An opinion poll is a better indicator of popularity than an election

By MUNGAI KIHANYA

The Sunday Nation

Nairobi,

31 July 2022

As we get closer to general election scheduled for 9th of August this year, many agencies have conducted political opinion polls and published their result in the mainstream as well as the internet media. A trend has been observed that the popularity of presidential candidates changes depending on which company did the survey. Hence, several readers have asked me if opinion polls are reliable.

In the run up to the 2007 general election, I wrote a comprehensive explanation on the reliability and limitations of opinion polls. In summary, if properly conducted on a random sample of the target population, the survey will give a very accurate measure of the popularity of candidates. However, it is not a good predictor of election results.

The reason is that elections are not determined by popularity alone; an equally important factor is voter turnout. And this is not just about the number of voters who turn up, but their preferences. One candidate might be better at mobilising voters than the other and this makes a big difference in the result.

Politicians know this very well and so they facilitate supporters to go to the polling stations. It’s not a case of bribery! It is fare to travel to the polling station or even compensation for a lost day of work. Therefore, it can be argued that the winner of an election is not necessarily the most popular person.

As mentioned in the 2007, I once won an “election” with just 10 per cent popularity! This was during my days as a university student in a foreign country. There were just 500 of us foreigners in a population of about 5,000.

We needed to make a change in the students’ union constitution. We knew that normal attendance at the monthly general meeting was around 200 people. So, we mobilised the foreigners – made them believe that their future well-being depended on the proposed changes.

On the day of the general meeting, there were over 500 people present – the usual 200 plus our 300 foreigners. When the vote on the motion to amend the constitution was called, it was passed by over 75 per cent!

That’s how a small well-organised group can determine the fate of the majority. For this reason, it might be argued that an opinion poll might be a better indicator of popularity than an election. The big question, however, is whom can we trust to conduct the survey?