Find us on Google+

Saturday, 11 October 2008

Giving rigging a chance...

When I first heard ECZ had decided to print 4.5m ballots, I immediately asked a friend how many registered voters we had in 2006 because I was too lazy to look it up myself. The answer came that it was 3.9m. So I subtracted 100,000 who I know may have passed away or simply never got round to re-registering (no new registrations are allowed for this election), and I concluded ECZ had printed 700,000 additional ballots. The question then came...why or why? I never blogged it, until today when I have just read that all political parties are equally worried.


  1. Does the ECZ have ballot spoilage figures from previous elections available? The number they have apparently printed would seem to assume both 100% turnout of registered voters and spoilage of up to 1 in 6 ballots at every polling place. This would seem to be a problem easily solved by a standard formula for all elections along the lines of #registered voters + %ballot spoilage, combined with accountability for accurate tracking of actual spoiled and unused ballots.

    In other words, the mere number of ballots printed should not be cause for concern in a transparent, impartial and uncorrupted electoral process. The fact that it does raise such widespread concern in the current circumstance I think says volumes about public perception of electoral fairness.

  2. I think the standard formula would work well, especially in Zambia with life expectancy at 35. Simply put I suspect the "deaths" since 2006 more than compensates for the margin of error of "% spoiled".

    You are quite correct that that the "mere number" should not cause concern, though the 1 in 6 number seems too unproductive. If a company produced ballots with that level of spoillage, you would have to assume the government does not pay for the addition 600,000 since it is a printing error? But which printing company would agree with that?

  3. My apologies, I was unclear, by "ballot spoilage" I mean ballot papers which are damaged or mistakenly marked contrary to the voter's intent prior to being placed inside ballot boxes, necessitating replacement. Now that you mention it, I suppose that number could also include any printer's errors. That said, 20% would seem abnormally high. "Spoilage rate" is sometimes also used to refer to ballots from boxes which cannot be properly counted after polls close, and so for clarity I am only referring to that portion of total spoilage that is detected by the voter themselves.

    Rates of voter-undetected spoilage (ie detected in the counting stage) for most national elections worldwide usually range from around 1% to 5% (US 2000 Presidential Election the rate was 1.8%). While I do not have access to figures on the relative number of detected vs undetected ballot spoilage, for the moment I find it reasonable to assume they are roughly equivalent (errors are often by their nature undetected). On that basis we would expect only a small percentage of replacement ballots required by polling places even with 100% voter turnout.


All contributors should follow the basic principles of a productive dialogue: communicate their perspective, ask, comment, respond,and share information and knowledge, but do all this with a positive approach.

This is a friendly website. However, if you feel compelled to comment 'anonymously', you are strongly encouraged to state your location / adopt a unique nick name so that other commentators/readers do not confuse your comments with other individuals also commenting anonymously.