Reading this article about teaching superdiversity in Singapore got me thinking about the official CMIO model. The Chinese Malay Indian Others racial classification system has been in place for decades, used to classify people by their ethnicity. As rigid models like these are wont to do, they have fallen short in addressing with the grey areas of life, such as inter-ethnic marriages and the children of such unions.
It is easy to say that a person’s skin colour is simply an accident of birth, not a reflection of who or what he is. It is also politically correct to say that race is a social construct. The natural corollary is that Singapore should abandon the CMIO model altogether.
But there is a reason why it existed to begin with.
The government has long advocated the use of social engineering techniques to foster and cement the notion of a diverse society. By deliberately mixing the population to create an ethnic soup everywhere in Singapore, they hoped to prevent people forming the cross-racial and cross-cultural barriers that led to the race riots of the turbulent 1960s.
Singapore’s housing policy, for instance, is structured along a quota system. Through the Ethnic Integration Policy, the Housing Development Board aims to prevent the formation of ethnic enclaves by ensuring a balanced racial mix. This, the HDB hopes, would lead to racial integration and harmony.
In countries without a similar policy, heterogeneous populations have historically clustered along racial lines, to ill effect. Arab enclaves, officially called banlieus and usually called ghettos, surround Paris, and are usually poverty traps that breed crime and resentment against the state. America’s major cities have housing projects, which tend to concentrate low-income blacks and give rise to high levels of crime and gang violence.
In particular, during the 1980s, Korean immigrants to Los Angeles tended to keep to themselves, forming tightly knit communities that tended to treat outsiders poorly. The 1991 shooting of Latasha Harlins by a Korean shopkeeper, Soon Ja Du, incited racial tensions between the black and Korean communities. When the Rodney King riots took off in 1992, the mobs targeted predominantly Korean businesses. Making matters worse, the police abandoned the streets, leaving Korean-Americans to defend themselves. This clash of communities could, at least in part, be attributed to African-American and Korean-American communities strongly identifying themselves along ethnic lines and turning against outsiders.
Conversely, Singapore has never experienced racial violence on the scale of the race riots fifty years ago. It is hasty to attribute this solely to the EIP, but the EIP did prevent the geographical clustering that led to the hardening of racial identities in America and Europe.
And the EIP uses the CMIO model.
At the macroscopic level, Singaporean society is seeing two major demographic shifts. Firstly, Singapore’s population is greying, accompanied by a declining birth rate. This spurred the second shift: an increase in immigration, encouraged by the government.
In this paper by the Ministry of Social and Family Development, there were much more elderly Chinese people than people of other races. This article from Migration Policy notes that the majority of immigrants to Singapore come from Malaysia, Hong Kong, Macau, South Asia, Indonesia and other Asian countries.
Singapore’s elders, it should be noted, did not necessarily have access to an English education, so it might be necessary to communicate with them using dialects or their mother tongue. The same applies for Singapore’s immigrants.
Meanwhile, CMIO also rears its head in Singapore’s education. In Singapore, every student is expected to be proficient in English and a ‘mother tongue’. This mother tongue is arbitrarily assigned based on a child’s race, though some biracial students have the option of choosing other languages. Dialects are not taught in school and officially discouraged by authorities; the current generation have to learn dialects through private programs, mass media or from their elders.
This confluence of demographics, education, language and culture would have significant impact on the ground. An immigrant from Hong Kong would be more likely to favourably respond to a Chinese person speaking Cantonese than a Malay person speaking English. Eldercare in Singapore is set to be a growth industry; with the majority of elders being Chinese, there will be many employment opportunities for Chinese and dialect speakers. Based on Singapore’s CMIO-driven education policy, that means employers would tend to hire Chinese people to meet the needs of their Chinese clients.
This also has implications for Singapore’s emergency services. Unlike most countries, the Singapore Police Force and the Singapore Civil Defence Force draws on conscripts for its frontline staff. These professions require personal interaction, which in turn require proficient language and communication skills as well as cultural appreciation and sensitivity. Being able to speak to people in their preferred language, and phrasing one’s message to suit the intended audience, is a vital asset. Singapore’s CMIO model means that it is easier for manpower planners to analyze the demographics of a given community, the skills required to interact with said community, how many people who possess such skills are needed in each station, and where to recruit them from.
That said, CMIO is a blunt instrument and does not accommodate all of reality. For instance, a Chinese police officer can study Bahasa Melayu to better interact with Malay citizens, and similarly an Indian can pick up Hokkien privately. Yet there are some cultural nuances that pure language courses do not cover. The Malay identity is intrinsically intertwined with Islam; among conservative households, it could be rude for a policeman to address a female at home without the permission of the (male) head of the family. I have also personally observed that people tend to use their mother tongue when conversing with people of the same race, and English when talking to people of other races. This suggests that a Chinese immigrant from Macau may be more comfortable conversing with a Chinese officer in Chinese but not necessarily a Malay officer in the same langauge. However, a Chinese immigrant from Kuala Lumpur could be fluent in Malay — which means it may not be prudent for Malay police officers to discuss in Malay the individual’s shortcomings within earshot.
The natural rebuttal is that first responders should naturally be given language and cultural education courses. To which I will say that there are only so many tax dollars and training hours available, and officers would resent being forced to take what they believe are unnecessary courses. Further, such an approach may not necessarily be available for private companies that need to turn a profit, so they have to rely on candidates already possessing such skills before they join up. The response to the latter, of course, that the education system should teach these skills before youths join the workforce, to which I have the same response: there’s only so many hours in the day and so much money to go around. There are pros and cons to every approach: the questions are what society will swallow, and whether the anti-CMIO crowd is prepared to explore and articulate the costs of their preferred option.
CMIO is a blunt instrument, and fails to cover the entirety of an increasingly heterogeneous population. But public policy has to paint with a broad brush to be efficient, and CMIO facilitates efficient administration and provides a predictive model for planning and policymaking.
This is not to say that CMIO should still be maintained, though. As I have mentioned above, there could be ways to work around the CMIO model. However, this does mean that it is not simply enough to say that Singapore should just drop CMIO and teach notions of superdiversity. CMIO underpins many of Singapore’s policies, such as housing, immigration, education and defence. Simply dropping CMIO would leave planners, analysts and policymakers scrabbling to find alternative tools to achieve policy objectives. Pundits and policymakers who wish to move away from CMIO need to go a step further, and start thinking about how to replace CMIO with what, and discuss the pros and cons of CMIO vs such theoretical approaches.
The devil, as always, is in the details.