Burma's census disregards the complex ethnic identities of its people. Could this breathe new life into sectarian conflict?
Next year, Burma will embark on its first census-taking process in more than three decades. It's an opportunity, but it's also a significant risk. On the one hand, the census could compel the state to finally recognize long-excluded people and foster a better collective understanding of the daily struggles that most Burmese face. But on the other, the census is set up to obscure Burma's incredible diversity by requiring that Burmese people choose just one ethnic identity, even if they identify with many ethnicities. This comes at a dangerous point in Burma's simmering ethnic conflict, especially since nationalists are now using conceptions of exclusive and timeless ethnicity to justify violence against populations suddenly deemed irrevocably "foreign."
Instead of fueling such demagoguery, politics around the census process should expose the inaccuracy of those narratives and highlight the wonderfully mixed-up nature of ethnicity in Burma. Otherwise, the census seems poised to be part of a new kind of Burmese state practice, one that simply goes from domination (direct and despotic) to a new kind of control (diffused and bureaucratic) that limits rather than enables Burma's people.
Burma has 135 "official national races" (in addition to the Chinese, Indians, Rohingya, etc., who have yet to be recognized as autochthonous despite their long-standing membership in Burma's society). Observers use this number to remark on Burma's incredible diversity -- but this categorization is often myopic. The current categories imply that every citizen fits snugly into one silo: only Shan, only Karen, only Burman. A closer look at Burma's ethnic make-up, however, shows a vast diversity not simply within the country, but within people themselves.
Over three months of field research in Yangon this summer, I asked dozens of Burmese about their lu-myo (race or ethnicity) and found that individuals often describe complex, mixed-ethnic genealogies. For example, a Burmese colleague explained that ethnic identity is highly dependent on context: "For people like me who live in cities and don't speak an ethnic minority language, don't have ethnic minority names, and who are Buddhists, I don't think it would be a problem to identify ourselves as ‘Bamar lu-myo' ['Burman'] at first. But as we talk more about ourselves we include more information about different ethnic roots we have.... I am Bamar, but I'm also Mon, Pa-O, and Chinese." As this suggests, in Burma, ethnicity is lived less as a pseudo-scientific racial category and more as a set of practices shaped by one's environment.
Because context matters, an individual's own lu-myo can also be "on the move," changing between generations or within individuals over their own lifetimes. For instance, a man told me about his father's shifting identity: he was born Rohingya Muslim, but after refraining from Islamic worship practices, marrying a Rakhine Buddhist, taking on Rakhine modes of dress, drinking habits, etc., he now is often considered Rakhine. There are countless examples of this phenomenon: a colleague identifies as Mon though a cousin of hers does not; another scholar found a Karen-identified brother and Kachin-identified sister.
And yet, against these mutating and elusive identifications, the recent conflict in Burma's western Arakan state between Rakhine Buddhists and Rohingya Muslims -- in which mobs of local Buddhists have left hundreds of Rohingya dead and 125,000 displaced -- relies on a concept of ethnicity that is more absolute. (In the photo above, Muslim residents of Rakhine state await aid after losing their homes in sectarian violence in early October.) I interviewed multiple Rakhine individuals who insisted that "all so-called Rohingya" were actually "Bengalis" (considered "outsiders") and should be expelled. At the same time, Wakkar Uddin, a prominent Rohingya activist, told an audience at Columbia University last year that the Rohingya were determined to expel illegal "Bengalis." Significant here is how Buddhist Rakhine reject any potential blurring of boundaries between themselves and Muslims (whether Rohingya or "Bengali"), while the Rohingya are doing the same with themselves and "Bengalis."
The Rakhine/Rohingya case shows that conflicts can ossify conceptions of ethnicity to the point where they are no longer fluid and flexible, particularly when ethnicity becomes in part a vehicle for accessing resources. International media coverage has focused on racist monks or shadowy military elites collaborating with Rakhine demagogues to foment unrest. However, interviews with Rakhine individuals suggest that the conflict is grounded in perceived struggles over resources, especially surrounding the recently completed Shwe Pipeline, which carries gas to China but has left Rakhine state the second-least developed in Burma. Moreover, Rakhine individuals told me they were afraid of "losing their land" to Rohingya, who are ostensibly able to win control of resources by utilizing the support of international Muslim communities.
Other Rakhine say that international development only benefits Rohingya and ignores Rakhine needs. One man asked, "Why do the NGOs always come to our land but provide nothing for us, only for the Rohingya?" As a Rakhine woman explained, in this context, "Rakhine" has come to mean something very particular: "If we had development, we might say we are just 'Myanmar' [citizens]. But we don't." Rapid and unequal development is making ethnicity a conduit for protecting access to resources, a phenomenon that appears to be spreading across the country.
Given that ethnicity is a fluid but potentially charged concept, the question becomes whether Burma's reform process will embrace the country's complexity, or choose to privilege mono-ethnicity. This is where the census comes in. Interviews with the United Nations Population Fund (UNFPA), the agency providing technical assistance to the census process, reveal that the census has been designed to ignore the existence of multiple identities. Respondents must choose only one of the official 135 ethnicities, or check the "Other" box and write in their ethnicity. If a person with multiple identities refuses to choose one, the census defaults to their father's ethnic identity.
This may have serious political consequences. If people who claim multiple identities choose to report only, for instance, a "Burman" identity, hyper-nationalist movements may argue that these data "prove" that Burma's ethnic issues were always overstated and demand that the government grant collective resources to the besieged majority. Alternatively, if people report only non-Burman identities, the same movement could use those data to construct an equally dangerous argument: "We Burmans, the rightful sons of Burma's soil, are being bred out by the ethnic minorities. We must fight back." Burma's current monk-led, anti-Muslim "969 movement" can be seen as an inchoate version of such politics.
Why, then, would the state choose to implement the census this way? Is this a government conspiracy, a project to foment extremism while displacing official recognition of diversity? It appears not: UNFPA's technical advisors say that it is simply logistically difficult -- for both the census enumerators and its respondents -- to record multiple ethnicities.
But this could have drastic consequences. Comparative historical evidence shows that state census projects can intervene in sociological reality, creating the very categories they count. Indeed, a closer inspection of Burma's current 135 official races show them to already be arbitrary and confused, asserting phantom ethnicities on one hand and eliminating existing identities on the other. As scholar Mufti Myint Thein shows, the government concocted the number 135 in 1982, when many Muslim ethnicities were removed from official recognition (link in Burmese). These acts of reduction provide the grounds for exclusion: as in, "you are group x, and group x is not part of us."
How will Burmese people respond to such a project? During the long years of military control, state messages were often disregarded or ignored by a wary or disinterested populace. But now, Burma's state elites are busy reforming health, education, legal, and tax sectors, and much more, promising a transition from military authoritarianism to an aspiring Weberian-bureaucracy. When institutional changes actually affect people's daily needs, they have reason to listen; when these changes hinge on ideas of ethnic belonging, ethnic conflict may follow. Since Burma's most recent constitution guarantees special political representation if a lu-myo achieves 0.1 percent of the population, ethnicity will be a powerful means for groups to fight for their interests -- but only for the ones that qualify. The census, then, will help determine which groups matter in Burma, and which don't.
So what can census makers do to fix this problem? The best option seems to be to change the current format to allow citizens to select multiple identities to accurately represent their experiences. Even then, this may not be enough to dampen the socially fragmentary effects of Burma's current scramble for development.
Indeed, whether the census is reformed or not, what ultimately matters is how this census information is turned into political narratives about legitimate political belonging. Contesting ethnic violence in Burma will require messages that stress that the military regime was abusive to Burmese people of all ethnic backgrounds -- but that people from these varied groups are still able to forge relationships based on mutual respect and benefit, and are all committed to participating in a future Burma.
In other words, the census can certainly make things worse, but it cannot make things better on its own. Political leaders and citizens must together craft a new concept of citizenship in Burma, one based on the shared politics of daily life there that embraces all of Burma's diverse people without eliding any of their particular identities.
Soe Than WIN/AFP/Getty Images