Many analysts have written claiming that the votes cast in the recent presidential election testify to the ethnic divide. Others claim that the vote is an affirmation of the obduracy of the Tamils and Sinhalese who do not want to cooperate with each other. Even Douglas Devananda has claimed that he should step down because the Tamils did not vote en masse for Rajapaksa. This impression is largely because people look at simple-minded colour maps which show a large green area in the North and East, without looking at the very sparse number of voters, and how these votes got distributed. A more accurate voting map with the colours calibrated to the number of votes may be seen at http://dh-web.org/place.names#maps. Even that is not enough, and the numbers have to be analysed within a more sophisticated model.
An analysis of the polls, using post-Mavil oya provincial elections data tells a completely different story. However, to keep the story simple, let us look at just the presidential election results. The total number of votes cast in Jaffna, Vanni and Trincomalee for Sarath Fonseka add up to 272,000, while Mahinda Rajapaksa got 143,000. That is, Rajapaksa has garnered nearly 35% of the Jaffna-Vanni-Trinco vote. Trincomalee, or ancient Gokanna includes the Seruwila area. It is a more mixed population and its 44% vote for MR is not indicative of a corresponding Tamil vote.
In the Madakalapua (Batticaloa) district, we have once again 56,000 for MR, against 156,000 for SF, giving MR a 26% vote. The votes in the central hills and the Colombo district contain multi-ethnic components. They can be unraveled using a vector diagonalization scheme that we cannot present in this note. However, we find that the Tamil support for MR in Colombo and in the central Hills can also be shown to be stronger than in the Jaffna-Vanni area, hovering around 35-45% of the Tamil population.
We must not forget that there has been no democracy practiced in the North and the East for decades, and that there were only 325 votes cast in the 2005 election in Jaffna. This time we have more than a 100-fold increase in the vote. Thus democracy has returned to these areas. However, if the people were truly enthusiastic about the Tamil National Alliance (TNA), or even of Mr. Shivajilingam, then they would have come out massively. Instead they voted, more or less as their handlers told them, and without strong enthusiasm.
But they were willing to give 25-35% votes to Rajapaksa who was presented by the TNA, and the vast majority of Tamil newspapers as a person who should be rejected.Thus, in spite of all this anti-Rajapaksa propaganda, the Tamils of the North and East have given the President a strong supportive hand. This would indeed be the most fruitful path available to Tamil Cultural Nationalism, as I argued in a pre-election article that was published in the Transcurrents (http://transcurrents.com/tc/2009/12/the_presidential_election_and.html) electronic journal.
Unlike in the West, in Sri lanka we do not have a floating vote. This is particularly so just after a very emotional war where the thinking of the electors is strongly rooted in their polarized positions. Thus the Sinhalese voted as they did in the recent provincial elections, and so did the Tamils, as has been pointed out in detail in a recent article by "Gam Vaaesiya" in the Sri Lanka Guardian (http://www.srilankaguardian.org/2010/01/presidential-election-and-what-it.html).
Pre-election commentators in the West, and in Colombo erroneously assumed that there was a very significant floating vote which could be swayed using the "Fonseka factor".
Similarly, Marxist analysts have as usual found themselves to be completely off the mark.