The good, the bad and the ugly: How plausible are opinion polls?

Following the outcome of the August 2017 General Election, a mixture of anger and contempt has been showered on the pollsters and their poll results by all and sundry across the political divide. But arguably, election polls still remain an important source of political information in all modern democracies.

The margin difference in the poll results for the final survey done days before the Election Day between the two major pollsters in Kenya raised questions on the credibility of either of two pollsters and sought for more clarifications on the methodologies and sampling done.

While Ipsos Synovate put Jubilee candidate Uhuru Kenyatta at 47 per cent ahead of NASA’s Raila Odinga at 44 per cent, Infotrak set Odinga at 49 per cent ahead of Uhuru at 48 per cent.

Whereas the outcome in these polls could have shown a reduction in the percentage of the undecided voters as implied by Infotrak’s results, more queries will rise on the clarity of the questions asked and the limitations of the surveys by both pollsters.

Tools of democracy

That said, before we could demonise polls in general and accuse them of being manipulated, it is imperative to recognise how we generally acknowledge these political surveys and sometimes fail to affirm their legitimacy for reasons we often don’t say. Admittedly, polls are very important and useful tools for democracy especially in nations where the leaders want to see or wish to be seen as supporting the will of the people and having their voice heard.

To a large extent, polls will change perceptions and opinions of others, not just about whom to choose as their next president but also on the course of action in making economic and diplomatic policies.

Intriguingly, in Kenya’s opinion polls, legitimacy depends on which side of the political divide you support and how the polls portray your preferred candidate. In the instance that your candidate seems to be doing well in the polls, the polls become ‘credible’, and they fail to be so if your candidate is portrayed to be losing.

In a research article published in June 2003 by Thomas E. Mann, people are always skeptical about polls when opinion moves in the “wrong” direction. This, of course, is accompanied by questions such as; did they even ask the right questions? Notwithstanding, we will often find ourselves using the same polls to push agendas, often when they sound favourable.

Politicians, business leaders, and journalists should pay attention to the polls since they do matter and can shape their course of action. If therefore, public opinion about polls is largely affirmative, there must be an inherent intruder that invades the credibility of the polls and comes on and off only to see that their interests are met. In that regard, persistent and consistent professional work in the forecasts has never been more desirable.

When interested parties in the outcomes of the opinion polls are involved in seeing through the success of their preferred outcomes, then we lose the legitimacy of the polls and ultimately the voice of the public is not truly articulated as it ought to.

When we trust our pollsters but then do not receive a collateral and corresponding expected level of legitimate accomplishments from them, questions of manipulation arise.

In the 1980 US elections, when the incumbent President Jimmy Carter was veritably crushed in the race by his challenger, Ronald Reagan with a landslide victory, the politicians and the public alike revisited the polls to anatomise and lay bare how pollsters had such skewed and inaccurate predictions.

Political agenda

Whilst the Kenyan pollsters in this year’s elections made efforts to ask questions related to the particular presidential candidate and the economic challenges that the electorate wished their next president should pay more attention to, they occasionally failed to drive the political agendas to levels of swaying both the politician and the electorate for common good. Actually, the opinion questions might have aided in retaining the country’s ethnic-based voting pattern.

In addition, we would need more clarity on the selection of the respondents taking into consideration that the country’s voting pattern where the normal sampling methodological ways might not work effectively to show the voice of the nation.

The use of old provinces to assert the voice of the electorate was also not consistent with the 2010 Constitution’s requirement that the winner garners at least a 25 per cent vote in at least 24 counties.

While we cannot do away with the polls, we need to strengthen the regulations of the pollsters. But remember that polling is a tool, not a principle.

Mr Boi is a Monitoring and Evaluation Specialist. [email protected]

Business
Premium Ruto's food security hopes facing storm amid fake fertiliser scam
Real Estate
Premium Affordable housing: Will State's data-backed action now pay off?
Business
Premium Nairobi business community plans protest as over 700 containers held at port
By Peter Muiruri 40 mins ago
Real Estate
Premium Building to the skies, but at what cost?