communications between affected areas were poor. Yet Rohingya witnesses also reported people being taken away during security operations, bodies being removed and families receiving no information about detained relatives. In those circumstances, the boundary between a casualty list and a missing-person list could remain unresolved for months or longer.
This uncertainty created a methodological problem that official statistics could not solve simply by publishing a lower number. If a man disappeared after being arrested, his family might record him as missing while the government did not record him as a death. If bodies were removed before relatives could identify them, they might never enter a community's confirmed list. Conversely, community estimates produced amid panic could include people later found alive elsewhere. Both possibilities must be recognized when reconstructing the record.
The government and Rakhine community also documented Rakhine deaths. Rakhine civilians were killed during the June violence, including people attacked during the early confrontations and subsequent communal attacks. Human Rights Watch confirmed that serious violence was committed by members of both communities. Those deaths belong in the historical record and should not be excluded in an attempt to demonstrate the scale of Rohingya suffering. The central dispute concerns whether the official combined total adequately captured the much larger number of deaths and disappearances alleged by Rohingya communities.
International agencies faced similar limitations. The United Nations and humanitarian organisations could count people arriving at displacement sites and assess humanitarian needs, but they did not possess unrestricted access to every location in which killings were alleged to have occurred. Humanitarian population figures therefore provide valuable evidence concerning displacement but cannot automatically be converted into a complete casualty count.
The problem is particularly acute when comparing figures produced at different dates. An estimate issued on 10 June could not include people killed or discovered missing later in the month. A government total issued on 30 June might use different criteria from a Rohingya organisation counting missing relatives as probable victims. A human-rights report published months later could incorporate additional testimony but still lack access to official detention records, burial sites or every affected village. Apparent numerical contradictions can therefore reflect differences in methodology as well as genuine disagreement over the scale of the violence.
For Rohipedia, the responsible approach is not to select the lowest figure because it was issued by the government, nor automatically to select the highest figure because it originated from Rohingya organisations. The historical task is to preserve the Rohingya record in full while distinguishing estimates from independently corroborated counts. Rohingya organisations were documenting a catastrophe as information reached them from communities under severe restrictions. Their figures are historical evidence in their own right, even where the exact totals cannot now be independently established.
Several conclusions can nevertheless be drawn with confidence. First, the government's figure of 78 total deaths was disputed from the beginning and was subsequently regarded by Human Rights Watch as likely a gross underestimate. Second, Rohingya organisations consistently reported substantially greater Rohingya losses, ranging from more than one hundred deaths during the earliest days to hundreds and, in some reports, many thousands as additional information was claimed to have arrived. Third, the existence of large numbers of missing people makes any simple death count inherently incomplete. Finally, the restrictions on independent access mean that no surviving source can credibly claim to possess a complete forensic accounting of every person killed in June 2012.
The absence of one universally verifiable number should therefore not be mistaken for evidence that the losses were small. It is instead evidence of the conditions under which the violence occurred: restricted access, mass displacement, disrupted communications, contested state reporting and families unable to determine the fate of missing relatives.
For the Rohingya historical record, the names and testimonies behind these numbers are ultimately as important as the totals themselves. A future evidence archive capable of identifying victims village by village, family by family and from contemporaneous records may provide a more complete picture than any single estimate produced in June 2012. Until such documentation is complete, Rohipedia should preserve the range of reported figures openly, explain their origins and limitations, and resist converting uncertainty into an artificially precise casualty number.
The numerical record of deaths and disappearances is only one part of establishing the scale of the June violence. Physical destruction left another form of evidence, one that could, in some places, be observed independently from above. Burned villages, destroyed urban quarters, homes, shops, mosques and other religious buildings, together with satellite imagery, provide a second means of reconstructing what happened even where the number of people killed remains disputed.
Section 2. Measuring Destruction: Villages, Homes, Mosques and Satellite Evidence
If the number of people killed in June 2012 remains deeply contested, the destruction of the physical landscape provides another body of evidence. Burned houses, emptied neighbourhoods, destroyed shops and damaged religious buildings could be counted, photographed and, in some locations, independently examined through satellite imagery. This evidence does not resolve the disputed casualty figures, but it demonstrates that the violence produced destruction on a scale extending far beyond isolated clashes between individuals.
The Myanmar government's own figures acknowledged extensive physical devastation. By 30 June, the authorities reported that 4,822 houses had been destroyed during the violence. Government figures also recorded the destruction of 17 mosques, 15 monasteries and three schools. These totals covered destruction affecting both Muslim and Rakhine communities and did not establish who was responsible for every individual building destroyed. Nevertheless, even the official statistics demonstrated that thousands of civilian structures had disappeared within a matter of weeks.
For Rohingya organisations, however, the official property figures did not adequately communicate the scale or character of destruction experienced by Muslim communities. Contemporary Rohingya reports described entire clusters of houses being burned and repeatedly recorded attacks on mosques, religious schools, shops and Rohingya residential areas. Community organisations attempted to collect information from affected villages despite severe restrictions on communication and access. Their reports frequently contained locations and descriptions that were absent from the short statistical summaries issued by the authorities.
Religious buildings occupied a particularly important place in Rohingya documentation. A mosque was not simply another structure lost to fire. In Rohingya villages and urban neighbourhoods it functioned as a centre of religious and community life and represented the established presence of a Muslim population in that location. Reports that mosques were deliberately attacked therefore carried significance beyond the monetary value of the buildings themselves. Contemporary Rohingya sources described mosques being burned or damaged in affected areas of Maungdaw and Akyab/Sittwe, while government statistics themselves acknowledged that mosques were among the religious structures destroyed during the violence.
The government figure recording the destruction of Buddhist monasteries is equally important. Rakhine religious property was also attacked, demonstrating again that the physical destruction of June was not confined exclusively to one community. Rohingya documentation should not erase such losses. The central historical question is instead how destruction was distributed geographically, what happened to the populations whose neighbourhoods were destroyed, and whether residents were subsequently able to return and rebuild.
This distinction becomes especially important when counting individual houses. A total of several thousand destroyed homes does not reveal whether they were scattered evenly across many places or whether entire neighbourhoods were eliminated. Nor does it show whether a destroyed population subsequently returned. A house count can therefore underestimate the social significance of concentrated destruction. The disappearance of hundreds of adjoining homes can remove an entire community from a particular location even if the same number of structures destroyed individually elsewhere would not produce the same demographic result.
This is where satellite imagery became especially valuable. Human Rights Watch obtained and analysed high-resolution satellite images of areas affected by the 2012 violence. Satellite evidence provided an independent means of identifying large burn zones at a time when journalists and investigators faced restrictions on access. Structures visible before the violence could be compared with the same locations afterward, allowing researchers to document destruction that did not depend solely on government or community casualty reporting.
In Akyab/Sittwe, Human Rights Watch's satellite analysis identified approximately 2,558 destroyed structures across the destruction zones it examined following the June violence. The imagery showed extensive areas where dense concentrations of buildings visible before the violence had disappeared. The destruction of Narzi, examined in detail in Chapter 5, formed part of this physical record. Its importance here is quantitative rather than narrative: satellite imagery independently confirmed that destruction in the capital occurred across large contiguous areas and involved thousands of structures.
The figure of 2,558 destroyed structures in Akyab should not automatically be equated with 2,558 Rohingya houses. Satellite imagery identifies physical structures; it does not by itself determine the ethnicity of every owner, distinguish a house from every other type of building, or establish who burned each structure. Ground testimony and demographic information are necessary to interpret what the imagery shows. Nevertheless, when satellite evidence is combined with witness accounts identifying particular areas as Muslim or Rakhine neighbourhoods, a much more detailed reconstruction becomes possible.
Physical evidence elsewhere was more difficult to quantify comprehensively. Northern Arakan contained numerous villages spread across a large rural landscape, and not every settlement was covered by the same quality of satellite imagery or investigated independently. Rohingya organisations reported destruction in and around Maungdaw, including residential property and religious buildings, while Rakhine communities also documented burned homes and villages. The absence of satellite analysis for every reported village should therefore not be mistaken for evidence that no destruction occurred there.
A village-by-village record is particularly important for Rohipedia because aggregate statistics can conceal local history. A statement that “thousands of houses were burned” preserves the scale but loses the identity of the communities affected. Wherever reliable evidence permits, the historical archive should ultimately record the township, village tract or ward, village or neighbourhood name, type of structure destroyed, reported number of buildings, source, date of report and whether the information has independent corroboration. Such a database would allow future researchers to distinguish verified destruction from estimates while preserving Rohingya community evidence that might otherwise disappear.
Contemporary Rohingya documentation is especially valuable for this purpose because international organisations were unable to visit every affected settlement. Local reports sometimes preserve the names of neighbourhoods, mosques or villages that received little attention in international media. These sources should not automatically be treated as independently verified merely because they contain detailed information, but neither should they be excluded simply because international observers could not reach the location. The appropriate method is to retain the report, identify its source and seek corroboration through photographs, satellite imagery, additional witnesses or later investigations.
The destruction of shops and businesses also deserves attention. Government totals focusing on houses cannot fully measure the economic losses suffered by affected communities. Markets, commercial premises, stored goods and family businesses were destroyed alongside residential property. For Rohingya traders, particularly in Akyab, losing a shop could mean losing both accumulated capital and access to the commercial networks through which a family survived. Subsequent movement restrictions could then prevent displaced owners from rebuilding their businesses even where they retained some assets.
Documentation of religious destruction requires similar precision. A mosque reported burned by a Rohingya organisation should ideally be recorded by name and location rather than absorbed into a single statewide total. The same standard should apply to Buddhist monasteries and other religious buildings. Preserving both records allows the archive to document the full violence while also examining whether patterns differed between communities and locations.
Photographs constitute another category of evidence, but they require careful verification. Images of burning houses and displaced Rohingya circulated widely during and after the 2012 crisis, while photographs from unrelated conflicts or later episodes of violence have sometimes been incorrectly presented online as evidence from June 2012. Rohipedia should therefore attach a date, location, photographer or originating organisation wherever possible. An image whose origin cannot be established should not be used as definitive evidence for a particular village or incident.
Satellite evidence is less vulnerable to this particular problem because image acquisition dates and coordinates can be independently recorded. Yet satellite imagery also has limitations. It can demonstrate that buildings disappeared or were burned between two acquisition dates, but it cannot identify the person who lit a fire, determine how many people were inside a structure, or establish motive by itself. The strongest historical reconstruction therefore emerges when satellite evidence, contemporary Rohingya testimony, independent investigations, government records and photographic evidence converge on the same location.
The physical record also helps explain why displacement became so extensive. Tens of thousands of people were not displaced merely because they feared another confrontation. Many had no intact house to which they could immediately return. In areas where whole clusters of buildings were destroyed, the physical basis for return had disappeared. Where authorities subsequently restricted return or movement, temporary homelessness could become prolonged displacement.
This is especially significant from the Rohingya perspective. Destruction cannot be measured only by asking how many buildings burned. The more consequential question is what happened to the people whose buildings were destroyed afterward. A burned house that is rebuilt and reoccupied represents one kind of loss. A burned neighbourhood whose residents are prevented from returning and subsequently confined elsewhere represents a fundamentally different historical outcome.
The available evidence therefore establishes several points even where exact building totals remain subject to revision. Thousands of structures were destroyed during the June violence; Muslim and Buddhist religious buildings were among the losses; both Rohingya/Muslim and Rakhine civilians suffered property destruction; and satellite imagery independently confirms destruction on a massive scale in parts of Akyab/Sittwe. Rohingya community documentation further records destruction in locations that international investigators could not examine comprehensively.
The next evidentiary problem is consequently not whether mass displacement occurred—it unquestionably did—but how many people were displaced, from which communities, where they went, and whether displacement was temporary or became effectively permanent. Government figures, United Nations assessments and Rohingya organisations again produced different numbers, partly because the displaced population was changing rapidly as the violence continued.
Establishing that displacement record is essential because it connects the human and physical evidence. The dead and missing represent one measure of the violence; destroyed homes and religious buildings provide another. The tens of thousands of people forced from those destroyed communities provide the third.
Section 3. Displacement and the Problem of Establishing a Final June Record
Displacement provides the third major measure of the June 2012 violence. Unlike destroyed buildings, however, displaced populations were constantly moving. Families fled from one neighbourhood to another, gathered temporarily with relatives, entered makeshift shelters, moved toward formal displacement sites or attempted to cross into Bangladesh. A person counted as displaced on one date might be recorded at another location days later. Government agencies, humanitarian organisations and Rohingya groups were therefore attempting to count a population that had not yet stopped moving.
By 30 June 2012, the Myanmar government reported approximately 52,000 people displaced by the violence. Like its casualty and destruction statistics, this was a combined figure rather than a count of Rohingya displacement alone. It included people from both Muslim and Rakhine communities who had been forced from their homes. The figure established that displacement had already reached tens of thousands even according to the authorities, but it did not settle the question of the actual number uprooted.
Humanitarian estimates subsequently rose as agencies obtained greater access and additional displaced populations were identified. United Nations reporting during the following weeks described tens of thousands of displaced people across Arakan State, with both Rakhine and Muslim populations requiring emergency assistance. By July, humanitarian planning figures were approaching or exceeding 70,000 displaced people, and the number would rise further as later violence created additional displacement. These figures should not be confused with the eventual displacement totals for the entire 2012 crisis, which included the much larger second wave of violence in October.
Rohingya organisations were reporting higher Rohingya displacement figures during June itself. As noted earlier, BROUK President Tun Khin stated on 28 June that information reaching his organisation indicated more than 80,000 Rohingya had been displaced, alongside its estimates of hundreds killed and approximately 1,200 missing. Other Rohingya organisations similarly described very large numbers of people driven from their homes. As with their casualty figures, these estimates were assembled from community networks operating under conditions of restricted access and disrupted communication rather than through an unrestricted population census.
The discrepancy does not necessarily mean that one figure was simply fabricated while another was correct. Different organisations were defining and locating displacement differently. Government counts could concentrate on populations registered at recognised displacement locations. Rohingya organisations could include people sheltering informally with relatives, hiding in villages, temporarily displaced within their own village tracts or otherwise absent from government camps. Humanitarian agencies, meanwhile, tended to count populations they could identify for assistance. These methods could produce substantially different totals even when all were attempting to describe the same crisis.
Rohingya displacement also had a distinctive geographical pattern. In Akyab/Sittwe, destruction removed large numbers of Muslims from central urban neighbourhoods and pushed them toward Muslim-majority areas and displacement sites outside much of the city's ordinary urban life. In northern Arakan, displacement could take different forms: families moved between villages, hid away from roads and security operations, sought shelter among relatives or attempted to reach the Bangladesh border. These different forms of displacement could not all be captured by counting tents or registered camps.
The Bangladesh frontier illustrates this problem particularly clearly. Rohingya attempting to cross the Naf River were not necessarily included in Myanmar displacement records, while those intercepted or pushed back by Bangladesh could move repeatedly between the river, border areas and villages inside Arakan. Human Rights Watch and the Equal Rights Trust documented Bangladesh's refusal of entry to Rohingya fleeing the violence. The resulting population was displaced but simultaneously denied the possibility of becoming formally recognized refugees across the border.
For Rohingya families, displacement meant more than changing physical location. People lost access to homes, land, shops, documents, markets and community institutions. Children were separated from schools; workers and traders lost access to livelihoods; patients faced obstacles reaching medical care. These consequences would become increasingly severe as displacement continued. Their detailed development belongs to Chapter 12, but they are important here because they demonstrate why simply counting displaced persons cannot capture the full scale of the loss.
The composition of the displaced population also matters. Rakhine civilians were displaced in substantial numbers, particularly from locations where Rakhine houses and villages had been attacked or burned. Humanitarian assistance therefore served populations from both communities. The historical difference that subsequently emerged concerned the possibilities for return, movement and reintegration. Many Rohingya and other Muslim displaced people faced restrictions that increasingly prevented them from returning to their former urban neighbourhoods or freely moving through areas from which they had been removed.
This is one reason that the final June record cannot be reduced to a table containing deaths, burned houses and displaced persons. Those numbers measure immediate damage but do not show whether the consequences were reversible. The destruction of a house is one event; preventing its inhabitants from returning is another. Initial displacement can be temporary, but when restrictions prevent return it becomes a mechanism through which the demographic geography of an area is permanently altered.
The evidence reviewed across this chapter therefore produces ranges and competing records rather than one definitive set of numbers. The Myanmar government reported 78 deaths by the end of June; Rohingya organisations reported substantially higher Rohingya casualties and large numbers missing; Human Rights Watch considered the official death toll likely a gross underestimate. Government figures recorded thousands of destroyed houses and numerous religious buildings, while satellite imagery independently demonstrated extensive destruction in Akyab/Sittwe. Government displacement figures reached tens of thousands, while Rohingya organisations reported a still larger Rohingya displaced population.
These figures should not be artificially reconciled. A historical archive gains credibility by preserving disagreement where disagreement genuinely exists. Rohipedia should therefore record who produced each number, the date on which it was produced, the geographical area and population it purported to cover, and the method or evidentiary basis where known. A figure issued while violence was continuing should not be presented as though it were a retrospective forensic investigation, and an organisational estimate should not be stripped of its attribution.
The same principle applies to Rohingya evidence. Preserving the Rohingya perspective does not require converting every community estimate into an uncontested fact. On the contrary, preserving the original estimate together with its source and date protects the historical record from later erasure. If Rohingya organisations reported hundreds killed, more than a thousand missing and tens of thousands displaced, those claims belong in the archive even when independent investigators could not verify every individual case. The reader should be shown the evidence and its limitations rather than given an artificially simplified number.
At the same time, uncertainty must not be used to minimize what is established. Mass killing occurred. Thousands of structures were destroyed. Religious buildings belonging to both communities were damaged or burned. Tens of thousands of civilians were displaced. Rohingya communities reported losses far exceeding the government's acknowledged death toll, and independent human-rights investigators themselves concluded that the official casualty figure was likely substantially understated. Those conclusions do not depend on accepting the highest estimate.
A stronger future accounting would require evidence that was unavailable to investigators in 2012: complete detention registers, hospital and morgue records, burial information, village household lists, missing-person records, property registers, contemporaneous photographs and testimony systematically cross-referenced by location. Rohingya community archives may contain material capable of filling some of these gaps. Preserving those records is therefore not simply commemorative work; it can contribute to a more accurate historical reconstruction.
For Rohipedia, this creates an important long-term task. The June 2012 archive should eventually move beyond aggregate figures toward a victim- and location-based record: names of those reported killed or missing; village, ward and township; date and circumstances where known; destroyed houses and religious buildings; source of the information; and degree of corroboration. Such documentation would allow later researchers to revise totals without losing the individual people represented by them.
The June record therefore remains incomplete, but it is not unknowable. Government statistics provide one layer. Rohingya organisations preserve another, often much larger, account of loss. Rakhine records document their own casualties and destruction. United Nations and humanitarian assessments help establish displacement, while Human Rights Watch, Amnesty International and satellite imagery provide independent corroboration of important aspects of the violence. None of these sources alone contains the whole history.
Taken together, however, they establish that the June 2012 violence caused a human and physical catastrophe considerably more complex than the early official figures suggested. For Rohingya, the uncertainty surrounding the dead and missing remains part of that catastrophe itself: families were not merely deprived of relatives, but in many cases deprived of certainty about what had happened to them.
By the end of June, the immediate destruction was already being followed by another political development. Calls for the physical separation of Rohingya and Rakhine communities were gaining strength; restrictions imposed during the emergency were becoming harder to reverse; humanitarian access was contested; and displacement was beginning to harden into segregation.
Continue reading → Part 13, Chapter 8. Political Mobilization and the Institutionalization of Separation, June–September 2012.