The persuasive power of numbers (Part 2): “54 per cent of Tamils live outside the North and East”

My earlier article (The persuasive power of numbers and the mystery “8.5%” figure) highlighted the dangers of using incomplete census data and how it has been manipulated to bolster political agendas.

Another persuasive number that is routinely cited is the figure of “54%”, used in connection with the proportion of ‘Tamils’ living outside of the North and East. Here is a recent example:

“At least, the West has realized that 54% of ethnic Tamils are now living harmoniously in the Sinhalese-majority districts in other parts of the country, but Jehan Perera is in a fantasy world.” (Sri Lanka Tamil Tiger spokesman Tamilselvan’s death a set back for peace laments Sri Lanka’s peacenik, 3 November 2007)

There are other commentators who have used the number in such a way as to seek to justify that since “54%” of Tamils now live amongst the Sinhalese, there is no ‘ethnic conflict’:

When an untruth is repeated over a period of time, it usually appears to end up as an unassailable truth…It appears that the Boston Globe is unaware that today, the majority of Tamils in Sri Lanka, (about 54%), lives outside the north and the east of the country, among the Sinhalese, Muslims and other communities that blend into Sri Lanka’s multi-ethnic social fabric. This demographic transition shows that Sri Lanka’s conflict is not an “ethnic conflict” as the world at large conveniently categorizes it. (Global media’s ignorance of Lanka’s complex issues, Rajmohan Gomez, 19 November 2007)

But it is not just Asian Tribune writers that propagate the number. Bernard Goonetilleke, Sri Lanka’s ambassador to the US, in his attack on the Boston Globe article referred to above had this to say:

“[The Boston Globe] says Tamils “live mostly in the island nation’s north and east”. However, 54 percent of Sri Lanka’s Tamils now live outside the north and east, among Sinhalese, Muslims, and others. (Editorial misreads Sri Lankan government’s situation, Bernard Goonetilleke, 19 November 2007)

President Rajapakse has also referred to the “54%” figure in his various addresses to underscore the view that Tamils have “migrated” to the Western Province, where they reside “happily”. The following is an extract from a speech before the United Nations General Assembly:

“Today, the innocent Tamil people in Sri Lanka are migrating to the Western Province in large numbers. 54% of the total Tamil population is now living outside the North East. Especially in the Western Province they live and work happily. It is no secret to the world that the LTTE is a murderous outfit.” (Speech of Mahinda Rajapaksa, Sixty-First Session of the United Nations General Assembly, 20th September 2006)

It is worth noting that in some instances the reference is simply to “Tamils”, but in other instances to “Sri Lanka’s Tamils” (Goonetilleke) or to “54% of ethnic Tamils”, as in this example:

“The life line is given to a terrorist movement which has no acceptance over a majority of the 12 percent Tamils. Those who are providing the life line in their acts, pronouncements, policy decisions and behaviors do not want to know that 54% of ethnic Tamils in Sri Lanka now live with the majority Sinhalese in other provinces…” (A joint ‘life line’ to Sri Lanka’s Tamil Tigers, Daya Gamage, 23 September 2007)

Interestingly, in the above example, Gamage cites both the correct figure of 12% from the DCS 2001 census estimate (see earlier article) and the “54%” figure, which of course makes his point all the more persuasive.

So where does the “54%” derive?

All references to the statement that “54% of Tamils live outside of the North and East” do not provide any supporting evidence. The recent report of the International Crisis Group in its study on Sinhala nationalism also highlights the lack of evidence for the “54%” claim and how it has been used to argue that devolution of power and federal solutions are irrelevant:

“More recently, the president and others have taken to claiming that 54 per cent of Tamils now live outside the north and east. That figure – for which no evidence is given – is used to argue that devolution is not relevant, since many Tamils no longer have a connection to their supposed homeland and in any case prefer to live in the south, where a federal solution would not benefit them. That so many Tamils live outside the north and east is also offered as proof they are well treated by the government and the Sinhala majority. “There is no ethnic conflict”, Gammanpila said. “If you come to Colombo, you see every ethnic community living in harmony….Outside of the north and east, all communities are living in peace and harmony”. (Sri Lanka: Sinhala nationalism and the elusive southern consensus, 7 November 2007)

Having previously explored the material surrounding the “8.5%” figure, I had a hunch that the “54%” figure that is cited by so many from the President down to mere Asian Tribune writers, actually stems from the same error: the source was the same flawed and incomplete CIA data highlighted in my earlier article.

When a colleague adept at numbers looked at the figures for me, he very quickly arrived at the same conclusion. Those not interested in the maths, can skip to the next paragraph. Perversely, the “54%” figure seems to be arrived at by ignoring the Sri Lankan Tamils living outside of the North East (where the 2001 census was able to capture their numbers – 3.9% according to the flawed CIA Factsheet) and solely take the percentage of Up Country Tamils (4.6%) as a percentage of 8.5%, giving the “54%” figure. In other words, or rather put as an equation, this would be 4.6 / 8.5 x 100 = 54%. Could our thinking behind this rather dubious approach to the calculation be right? I invite alternative explanations as to how the “54%” figure might have been calculated.

Some reflections

The use of the “54%” figure raises several important issues. It is in the first instance another example of misleading use of incomplete census data and journalists need to be be cautious when relying on this figure and its implications. But what is more interesting is to look at what underlines the thinking behind the “54%” number by its proponents.

The rationale behind the use of the figure is clearly to include data relating to so called ‘Up Country’ or ‘Indian Origin Tamils’. But it is clear that when dealing with issues directly pertaining to the North East, such as a discussion on the concept of a Tamil ‘homeland’, issues of devolution or federalism or as to whether there is even the existence of an ‘ethnic conflict’, using such data seriously lacks credibility. The International Crisis Group rightly points out that “including ‘Indian Tamils’ weakens the [president’s] point, since they have never been involved in the struggle for a Tamil homeland or autonomous region in the North and East”.

These challenges to the “54%” figure dismiss many arguments put forward by those pushing certain political agendas. Of course the figure is very persuasive. But replacing the figure of “54%” with the ‘true’ or truer figure for the number of Sri Lankan Tamils actually living outside the North and East, would give us a very different way of looking at the picture. The footnote to the above extract from the ICG report neatly sets out a rough estimate of the number of Sri Lankan Tamils living outside of the North and East, using data compiled by the Northeast Provincial Council from District Secretariats’ data and arrive at a figure of 27 per cent, not 54 per cent. Replacing the figure “54%” with the figure of “27%” does not really have the same impact does it?

But regardless of the numbers or their accuracy, these statistics linked to demographic changes in the country are being used to sidestep important issues that are at the heart of the conflict. The fact that a group of people may have moved away from a particular region – a region that is experiencing severe hardships as a result of cycles of sustained conflict – is being offered as an indicator of preference. This raises a number of questions. Have these people voted with their feet? And what have they voted for by moving? Have people of different ethnic groups who have moved out of areas prone to conflict done the same? And what are the political implications of the movement of other ethnic groups from conflict areas?

It is clear from the above observations that politicians and certain media organisations are cleverly using the “54%” as a particular political manipulation in such a way as to undermine a different political agenda. But then, that is the persuasive power of numbers.

[Editors Note: For more information on the evolution of conflict and peace in Sri Lanka visit PACT. PACT seeks to help those with an interest in the Sri Lankan conflict gain a deeper understanding of the conflict's roots, manifestation and trajectory and to promote discussion around events, themes and experiences of peace and conflict related events.]

  • Sham

    Dear unionblack

    As can be seen the number 54% seems to have been arrived from the way that you have detailed.

    but like you said, by the fact that they actually have moved to western province shows that they have voted with their feet.

    It would be interesting to see if they will vote for the TNA or to another party at the next election.

    on another note, are you suggesting that the number of % who are living outside NE is higher than 54%? can it not be 54% by a mere chance?

    i dont seem to understand your argument, as you can’t use the above logic to prove that the acutual number living outside is differing from 54% , or as aindication in which direction (UP OR DOWN) it should move, and you have no clue as to the number?

    SO basically whats you say is 54% may be wrong as its a simple calculation but there is no other calcualtion, which gives a better picture? and actually you may be wrong and it may be acutally close to 54%.?

  • Veedhur

    Union Jack: Brilliant and much needed approach to the questions at hand.

  • Sinaha

    How did you arrive at this figure “54%” when census are not carried out in the North and East to know how many Tamils actually now live in Sri Lanka. If the author is concerned about others using his raw data, then first authenticate the basis for this figure.

    Other stats say that there are only 4% Tamils left in the country. Do we take are seriously too?

  • Raj Sreetharan

    This trend, we find elsewhere in chronic discrepancies in statistics across many sectors in the combat zone, in this case, in demography, how many Tamil in N-E.

    Also, in others: how many refugees, how many displaced including shades of displacement, how many actually killed by SLA, LTTE, paramilitary, depressed/inflated casualty tolls of cadres from all armed actors. This list is a running one.

    Sham, I think union black’s point is simply, 54% is not an empirically immutable figure, thus it should not be used – for example, by a President – as empirical evidence to disprove ethnic demographic majoritarianism of Tamils in the N-E, if that is the case or not.

    If 54 is correct, above, below, is less relevant. If these figures are used with a priori knowledge they may not be 100% correct, to prove a political point, it would be suggestive of mala fides, no? The same logic would apply to Tamil nationalist advocates on the murky issue of “demographic patterns” and state-sponsored resettlement schemes in Amparai.

    Which actually touches on a much larger problem of Tamil-Muslim historiography, past and modern, in Lanka – where much of the information coming out of the region is perceived biased before it is read, tainted before reported, on both sides of the ethnic divide, a simplistic example, from Tamilnet to AsianTribune. Both are perceived biased by the perspectives they do not push.

    In general, the modern historical narrative is versioned, aside from a few media columnists and writers doing the best they can with the access they have to report the truth. in the aftermath, until there is some sort of internationally-mediated TRC in a distant, distant post-settlement future where the plausibility of co-existence does not seem depressingly laughable – Lanka’s inclusive history – Sinhala, Tamil, Muslim – will remain ambiguous, disputed. History is always disputed, but in Lanka’s case, there is room to diminish the divide, and find areas of convergence.

  • Sham

    Dear Unionblack /Raj

    Please accept my appologies.
    what i was acutally saying is that in the absence of anynumber, some aproxiamte number would do, till it’s refined and cut and chopped till it come close to the actual number.

    but actually , interetsing fact is that CPA, has carried out a survey, and they seems to have the ethnic figures as per the people who conducted the survey.

    the article is titles “Poll results: How do you think we can end the war and attain peace in Sri Lanka?”

    and in return to a question i posed a staff member had replied that “However, the data is weighted before any analysis in order to make sure that the survey reflects the actual ethnic and geographical compositions of the surveyed population (in this case, the survey areas mentioned in the report). ”

    it may possible to get the exact figures and % from them. good luck

  • unionblack

    I tried to underscore in my article that one needs to look beyond the numbers and not “sidestep important issues that are at the heart of the conflict” by using these statistics, flawed or otherwise.

    But at the risk of weaking this important point and since a number of responses (Sham, Sinaha) have asked about the numbers, I respond as follows (most of which is found in the main article).

    As I said in my post there is no evidence given for the “54%” figure by its proponents (President Rajapakse, Asian Tribune etc). My hunch was that the very persuasive “54%” figure is based on a flawed interpretation of census data, together with a misguided approach to North East conflict issues.

    However (as I noted in the article) one estimate for the actual number of Sri Lankan Tamils living outside the North East is given in the recent ICG report on Sinhala nationalism. The ICG uses recent data compiled by the Northeast Provincial Council from District Secretariats’ data. Rather than paraphrase it, I reproduce below the footnote of the ICG report:

    “President Rajapaksa, addressing the World Affairs Council in Los Angeles, stated that “54 per cent of Sri Lanka’s Tamil population now lives in areas other than the north and the east of the country, among the Sinhalese and other communities”. Accurate population figures for Sri Lankan Tamils are difficult to come by, given the 2001 census could not be held in the Northern and Eastern Provinces and given ongoing war-related displacements. Calculations are further complicated by the fact that many Indian or Up-Country Tamils categorised themselves as “Sri Lankan Tamils” in the 2001 census. Nonetheless, one can come to a rough estimate by combining the 2001 census data for Sri Lankan Tamils living outside the north and east – 622,961 – with figures for Tamils living in the north and east – 1,693,751 – compiled in 2004 by the Northeast Provincial Council from district secretariats’ data. Based on these figures, 27 per cent of the total Sri Lankan Tamil population lives outside the Northern and Eastern Provinces, not 54 per cent.”
    (Sinhala Nationalism and the Elusive Southern Consensus)

    The “27%” figure calculated by the ICG as the number of Sri Lankan Tamils living outside the North and East would seem an accurate figure, but I reiterate the need to look at issues at the root of the conflict and reasons for the movement of Tamils, rather than give too much credence to these seemingly persuasive numbers.

  • ealem boy

    54%,, 27 %,, 8.5 %,, 4% doesn’t really matters now, what really matters is before the problem escalated when the Tamils in 1977 endorsed Tamil Elam by electing TULF with over whelming majority from Tamil areas that’s when it’s matters. Sure, like you people say, may be Tamils living outside north and north east may be more than people living in north and northeast, but do they have the freedom or choice to live in their in their own land or how many of them are living outside are forced to live there for fear of being killed or kidnapped by army or SLAF or the thugs who have been on the GOSL pay roll (suppose to be the moderate Tamils).

  • apec

    Dear unionblack I am sure your friend who is good with numbers could actually extrapolate a more accurate number of Sri Lankan Tamils living outside the North and East, using some good estimations. Here are some curious calculations drawn from the numbers readily available on the DCS website.

    Some data from the 2001 Census

    732,149 SL Tamils in the 18 districts outside of the N&E. This is 4.3% of the total population outside the North and East (which is 16,929,689).

    16,929,689 is the total population outside the N&E

    18,797,257 total population of Sri Lanka

    1,867,568 population of the N&E

    If we decided to try and calculate the 54% percentage differently and assumed, since we don’t have the statistics, that all the population of the N&E were SL Tamil, (which of course they are not) then in 2001 there would have been 1,867,568 + 732,149 SL Tamils in all of Sri Lanka, and the proportion out of them living outside of the North and East would be 28% . I am sure you could come to closer estimates using other figures, but not sure it would give you a figure of 54%.

    If we assume the 54% number and calculate the number of SL Tamil people in Sri Lanka, then you get a figure of 7.2% which is lower than your 8.4% and a huge drop from 12.7% in the 1981 census. Could calculations based on natural population increase over two decades, migration and death account for such a huge variance?

    Another point is the juxtaposition of dubious numbers with more credible facts that lead to completely distorted messages – for instance, President Rajapakse’s contention that “Today, the innocent Tamil people in Sri Lanka are migrating to the Western Province in large numbers…especially in the Western Province they live and work happily” is backed by migration statistics that show that the largest migration to Colombo is from Jaffna, though innocence, Tamilness, or ‘living happily’ are assumptions and leaps of faith!

  • Dush Kurera

    The main points to be taken on 27% or 54% is LTTE has been upto ethnic cleansing by removing most of the Muslims and the Sinhalease from the North & East of Sri Lanka. This, they have not stopped at that but the Tamils too has run from thier dictotorial rule of Gun Culture and Forced proscription of the young. The LTTE WILL NEVER ALOW TO TAKE A PROPER CENSUS … Please note.