Projections, not predictions

By Raffique Shah
September 14, 2015

Raffique ShahMy last two columns, one titled “Rowley rising” and the other “lament for a falling leader”, were seen by many of my readers as being almost prophetic in the wake of last Monday’s election results.

Had I made public another document in which I analysed the results in all 41 constituencies from 2007 onwards, using historical data and trends, and projecting the 2015 results (which I circulated only to close friends), I might have been accused of being an “obeahman” or “seer”.

As someone who steers clear of the spiritual world and who certainly does not believe in “obeah”, I must say that my accurate projections (not predictions) were based on a scientific analysis of hard data, on my ability to read the mood of the electorate by monitoring public meetings and similar indicators, by assessing the two contending parties, their strategies, their leaders, and most of all by being objective.

Note well that I had long dismissed Jack Warner and the ILP, writing as much repeatedly, saying that if Jack polled 1,000 votes anywhere, I’d be surprised.

Essentially, I argued that the election will be won on the results from eight constituencies, ranked by degree of marginality as follows: San Fernando West, Tunapuna, St Joseph, Barataria/San Juan, Mayaro, Pointe-a-Pierre, Moruga/Tableland and Chaguanas East.

La Horquetta/Talparo and Toco/Sangre Grande were never marginal. They were won by the PP in 2010 in an extraordinary general election that saw the coalition ride a wave of popularity, taking other PNM strongholds such as Arima, Lopinot/Bon Air, the two Tobago seats, and scoring very well in all three Diego Martin seats as well as in other similar constituencies.

Following its manifest decline in popularity that showed up in the 2013 local government elections, which pushed the PP back into its base (and even there it was under threat), the UNC, with the COP being a spent force, went into this general election with 14 safe seats, having to fight to win all eight marginal and sub-marginal seats if it were to win the election.

The PNM, on the other hand, having wrested all of the East-West corridor corporations as well as San Fernando from the PP in 2013, now had 19 safe seats, and needed only two to take it over the top.

I saw those two as San Fernando West and Tunapuna (which it won comfortably). Anything beyond that was gravy-which it got in St Joseph (by over 1,500 votes) and Moruga/Tableland (a close 533).

By similar token, The UNC won Barataria/San Juan by 540 votes, and Pointe-a-Pierre and Chaguanas East by approximately 1,500 votes.

One constituency in which I was dead wrong is Mayaro: I thought Clarence Rambharat had done enough, and with the UNC finally naming its candidate on Nomination Day, the PNM stood a good chance of winning. Clearly, the reshuffling of polling divisions before the election benefitted the UNC, which won the seat by 2,900 votes.

Now that the dust has settled—we can forget the UNC’s spurious challenge: if they do get the courts to rule in their favour and the country has to return to the polls, people’s anger will wipe them off the map—there are some questions to be answered.

Among the recognised pollsters, only Louis Bertrand called it right. The well respected Nigel Henry cannot claim accuracy, having categorised La Horquetta/Talparo and Toco/Grande as marginal on the eve of the election.

Also, to the end, he stuck with 30 percent of the electorate being undecided, and worse, with hours to go, he asked people who they would vote for if the election were called then. I think Nigel needs to study Trinidad’s voters, and Tobago’s (there are differences), more closely.

Others like Vishnu Bisram (NACTA) and certain analysts and columnists lost all objectivity, calling the election results the way they wanted them, not the way they were seeing them.

If they were bad, UNC apologists were worse. One Kama Maharaj wrote of the “whooping (sic) 147,000-vote licks” the PP put on the PNM in 2010, decreeing there was no way the latter could overcome that.

Political scientist Dr Hamid Ghany saw the PP leading in St Joseph, Tunapuna, Point Fortin (??) and Talparo, with only San Fernando West and Toco too close to call. Is this a case of the “nutty professor”?

And on Election Day, UNC bloggers wrote of “exit polls” showing the coalition having a “commanding lead”. Really? Who conducted this US-style exercise, circus clown Rodney Charles?

Maybe it was he who had Kamla Persad-Bissessar say that her party was leading at six o’clock but losing at seven!

Off with his head, Kamla: he had you looking delusional in defeat.

3 Responses to “Projections, not predictions”


  • More than three months ago I challenged the figures that Nigel Henry was publishing. There is no way that there can be a thirty per cent undecided in any election. Personally, I do believe that Nigel knows this. He is not dumb nor is he unaware. What I believe is that he committed himself (or someone) to the idea that the election is close and it is too close to call, whilst using the thirty percent figure for cover.
    Is it possible for thirty percent to say that they have no preference? The answer is yes but an experienced pollster does not stop there. When people air uncertainties in polling, there are different paradigms that are used to exact their preferences.
    They may employ hypotheticals as one method and similes as another. These methods do not put the person being polled on the spot but acts in a way that would equate to what the pollster really wants to get from them. Vishnu Bisram also knew differently but he too probably made promises and had to stick with his promises. Hamid Ghany? Well he is part of the UNC pact and no one expected his polling to be different. Same for Sagewan-Ali, there is no doubt that she too favored the UNC so we expected her call for the UNC. Bertrand was definitely professional. His method of polling was on spot. He used the cross-overs from 2010 as a guide and determined how these constituencies were going to vote and he was spot on. Nigel has not done himself a favor professionally. I do not believe for one moment that he was fooled by the result. But he did his lot to present the predictions as presented.

  • Vishnu Bisram is an unapologetic ethnicist, a member of GOIP, and infamous for his fake polls and fake association. I have never found NACTA on line.

  • Predictions, polling and undecided voter are now things of the past for the 2015 General Election. The present reality is the PNM won the election fair and square, and we must now engage in the process of nation building. Remember Trinidad and Tobago is a Democracy and no party should think that political leadership is a birthright like a tyrant government – but depends on the will of the people!

Comments are currently closed.