Why opinion polls really matter in Kenya

It's the season of pollsters. The next General Election is just a year away. And Ipsos-Synovate's Tom Wolf has wasted no time wading into controversy. Maggie Ireri of Trends and Insights for Africa (TIFA) has also joined the fray. These two are cut from the same cloth. But that's a matter for another day.

Mr Wolf has been strongly censured by allies of Raila Odinga, who accuse the pollster of being "obsessed" with the former Prime Minister. In his latest survey, Mr Wolf curiously delved into the issue of Mr Odinga's age and seemed to "tease out" from respondents the answer to the question of whether the veteran politician should now retire.

According to his findings, 35 per cent of those interviewed want Mr Odinga to retire from politics. Now this line of inquiry, Mr Odinga's allies argue, is conspiratorial and designed to undermine the CORD co-principal's 2017 presidential bid.

But Mr Wolf isn't exactly new to controversy. He has been, according to his critics, on numerous such fishing expeditions before. Mr Odinga's allies are now questioning both his impartiality and the "scientific-ness" of his surveys. They want stiffer laws governing the conduct of opinion polls to weed out rogue pollsters.

Mr Odinga's allies' beef aside, Mr Wolf has succeeded in doing exactly what opinion polls are designed to do: stimulate debate and speculation among the media, the public and politicians.
That debate is happening. And it's a symbol of our evolving democracy. Indeed media coverage of Ipsos-Synovate's opinion polls, including those of all other pollsters, should be regarded as an integral part of free speech - a right enshrined in our Constitution.

It would, however, be wrong for pollsters to sacrifice professionalism at the altar of sensationalism. Opinion polls can arouse very strong passions. Generally, opinion polls focus on voting intentions and voters' leadership preferences.

They're more about a candidates "likeability" and "electability", which might seem like an oversimplification of a very complex and grave matter of picking the right leader.

This concern is aptly captured by Tifa survey of the Nairobi gubernatorial race, which casts the contest as between Governor Evans Kidero and Senator Mike Sonko, if elections were held today.

It illustrates how opinion polls can distract from the real issues by reducing an election into a populist contest. In my view, this Tifa survey merely plays to the gallery. It is bereft of a detailed interrogation of policies, which is what should set candidates for Nairobi gubernatorial race apart.

Yet opinion polls shouldn't be a matter of life and death. First, they can be unreliable. And this goes many years back. In the 1936 US presidential race, when the science of opinion polling was still in its infancy, a pollster predicted an Alf Langdon win. Instead, incumbent Franklin D Roosevelt romped home with a landslide.

In more recent times this fickleness has been witnessed in the UK — not once but twice; in the 2015 general election, when they predicted a hung Parliament, and during the 2016 referendum on the EU, also known as Brexit, in which Britain voted to leave. Other examples abound.

Secondly, and this is by far the most important lesson, which according to researchers studying the science of opinion polling, Mr Odinga's allies can take away from the "offending" Ipsos-Synovate poll.

Opinion polls, they say, can offer immediate feedback to politicians about their party's, or their individual performance. This may in turn positively influence decisions by, and concerning, the party leadership.

The question then is: why would Mr Odinga's allies be so upset with Mr Wolf, if his survey, spiteful as it may seem, can only be beneficial to them as it could possibly trigger strategic internal party re-think? Or, why should Kidero be all smiles after topping the latest survey on the Nairobi gubernatorial race?
Opinion polls can potentially influence voting intentions and behaviour, according to researchers.

So politicians who commission them don't just put their money on an endeavour without returns. They're aware of this impact.

In her 2011 study on public opinion polling in Australia, researcher Rachel Macreadie cites opinion poll commentators, who argue that a 'bandwagon' effect can take place, when polls indicate a clear leader, which suggests that some voters may be inclined to vote for 'the winning side' and therefore vote according to the party or individual with majority support.

So Mr Odinga's allies have reason to be concerned. In a sense, therefore, these findings lend weight to the quest for integrity and professionalism in Kenya's nascent polling industry just as much as they recognise the crucial role of pollsters in electoral democracy.