Winning an election is as easy as making githeri!

By MUNGAI KIHANYA

The Sunday Nation

Nairobi,

04 November 2007

If any the leading presidential candidates can attract 50,000 supporters to their respective campaign launch rallies, are the recent opinion polls credible in indicating wide gaps in the popularity of the aspirants?

To understand what is going on, suppose you were making some githeri. After mixing the maize and beans in the pot, how would you determine the ratio of the seeds quickly?

You can scoop out a cupful of the mixture and count the number of maize and beans. Then the ratio of the seeds in the cup would be the same as that of those in the pot. This method is similar to an opinion poll: You scoop out a cupful (small sample of respondents) and use the ratio of seeds (candidate preferences) in the cup to determine the ratio in the pot (the whole population).

Now suppose further that you have already mixed the maize and beans in the pot and before putting it on the fire you remember that you need a cupful of beans for another meal. How do you that get them out of the mixture? It’s easy; you carefully select the beans (one-by-one) and put them in the cup…until it fills up.

This is similar to a political rally for one candidate: only the desired type of seed (the people who support that candidate) goes into the cup (the rally venue). As demonstrated last week, the Nairobi venues take about 50,000 people – that’s an easy task for the main presidential aspirants.

Now we have seen an opinion poll and a political rally, but how would the githeri simulate an election? Well, suppose you have mixed the seeds in the pot – with more maize than beans (by the way, this makes better githeri!)

Then you call two people and ask one of them to pick out the maize and the other the beans in a duration of, say, 30 seconds. The time limit is important because it will ensure that they don’t empty the pot.

Now, although there is more maize than beans in the pot, the person picking the beans could have faster fingers and therefore end up with more seeds at the end of the exercise. This is what happens in an election – the candidate who can get more of his supporters to go out to vote (faster fingers) wins. (I once won an election with only 10 percent support!) For this reason, some statisticians contend that a random sample opinion poll is a more valid measure of preference than an election.

This also explains why politicians give money to voters on Election Day. It is not a bribe to change their choice. No. The cash is given to supporters – not to the opponents! It is simply compensation for going to the queue to vote. Remember, this is a country where half the population survives on less than Sh60 per day. Spending Sh1,000,000 can easily get a politician 10,000 votes (Sh100 each) – enough to win a parliamentary seat!