Steady Steadman? PDF Print E-mail
Written by Don Seda   
Tuesday, 16 October 2007

Earlier this year, when the Steadman polls placed Raila Odinga's ratings at a mere 12%, against President Kibaki's 45%, and Kalonzo Musyoka's 20%, the pollsters came in for a bashing from Raila's backers.

They were they said, alert to the bare-faced effort at boosting the President's chances at the polls, making him look like he has an unassailable lead. The ODM-K party and its supporters promptly pronounced George Waititu and Tom Wolfe as frauds, government agents hired to kill off the oppositions chances.

Neither did it go unnoticed that Mwingi South MP Musyoka had a higher rating than did his Langata counterpart Raila Odinga. This we surmised was part of the scheme to foment discontent and open cracks in the solid ODM-K fraternity, as the polls reflected a view dissonant with the reality in ODM circles.

polling

Kalonzo's supporters in a show of unity spoke with a voice consonant with that of their counterparts in the opposition movement, joining Raila's supporters in discounting the rectitude of Steadman. It was the widespread belief then that the organization was a government front, working to garble the mood on the ground by amplifying the President's popularity.
Still, deep in their hearts of hearts, Kalonzo's supporters had their calculators out, crunching the numbers and building plans on the basis of Kalonzo's 20% reserve. Extrapolated, these figures were a harbinger to a General Election face-off with President Kibaki and Raila Odinga was suddenly the bad man, the obstacle on the path to Kalonzo's date with destiny.

The President's faction on the other hand heralded the poll results as a true reflection of his popularity, and an endorsement of the economic progress made under his government. Those discrediting the figures were described as passionate and unrealistic fanatics of the Langata MP's, out of touch with reality and ignorant on the workings of opinion polls.

Six months down the line, the good men at Steadman now place Raila at a colossal 53%, sixteen points ahead of Kibaki's 37% and above Musyoka's 8%. These results now seem to be received with much greater approval in ODM circles, the difference in the two figures a declaration of the success the party has achieved in seeking to promulgate its promise. At the same time however, the Kalonzo team and the President's backers are seeking to dismiss the latest opinion polls, labeling them fabrications, that do not portray the true picture of the political developments in the country.

poll.jpg

So what to make of these polls and what really are Steadman up to? Who pays them, and who do they earn this pay? These are but a few of the questions relating to the riddle of the pollsters, and the efficacy of their methods. The vacillations are certainly out of step with global trends, and Kenya is either an exception to global rules, or something else is at play.

The first and perhaps most readily advanced answer is simply a natural and stubborn inclination by political factions to see only the reality that they desire. Loyal supporters will therefore resist any attempt at portrayal of their candidate of choice as trailing or having fallen behind their competition. Following this thread, the reasoning is that parties will validate the results whenever their respective candidates are rated highly, but treacherously disown the very same pollsters under when the ratings dip.

The second idea, mostly from supporters of the ODM candidate is that Steadman was previously shoddy, disorganized and even biased in its polling. Pressure from the public and from the Odinga led party however, woke the pollsters up to the reality, and they quickly saw that it would be dangerous to their continued relevance if they posted figures that were clearly not reflective of the mood on the ground.

The third idea, related to the second posits that the public should maintain its distrust of the Steadman group, as it is a partisan advocacy group working in the interests of the Kibaki re-election campaign, by providing push-polls to its benefit. As evidence, it suggests that the ODM leader's ratings did jump after the mammoth KICC launch but that his figures were never as low as portrayed to begin with. To this group, Steadman is culpable for creating a mythical perception that a Kalonzo Presidency was a possible reality, what perception placed a stumbling block on Raila's path to becoming the ODM-K nominee.

miracle_man.jpg
 miracle man?

This version of events then sees Steadman continue in their dirty work, by portraying President Kibaki as having built an insurmountable lead over his competition. This tactic accounts for the massive 45% Kibaki enjoyed over Kalonzo's 20% and Raila's 12%. So suddenly, the public is seen to have woken up to the government's potent economic credentials as attested to by his improved approval ratings. These impressive figures would also play the auxiliary role of creating a solid magnet that would attract political alliances from left right and center, thereby making it virtually impossible for any opposition traction to be built. Our fickle politicians, backing as they always do the winning horse, would be attracted to the president's team, persuaded by the polls that his win was a foregone conclusion.

According to this theory, the Waititu outfit, now acknowledging their prior miscalculations, has finally reverted to one last "bait and switch" move. This has them show Raila's true numbers (53%) to the public, a move that serves to tranquilize an electorate that still distrusts Steadman and then, with elections drawing ever closer, to bank on this trust to raise President Kibaki's numbers steadily while pushing those of Raila to a nose dive. Hypnotized and confused by Steadman's previous friendliness to Raila, the public will not see the trick and will then cast their votes based on a perceived change of attitude towards a Kibaki presidency.

The fourth and most ugly alternative explanation is that the pollsters are part of a colossal fraud that is preparing Kenyans for a rigged election. It follows the exact same trail as the third explanation except that Steadman's motive would be to set a poll rating for the President, somewhere in the region of what would work to validate rigged polls. The final numbers are therefore, unlike in the previous scenario, not intended to persuade voters, but instead to corroborate the rigged numbers at the actual elections.

We are it is true new to the concept of public polling, but the wild swings in the polls are a reason for extra vigilance. It remains to be seen exactly what sort of ride Kenyans are being taken for.





Digg!Del.icio.us!Google!Facebook!Technorati!StumbleUpon!Newsvine!Yahoo!Ma.gnolia!Free social bookmarking plugins and extensions for Joomla! websites!
Trackback(0)
Comments (7)add
0
...
written by Timothy Wainaina , October 16, 2007
What a grand conspiracy!! There is of course a law against push-polling and any other such naughtiness. Anyone who believes Waititu's tactics are not up to standard may report them to the police, or even take them to the courts themselves.

It is after all, ILLEGAL, to post fraudulent opinion polls. Here's a link to that law.
report abuse
vote down
vote up
Votes: +0
0
...
written by Gichangi , October 16, 2007
I admit to having had taken Steadman seriosuly when their polls whowed results favourable to my favoured candidate. However, we must be careful at paying too much attention to these results- not just Steadman's but all of them- especially because in urban areas, it is hard to give a view that is reflective of the true composition of that constituency.

For example, if I was polling in Mombasa, how do I ensure that I seek out business people, the Akamba of Changamwe, the Luo of Kisumu Ndogo, the Arabs and the Mijikenda resentful of both the presence of Luos and wabara wengine, and all those squatters with a grudge against the likes of Balala?

Is it possible to reflect all these constituencies in a sample of 60 people? Nairobi is even worse! Like we saw with the Referendum, thse polls will vary wildly from the final result. ODM supporters tayari are making excuses, so they can go to court for five years alleging irregulations.
report abuse
vote down
vote up
Votes: +0
0
not the author
written by Don , October 16, 2007
Why are we singling out Steadman, when three other opinion polls have shown similar trends? Unless all these pollsters use common sample and sampling frame, it'd be foolhardy to purport data fudging. The swings in Steadman numbers, apart from the 51% Kibaki rating, have been plausibly explained before. At first, the political map was fuzzy with Kibaki pitted against numerous aspirants drawing from a common vote bank, the ODM family; hence the joke that Steadman was rating a "Kibaki competing with himself." Even then, Kibaki's numbers consistently remained a minority of the total. Now that things have crystallized, the common pool vote has consolidated into two candidates, raided the Kibaki vote bank, and drawn in the hitherto undecideds. To a statistician or an opinion polling expert, the numbers are therefore hardly surprising.

Since 2002, Steadman's political barometer has done pretty well. It was right on the money for the 2002 GE and the 2005 referendum. In both instances, the final Steadman poll and actual results differed by exactly the margin of error, which positively swung the opposition way.

I expect accusations of being a Raila supporter hence my defense of Steadman, that's fine. But for those who care, my experience in applied survey research and statistics are vast enough to give me confidence in what I'm saying. Of course, statistics is just that, a game of averages. Someone once illustrated the folly of statistics thus: "if you take a live person, put half the body in a freezer and the other half in the oven, statistics will report that the body temperature is average." It's possible that Steadman is reporting an average temperature for a "dead" man!!
report abuse
vote down
vote up
Votes: +0
0
...
written by emmo opoti , October 16, 2007
The size of the crowd at the Kalonzo rally tells me that many Kenyans are still sizing up the candidates and although they may be bent one way or the other, their minds are not yet made up.

In addition, the gains candidates may make in these campaign times, will take a while to percolate to the rural areas and even to build up any momentum in the urban areas. This is especially true of those in the rear who are working hard to catch-up. In the same way, Raila's trip to the NSE, for those who it will win over, is not likely to register immediately but over the next two weeks, some investors who were previously wary of the ODM leader may see those doubts beginning to subside. We have after all got one dominant picture of these candidates, and the campaign period gives them an opportunity to portray themselves in a very different light.
Anyone who thought Kibaki was an indolent geriatric will have had such doubts questioned by his walkabouts, and so on. This process will continue on and on until election day, especially because there really is no difference in the composition of the two parties and neither do the majority of the wananchi take their pledges too seriously.

It is for this reason that I think the ODM would be properly terrified by a delayed election. Already there are jitters about where Raila should stand, and Musalia aside, many in the Pentagon may not be going back to Parliament.
report abuse
vote down
vote up
Votes: +0
0
...
written by Tim Norwood , October 17, 2007
Don,
Your language says it all. Why is it all the polished people in the ODM camp are ashamed of being there? :D
Going along with your most excellent explanation, I'd say the alignments are not over yet. The dust has not yet settled down on the Kibaki team's formation either, and while that may not hold much hope, it nevertheless may shift a large number of votes.

There's an assumption in many pundits minds that Kenyans are rigid about this thing we are calling change. They really are not, if they were Kibaki would never have enjoyed those high approval ratings you yourself attest to. I think we are still in for major major movements, and potentially for events in the next months that could decide the outcome once and for all.

P.S. I am not really backing Kibaki's bid either, it is just the mayhem and the psychosis I am afraid of.
report abuse
vote down
vote up
Votes: +0
0
...
written by Stephen Wanyama , October 17, 2007
Polling is hard work, and in a country like ours where not too many people have phones in their homes, this is not an accurate science. There is a margin of error on some of the polls, I do not know about the Steadman ones. As Gichangi says also, it is very difficult to predict anything in our urban areas, especially with voters who may not vote there and with such different interests. He points out Mombasa, but I think Nairobi is much more difficult to predict.

Finally, given that there are costs to polling, and a lot of trust is placed in the pollsters, is it not possible that pollsters fill some of those forms themselves? I have certainly had some of them do that in-front of me. Why go house to house, or stop people in Kisumu streets when you can very easily make up the numbers?

I don't think either of the two main parties relies too much on the polls, but the public does hence the relevance of Seda's theory on push-polling. Wainaina says it is illegal to play with polls, but I am not sure that matters too much in our country. What risk of being caught, and by who?
report abuse
vote down
vote up
Votes: +0
0
Reporting of Polls
written by githush , November 15, 2007
The discussion above is quite interesting and the difficulties of polling in Kenya quite clear. My main concern with the polls released in Kenya is the lack of clear information on methodology, respondent breakdow etc... I quickly perused the steadman website and found the most recent survey: http://www.steadman-group.com/..._09_07.pdf
whereas it has more information than what was reported in the news media, it is still lacking in detail, as opposed to the IRI/SPRR survey in September: http://www.iri.org/africa/keny...-Kenya.asp

The latter doesn't ask the "who'd you vote for" question, but the whole survey is (methodology, sample breakdowns etc.) is there to be seen and critiqued. Now that is how a poll should be reported.
report abuse
vote down
vote up
Votes: +0
Write comment

security image
Write the displayed characters


busy
Last Updated ( Tuesday, 16 October 2007 )
 
< Prev   Next >


Archives | About Us | KenyaImagine How To | Privacy Policy | ContactUs | Join KenyaImagine |  Advertise Here| Legal Disclaimer | Terms & Conditions | Directory
rss-2.png

 

Copyright 2009 KenyaImagine.com, the KenyaImagine logo and KenyaImagine.com are trademarks of  The Imagine Company