Monday, January 11, 2010

Are Indian students being targetted in Australia?

The spate of high-profile violent attacks on Indian students studying in Australia has been the source of an intense outpouring of public indignation in India.

Popular reaction to the attacks, mainly channelized through the trigger-happy Indian electronic media, have accused Australia of racism and the government of not doing enough to seriously address it. The Australian government have, while acknowledging the problem, been at pains to point to the fact that its cities are much less crime prone than many Indian cities.

Are the attacks racially motivated and has it assumed epidemic proportions? The answer probably lies in examining records and using statistical techniques, specifically Bayesian probability, to tease out the conclusions. In other words, we need to show that among those living in the same areas, Indian students experience a statistically significant higher probability of facing a violent attack than a normal Australian.

Since Indians are the largest foreign student community and since many of them work outside to finance their studies, it is natural to expect them to be the target of racial attacks. More so since they live on the outskirts and suburbs and the recession has adversely affected employment opportunities in Australia. However, the attacks on Indian students can be classified as racist and disproportionately large only if

1. The share of attacks on Indians is statistically higher than those on Australians living in similar suburban areas, and
2. The share of attacks on Indian students is higher than those on other Indians

Though I could not get statistics about these from Google, reasonable approximations can be made with some simplifying assumptions. Wikipedia informs that there are 97000 Indian students in Australia, and Australian population in 2006 was 19,855,288. The Australian Bureau of Statistics reports that 255 people were murdered in Australia in 2007, or a rate of one murder for every 77,863 people. Assuming the same rate, one would expect just one (1.25 to be exact) murder crime among the 97000 Indian students in Australia!

Working backwards, the estimated more than 500 violent suspected/alleged racial attacks on Indian students in Australia in 2008 means that one in 200 students (or 500 victims per 100,000 people) are likely to face a violent attack. Extrapolating this for the entire Australian population, there should have been 1,02,347 incidents of violent criminal attacks in Australia!

Now take a look at offical Australian crime stats for 2008 - 1.2 murders, 1.1 attempted murders, 3.7 kidnapping and abduction victims, per 100,000 persons, 77.2 robbery victims, and 2 blackmail victims each per 100,000 persons. Surely, 500 victims of violent attacks per 100,000, among Indian students is a statistically significant aberration as to merit some explanation that goes beyond commonplace urban crimes.


Imhotep said...

Hi - interesting observation. One question though - where did you get the '500' attacks number from? If from media reports, then your analysis may not be correct; some of the media's reporting has been fanciful, to say the least. I am aware of at least 2 cases where allegations were made which were later found to be completely concocted.

A couple of other points - most Indian students live in and around urban areas, so the non-urban Australian population should be deducted from the overall no. in order to arrive at a more accurate urban crime comparison.

If race is an element (and I can imagine it definitely is), then it would be interesting to compare the experiences of Indians versus black people/aboriginal people or Middle Eastern people to see if Indians are being particularly targetted or whether it's just anyone 'non-white'.

Urbanomics said...

thanks for the comment

500 is the lowest of the numbers i could pick up from Australian newspapers (one of them even reported of 1200 odd reported attacks in the Victorian capital). the papers are SMH, Age and Australian.

though there are many simplifying assumptions in the post, the final results do not change by much. crime is more likely to be concentrated around urban centers and Australia is one of the most urbanized countries around (more than 90% live in cities). by accounting for the non-urban population would change little.

you have point there about the race bit, we will need more broad-based data to claim that it is conclusively racist.

however, my feeling on this is that race (and racism) rises to prominence when the target group starts becoming a salient and important competitor to the locals in the fight for scarce (and lesser than normal in the present recession-hit times) resources (which can be jobs, land, houses, college seats etc). as I mentioned in the post, Indian students fit nicely into this bill. add the difficult (not-so-distant) racial history of Australia and the pointers cannot be much else.

so, I would not be surprised if the stats do not reveal any dramatic spurt in attacks against students from other areas.