The Truth About Tamil Statistics

The truth about Tamil statistics

Kathy English
Toronto Star
Apr 03, 2009 11:59 PM

When a human chain of Tamil Canadian protesters circled Toronto's city core last month, the Star reported the “fact” that Greater Toronto is home to 200,000 Tamils, the largest Sri Lankan Tamil community outside Sri Lanka.

Is that number in fact correct? Short of lining up and counting every Tamil in Toronto, the Star can't determine exactly how many have come from Sri Lanka to make the GTA home. The best we can do is provide readers with a fair approximation from authoritative sources.

In a city that's now home to a multitude of ethnic populations, who keeps count of the diverse numbers in the GTA? In reporting complex population statistics, who should journalists look to for the best sources of accurate data?

Those questions are at the heart of an ongoing newsroom debate sparked by the Star's library staff. They contend that the oft-reported statement about the GTA's 200,000-plus Tamil Canadian community is erroneous.

At issue here is a significant discrepancy between Statistics Canada's census data which points to 29,435 ethnic Tamils in Greater Toronto and the far greater estimate of 200,000 cited by various academic experts consulted by the Star's immigration reporter, Lesley Ciarula Taylor.

The Star's research library regards StatsCan as the “official” source of data on population in Canada because the federal agency has a legal mandate to provide accurate data based on Canadian Census information. “Statistics Canada provides an objective and comprehensive view of ethnicity in Canada,” says Star librarian Astrid Lange. “It is the only source that provides us with statistics on everyone in the country.”

For the most part, Star journalists do cite StatsCan numbers when reporting population data. City editor Lynn McAuley believes (and I concur) that the Star should regard StatsCan as its authoritative source to provide “some baseline of accurate data that we can use consistently in our stories.”

Still, reporter Ciarula Taylor makes a credible case for why she doesn't regard StatsCan data as providing a complete picture of Toronto Tamils. She consulted with academic experts who've extensively studied the Sri Lankan Tamil diaspora. One cites a well-regarded study that delves into the question of the discrepancy between the low number of ethnic Tamils reported by StatsCan and the “actual population.” A University of Toronto professor who organized a Tamil studies conference told the reporter that 200,000 is more accurate.

“In this case, I think the academics have made a good case for why a community made up of war-scarred refugees with an intense loyalty to Canada might not identify themselves as just Tamils,” she said, also citing the Canadian government's 2006 designation of the Liberation Tigers of Tamil Eelam (the Tamil Tigers) as a terrorist group as further reason why many Tamils might not identify themselves to census officials.

Now consider the StatsCan numbers provided by the Star library's “statistics pro” Rick Sznajder, who worked with 2006 census data of those who self-identified as being of Tamil ethnic origin to determine a count of 29,435 Tamils in the GTA. As well, he cites Immigration Canada stats indicating that 114,896 is the total immigration from Sri Lanka from 1983 to 2007. All of this leads to the library's conclusion that “it seems highly unlikely that the number of Tamils in Toronto is anywhere near the reported 200,000 figure.”

Further complicating this is the fact that Statistics Canada reports that 98,265 people in the Toronto CMA claim Tamil as their mother tongue. More evidence that stats are often bewilderingly complex is the reality that not all who speak Tamil come from Sri Lanka.

Based on this analysis, is a correction called for as the Star's library suggests? Should StatsCan be the final arbiter of accuracy here?

I think that publishing a correction would be appropriate if it could be determined that the StatsCan number is wholly accurate and the figure cited in the reporter's articles are absolutely incorrect. That's not the case here.

But readers were not given the most accurate or complete account. Given the significant range of difference in data, the reporter should not have stated as undisputed fact that 200,000 Tamils reside in Greater Toronto. Readers would have been better served by being informed of Statistics Canada's “official” tally and the extensive gap between that figure and the number deemed accurate by academic experts. This is essential context here.

Overall, even keeping in mind the caution about “lies, damned lies and statistics,” the Star's library makes a good case for consistently citing of Statistic Canada's “official” data in the Star's reporting on populations. But for the widest, most accurate picture of any population, other relevant data should not be ignored.


Comments :

I'm surprised The Star did not get the SL Consolate's opposing view points for this article… The majority of the crowd was from Toronto, but many more 50k were from the GTA “905” area, therefore that may be why…

Submitted by ramesh85 at 10:49 PM Saturday, April 04 2009

In such a case…

… I'd rather see something along the lines of “an estimated 30,00 to 200,000 Tamils” and then give some idea as to the sources of these widely disparate numbers. That would be the most correct way to do this, although not necessarily the most accurate way. By doing this, one of things that is implied is that there are different ways of counting. (Yes, I took some stats courses in university.) If you don't want that in the story itself, add those details in a sidebar.

Submitted by rpearlston at 1:58 PM Saturday, April 04 2009