Tracking polls capture the mood of the electorate

Dear Editor,

I refer to the letter captioned ‘Have Guyanese learnt anything from the American experience?’ by Mr Anthony Vieira. His letter contains several inaccuracies.

At the outset, I think Mr Vieira misunderstands my analytical assessment of the American presidential elections.  It appears he did not read my final piece published on election day in which I said Obama would win a landslide in electoral votes and popular votes that would be closer than what some polls had found.

On my polls in Guyana, Mr Vieira seems to suggest that I should not have conducted polls close to an election because it may sway opinion. I disagree. Polling is a dynamic phenomenon. Findings don’t stay static. That is why pollsters conduct tracking polls. For the US election, Gallup did daily tracking polls for about two months before the elections. Other pollsters tracked support every few days.  A poll conducted in August would not accurately reflect people’s attitudes in November.

Mr Vieira correctly stated that a poll I conducted about eight weeks before the 2006 elections showed the PPP with 42% of the votes. He suggested that no more polls should have been conducted because subsequent polls, he felt, caused PPP supporters (ie Indians) to rush to vote PPP so the party could win the elections.  I am baffled to understand why the poll galvanized PPP supporters and did not similarly galvanize PNC supporters (ie Africans) to defeat the PPP.  I conducted a series of four polls for the 2006 elections within eight weeks. PPP support increased during each one, while AFC support declined and PNC support increased. The polls were published exclusively in SN except for the final one. I approached Mr de Caires to carry the poll but he said he felt uncomfortable carrying it so close to the elections.  I proceeded to release the poll to all of the media.  It showed the PPP winning 51% with a 6% margin of error.

If I had not done additional tracking polls from the first one and had not conducted the final poll, I would not have accurately captured the changing mood of the electorate.  I should note that in the US polls are published even on election day, and there were dozens of polls released the day before the election. Contrary to Mr Vieira’s view, I do not think the polls had anything to do with a PPP victory.

On the US election, Mr Vieira is wrong about my assessment (I did not conduct polls but relied on the polls of others) on the closeness of the race. My assessment on the closeness of the US election changed with new polling data, and by election day I said Obama was heading for a landslide in the electoral votes as published in the Chronicle. Does Mr Vieira really expect what I wrote about the polls in August or mid-October would remain my view in November? In that case, I would not be a serious political analyst but a cheerleader for a candidate.  There were people who cheered Obama from 2007 to 2008 regardless of what the polls found. That is not analysis. In my final piece I maintained that the election would not be a 14% difference in popular support as the polls found. I wrote that McCain would get a minimum 46% support and Obama about 52%, and that was the election result.

On the electoral votes count, I indicated polls showed Obama getting between 314 and 386. I noted that in my analysis Obama was ahead in 18 states with 227 EVs and McCain ahead in 23 states with 189 EVs. Ten states were toss ups with 122 EVs which would give him 346. I noted that it was impossible for McCain to win any toss-up states. But I did award North Carolina to McCain by a squeaker. Obama won it by a fraction of a per cent and won the presidency by a landslide.

I did say that I did not think Obama would win the toss ups by the huge amounts the polls found. As it turned out, Obama won PA, NM, and Ohio big.  The others were very close. I should note that I was a volunteer for the Obama camp but supporting him did not compromise my objectivity in analyzing the polls. I did lay out a path in which McCain could win some swing states but still would not get enough votes to winning the presidency. That should not be interpreted as my saying the elections would be close.

Finally, I would like to support Mr Vieira in hoping that voters not only in Guyana but in other polarized societies learn from the recent American election.

Yours faithfully,
Vishnu Bisram