Ken Jones, Contributor

Ken Jones

Malaria aside, there is also the trace of election fever in the air, and although it is the season to be silly I would advise both friends and foes alike to neither exult nor be depressed by the inevitable diagnoses of the psephologists and their opinion polls. For my part, I will read their prophecies, but for entertainment only; and their findings will be consigned to the same file 13 that I use for horoscopes and the tips of tea-leaf readers, palmists and race track bookies. Of course all of these come out right some of the time, but a realist shouldn't plan his life or the fortunes of his country on this type of forecasting.

soothsayers jeered

Whether it be Anderson, Johnson, Stone or George Gallup, all soothsayers have come short of exactitude and have had their findings made objects of scorn and ridicule. Gallup, the daddy of this science of sampling was leader of the pack that predicted an easy win for Thomas Dewey who ran against Harry Truman for the U.S. presidency in 1948. He confidently predicted a Dewey victory margin of between 5 and 15 per cent and three weeks before the election declared that there was no point in further research. There was even an expressed concern that the polling might create a bandwagon effect that would make the margin even wider, but Truman won comfortably with a margin of 4.4 per cent.

Gallup's rise to prominence was itself the result of the major 1936 error made by his predecessor, The Literary Digest, which was at the time regarded as America's most influential polling organisation. On the result of ten million ballots circulated the Digest poll concluded that Franklin Roosevelt would be beaten by 14 points. Roosevelt won with a margin of more than 20 per cent; and Gallup who had earlier rejected the Digest's forecast was a winner with his correct prediction.

Mr. Bill Johnson's blunder in St. Lucia is only the latest of a string of incidents demonstrating the hazards of opinion polling. The claim that his published forecast was wrong because it was done too long before Election Day is revealing. If three weeks before the date is too soon for a reliable forecast how can anyone pay serious attention to the figures being trotted out a full year prior to the anticipated election!

To get closer to the real day of reckoning some practitioners have been using what is called an exit poll, which is based on answers taken from persons as they leave the polling booths on election day. But even this has been plagued by flaws and wrong guesses, to the point that the Washington Post's director of polling had cause to say, "The 2004 election may have finally stripped exit polling of its reputation as the crown jewel of political surveys ... this face-to-face, catch-the-voters-on-the-way-out poll has been revealed for what it is: just another poll, with all the problems and imperfections endemic to the craft."

These problems and imperfections are not readily admitted by any of our pollsters in Jamaica. Their defence lies only in the three per cent margin of error that is allowed for human and technical frailties of the business. Even with this safeguard pollsters must walk carefully through the thorny, hazardous pathways of prophecy. For instance, in the last two big elections in Jamaica pollsters stubbed their toes and dashed their feet against the stonewall of facts that emerged from the actual vote in 2002. In our electoral system the percentage of votes does not necessarily equate the percentage of seats, yet one pollster predicted that the PNP would win 40 seats. He was way off mark. The ruling party won 34 including three by margins below 300 votes. A year later they did even worse in forecasting the result of the local government elections. In that one, Bill Johnson had the PNP winning by near 10 per cent majority. In fact the party lost badly, finishing ahead in only one of 13 parish contests.

miscalculations

If those miscalculations are not enough to cast doubt on the accuracy of these opinion polls, then we can take the St. Lucian case into account; and we can marvel that despite the fears we are all experiencing, as well as the rising demand for more security guards and policemen, one highly rated pollster finds that 83 per cent of us feel safe. We may also wonder how in the middle of a major PNP internecine turmoil, while the JLP is enjoying unusual inner peace, the pollster finds that most Jamaicans believe that the PNP is more united than the JLP. Again, it may be just cause for concern when the poll taker reveals that the majority of Jamaicans although convinced that we are going in the wrong direction are still unwilling to change course. Could this be telling us softly that we are generally a blinking bunch of blinkered buffoons!

I have another basic reason for doubt. Scientific polling requires that the target group be representative of the whole. The pollster usually puts his questions to about 1,200 persons; and if he is trying to find out which party will win more seats he should, to be precise, select his audience from 60 separate constituencies, meaning an average of 20 persons in each. Each group of 20 must represent the average opinion of the voters in the relevant constituency. The trouble is that the voters' list in the constituencies are quite different as to size; and while people tend to vote along party lines each constituency is somewhat influenced by the personality of the candidate as well as the competence of his election machinery. Therefore, I'd like to know how the pollsters manage to draw a reliable, representative pattern out of these conflicting situations?

Then again, our pollsters seem to be apishly patterning the American presidential contests by asking about the comparative national popularity of Mrs. Simpson Miller and Mr. Bruce Golding. Our party leaders do not run directly against each other and in my view the personal popularity of the party leader in Jamaica has not mattered much since Bustamante swept the polls in 1944. He was the only politician who could effectively instruct voters to support anyone he personally chose to put up as a candidate; and indeed all the JLP candidates of '44 were named by him. Since then, party organisation and performance have been the more important and influential factors.

Despite his massive and enduring popularity Bustamante's party went into reverse after '44 and was beaten in 1953 largely because of the grassroots organisation done by the TUC and PNP field workers. Michael Manley was always far more popular than Edward Seaga, yet the PNP was swamped in 1980 and shied from battle in 1983. Mike's triumphant return in 1989 was due as much to the JLP's internal squabbling as it was to the PNP's skilful organising and its united thrust. This to me makes nonsense of the idea that Mrs. Simpson Miller's popularity with the opinion pollsters will have anything of significance to do with the outcome of a general election.

Another complicating factor is the sizeable group that includes all persons who, for whatever reason, obscure their opinion by simply saying, 'I don't know.' As any police investigator in Jamaica will tell you, this answer, rather than being a truthful statement is more often than not a shield against the possibility of negative consequences. The violent reality of the present situation induces many a voter to keep his opinion to himself. Then again, the measure of this caution is sometimes determined by the gender of the questioned. So, to get a correct assessment from a 'representative group', the pollster has to be careful about how many males and how many females are saying 'I don't know'.

There are occasions on which persons give the pollster a definitive answer when they should truthfully say, 'I don't know'. For instance, when one pollster asked which party is the better at governing, he admittedly included individuals whose personal experience was confined to one party only. This is not the sort of base upon which serious plans should be laid. Still, it was held up by the media as an example of Jamaican judgment. I wonder!

My criticism does not mean that there is no place for poll-taking. It is indeed useful in a democracy to know how people feel about important issues. The people are the source of power and their opinions should be taken into account by government. Still, we must be careful about how we draw conclusions; careful not to confuse opinion with fact, or be satisfied with the quick answers of unknown persons rather than the reasoned arguments of political leaders seeking a mandate to chart the course of our affairs.pollsnot a shield against the possibility of negative consequences. The violent reality of the present situation induces many a voter to keep his opinion to himself. Then again, the measure of this caution is sometimes determined by the gender of the questioned. So, to get a correct assessment from a 'representative group', the pollster has to be careful about how many males and how many females are saying 'I don't know'.

There are occasions on which persons give the pollster a definitive answer when they should truthfully say, 'I don't know'. For instance, when one pollster asked which party is the better at governing, he admittedly included individuals whose personal experience was confined to one party only. This is not the sort of base upon which serious plans should be laid. Still, it was held up by the media as an example of Jamaican judgment. I wonder!

My criticism does not mean that there is no place for poll-taking. It is indeed useful in a democracy to know how people feel about important issues. The people are the source of power and their opinions should be taken into account by government. Still, we must be careful about how we draw conclusions; careful not to confuse opinion with fact, or be satisfied with the quick answers of unknown persons rather than the reasoned arguments of political leaders seeking a mandate to chart the course of our affairs.

Ken Jones is a veteran journalist. Email: alllerdyce@hotmail.com.