The latest opinion poll by Infotrak suggests that in a presidential election held today Raila Odinga would receive 35 per cent of the vote, Uhuru Kenyatta 18 per cent and Musalia Mudavadi 8.5 per cent.

In such a scenario however, the election would go to a second round of voting with just the top two candidates going through. The Infotrak poll then shows Odinga beating Kenyatta by 52.4 per cent to 47.6 per cent, and Mudavadi by a larger margin of 53.9 per cent to 46.1 per cent, if Uhuru Kenyatta did not stand in the election.

Odinga supporters might be buoyed by the fact that he holds a substantial lead in the headline figures but he and his campaign advisors will be concerned that his poll rating isn’t nearer the 50 per cent winning line. Kenyatta advisors will surely be concerned that at present he is ‘flat lining’ in the polls. Mudavadi will be encouraged that although his headline support figure is low he does seem to be the second preference for voters from across the spectrum.

THE STAR’S CONFUSED REPORTING

The picture however is somewhat confused, not least given the manner in which The Star newspaper covered the publication of the figures.

It was either a very good story, or just as likely someone decided that they’d got it wrong first time round and need to try again, because the The Star ran coverage of the same poll over two days which would have been fine if only the figures they published added up or matched the graphics that went with them.

For example, The Stars weekend coverage of the first round poll figures had a front page graphic showing Uhuru Kenyatta on 17.3 per cent but a text that reported him as having 18 per cent support among voters at present.

The Star’s figures did not add up in other ways. The first round results, as reported did not at add up to 100 per cent but rather 88 per cent. Even if The Star had just discounted the ‘Don’t vote/won’t vote’ (seven per cent), it still didn’t add up. The second round figures published, however, did add up to 100 per cent which suggests The Star, or perhaps Infotrak, calculated or presented them in a different way.

INFOTRAK/STAR POLL – SOME QUERIES

The Infotrak opinion poll was based on a large sample – 2,400 respondents – and if conducted professionally, should be reasonably accurate (+/- 2 per cent) but the Forum at least questions some of the published findings.

The reported poll suggested that issues concerning voters most are corruption, mentioned by 22 per cent of respondents, and the implementation of the new constitution, cited by 13 per cent. These figures are totally at odds with other recent polls which, more plausibly to the Forum, show the major concerns among voters being the need for jobs, concern at rising prices, and economic security.

There are other facts the Forum would like to know regarding the poll and that it would be useful if Kenya’s newspapers would include in future in the coverage of such surveys. Were the figures ‘weighted’, for example, i.e., was the raw dated adjusted to match the demographics of Kenya’s electorate? And who commissioned the opinion poll? Was it The Star, Infotrak, a politician or political party, or an international organisation?

THE ELECTORAL OPINION POLLS BILL

This last question is important not just because it would be interesting to know, or useful to know so that readers can judge for themselves if they think there is any bias in what is reported, but also because it is now required under Kenyan law.

Last week President Kibaki signed into law the ‘Electoral Opinion Polls Bill’ that requires the disclosure of the sponsors of opinion polls and the banning of polls being published in the last 48 hours of the election.

The Kenya Forum hopes that The Star will now publish who sponsored the opinion poll reported by them at the weekend and on Monday, clarify the figures that they published, and employ sub editors.

Forum readers may like to try and decipher The Star’s reporting of the poll for themselves, see ‘Raila down Mudavadi up in poll’ and ‘Raila close favourite in runoff – poll’.