6 October 2012

Kenya: Opinion Polls Sends Raila Back to the Drawing Board

In modern democracies, political opinion polling or rather opinion polling has become the norm rather than the exception in canvassing public opinion on certain political, social and other issues at a given time.

In Kenya, this is also the case with a number of pollsters coming up with data on specific issues, especially voter preferences ahead of a general election.

On Tuesday, two such surveys by Ipsos Synovate and Strategic Research came out that gave rise to several questions on political polling in the country.

The Strategic Research focused on the status of governance and democracy in the country. The focus was on seven key result areas as follows: Enhanced national reconciliation amongst Kenyans; Working towards formulation of legal electoral and judicial reforms; Access to justice for the poor and marginalised in Kenya; Promoting people-centered devolved government in Kenya; Promoting economic, cultural and social rights (ECOSOC); Enhanced inclusion of previously under-served or marginalised groups and communities, and increased leadership accountability.

The Strategic Research poll, for some reason did not capture attention as much as the one by Ipsos Synovate whose focus was on the presidential aspirants and political parties support ahead of the general election in March next year.

This has become a staple and both The Star and The Standard newspapers led with the story on the results. Not to be left out were blogs and social media users. So let's look at the findings of the latter pollster.

The Ipsos Synovate survey found ODM's presidential aspirant Prime Minister Raila Odinga losing to his two deputies in case of a run-off despite scoring safely above the margin of error against TNA's Uhuru Kenyatta and UDF's Musalia Mudavadi.

Eldoret North MP William Ruto slipped sharply by 50 per cent from 12 per cent in July to six per cent, so was Vice president Kalonzo Musyoka who garnered five per cent in the Synovate poll.

There were also fringe aspirants such as Narc-Kenya's Martha Karua, Peter Kenneth, Eugene Wamalwa, Prof James ole Kiyiapi as well as Raphael Tuju among others.

Masinde Muliro University political science don Prof Frank Matanga weighed in on Ruto-Kalonzo poll numbers to offer the reasons for Ruto and Kalonzo's under-par achievement which indicated a negative shift in their base support.

"Politics is fluid and Ruto has been very much limited to Rift Valley. He has also not been on the campaign trail lately," Prof Matanga said. He, however, did not have kind words for the Vice President whose dwindling numbers he blamed on his perceived indecisiveness.

"Kenyans could be losing faith in the Vice President because of his politics of 'kupitia katikati' and this seems to have destabilised his base that voted for him in 2007," Matanga said.

Others have also argued that Water minister Charity Ngilu's entry into the race to succeed President Kibaki could work against the Vice President.

The campaigns of leading aspirants also pounced on the figures to question Ipsos Synovate's assumption that Mudavadi in particular with a seven per cent first round score against Raila's 36 per cent turn the tables to register 47 per cent in the run-off against Raila's 44 per cent.

Kalonzo's team has dismissed the survey in its entirety calling it 'mambo ya upuzi' (nonsensical). "There are only a few months to elections and then the people will know the truth.

'Hayo ni mambo ya upuzi'," the Kalonzo campaign team said. "Basically they are telling the seven per cent fellow not to run because he cannot make it to the run-off, in case we have one. Where does such a person garner the support to overturn such a huge first round deficit?" Raila for President Campaign spokesman Barrack Muluka said.

Muluka further argued that in a way the poll tries to force other aspirants to drop out of the race and form pre-election agreements around either Uhuru who scores 50 per cent against the PM in the run-off or Mudavadi in order to beat Raila.

"The pollster implicitly tells other presidential aspirants to drop out of the race and support either the TNA candidate or the UDF flag bearer," Muluka argued, backing his candidate's argument that the next general election will be a two-horse race.

TNA's Moses Kuria dismissed the notion of a two-horse race. On the polls, Kuria stated that TNA will take no notice even if they put his candidate in the lead.

"We don't give much interest to opinion polls. We just came from a real opinion poll, the by-elections because the opinion polls in this country still remain a tool of division rather than assisting Kenyans make informed and sound decisions as is the case in the developed world," Kuria stated in reference to the September 17 by-elections for Kangema, Kajiado North, Ndhiwa parliamentary seats and a number of civic seats where TNA candidates fared well.

TNA won the Kangema and Kajiado North parliamentary seats while ODM took the Ndhiwa seat. "We will continue ignoring them and instead build bridges because all these votes -whether they are attributed to the Vice President, William Ruto or any other candidate are ours (as G7) ODM."

The ODM wing while not dismissing the poll challenged the pollsters to stick to the Political Opinion Polls Act that was passed earlier this year. "The pollsters should answer whether they have stuck to the Act. But for us we take every poll seriously in spite of their flaws because they give some information," Muluka stated.

While deviating from the views of the candidates' campaigns Prof Matanga introduced a new element: whether the country is ready for another president from the Kikuyu community.

According to Prof Matanga, there is a "Kikuyu-phobia" which could work against Uhuru and give Raila and Mudavadi the edge. The university don argues that the pollsters should capture such a key factor that could essentially tilt the race.

"We have always said that it (the elections) will depend on who is in the race. The country is not ready for another Kikuyu president and I see Kikuyu-phobia," the political science scholar said.

In saying that, Prof Matanga was quick to add that he does not understand the Synovate analysis that gives Mudavadi the lead ahead of Raila in the run-off despite scoring 29 per cent lower than his former party leader.

"I don't see the logic of the (Synovate) analysis of Musalia Mudavadi turning the tables on Raila or Uhuru from a seven per cent score though he has been making forays in the Rift Valley and Central," Prof Matanga said.

TNA also asked pollsters to move away from using provinces that the 2010 constitution scrapped and go by the 47 counties created under the devolution plan to provide a realistic picture.

Side bar: Limitations of telephone surveys

The latest SPEC (Social, Political, Economic and Cultural Barometer) survey results released this week by Ipsos Synovate targeted all Kenyan adults aged 18 years and above according to the methodology released by the pollster.

The sample size was 2,229 respondents living in urban and rural areas. The margin of error, according to the pollster was +/-2.2. The pollster stated that the data of the survey was collected through Computer Aided Telephone Interviewing (CATI) and fully funded by the pollster.

However, experts have criticised the CATI method. First of all, the method's lack of representativeness since cellphone penetration in Kenya by urban, rural, age and other demographics is not 100 per cent. In such a case, the pollster interviews cellphone owners only.

Those without the gadget were not given any opportunity to participate in the survey thus the level of margin of error could actually be higher than the +/- 2.2.

Secondly, how big is the pollster's cellphone database in relation to the population and the number of cellphone users. If less than 100,000 only used over the last several years, they could have basically been recycling the same panel but without disclosing that this was a tracking poll.

Thirdly, there could be bias because households with no landline telephone or cellphones are not represented. Fourthly, in the era of cellphones, it is very difficult to ascertain the location of a respondent called on cellphone. How did they verify, for instance, the location of the voter?, or if the voter is over 18 or registered to vote?

Statisticians also wondered whether there is a relationship between cellphone users and registered voters or if this is just an assumption. "What this means is that any voter without a cellphone does not stand a chance of being picked for the interview hence the sample is not representative."

In addition, the urban-rural split in Kenya is 65 per cent and 35 per cent yet the pollster did not indicate whether the cellphone database used for the interviews represent this dichotomy and how it is guaranteed.

Furthermore, experts questioned the response and refusal rates which the poll did not indicate, or if its there, was not released to the public.

Ipsos Synovate managing director Margaret Ireri in responding to the limitations noted in the survey stated that "In summary, over the last year we have run several control and experiment surveys for the presidential aspirant support question.

This means that during the same period we have run the face-to-face survey and we have run the (same) question on CATI." "The results have been similar (minimal variance of 1%) . The CATI survey was used as a check to the face-to-face survey. This is part of our usual quality control," Ireri said.

Copyright © 2012 The Star. All rights reserved. Distributed by AllAfrica Global Media (allAfrica.com). To contact the copyright holder directly for corrections — or for permission to republish or make other authorized use of this material, click here.

AllAfrica publishes around 2,000 reports a day from more than 130 news organizations and over 200 other institutions and individuals, representing a diversity of positions on every topic. We publish news and views ranging from vigorous opponents of governments to government publications and spokespersons. Publishers named above each report are responsible for their own content, which AllAfrica does not have the legal right to edit or correct.

Articles and commentaries that identify allAfrica.com as the publisher are produced or commissioned by AllAfrica. To address comments or complaints, please Contact us.