A technician prepares a results board at the Electoral Commission of SA headquarters in Pretoria. Picture: REUTERS
A technician prepares a results board at the Electoral Commission of SA headquarters in Pretoria. Picture: REUTERS

For an organisation that only has one job to do and a lot of time to prepare for it, the Electoral Commission of SA (IEC) cannot be exactly proud of its showing in putting together the May 8 elections. 

Elections are by nature not the easiest things to organise, and SA has its specific challenges with voting  stations in diverse areas, from densely populated cities and informal settlements to nearly unreachable rural corners of the land.

For 25 years, IEC officials have navigated these challenges and have generally done a good job in ensuring that South Africans exercise one of their most basic democratic rights. There may have been glitches now and again, but the integrity of our polls has never been questioned.

But from the very first post-election media briefing this time, it became clear that something was amiss. And when Business Day took a closer look at the numbers they raised more questions.

Challenges were reported from early on, including voting stations running out of ballot papers in the Western Cape. This could be due to the fact that voters could cast their ballot at a district in which they were not registered.

Challenges were reported from early on, including voting stations running out of ballot papers in the Western Cape.

As a result, some voting stations recorded a voter turnout greater than 100%. Of those, Business Day managed to find four where the numbers simply didn’t add up.

At one station in Tshwane, the total number of registered voters plus the number of those who completed a section 24 form, which had to be filled in to apply to vote in a district where one wasn’t registered, fell 375 votes short of the total ballots cast.

The IEC was unable to immediately account for the discrepancy, which is a concern in itself given the widespread allegations of voter fraud. The numbers do not confirm foul play but do suggest, at the very least, flaws in the process.

Procedural inconsistencies expose the entire system to potential abuse. On the scale of an election, it becomes nearly impossible to ensure compliance if certain figures can only be verified by the word of an official.

While Business Day only managed to identify discrepancies at four voting stations that had a voter turnout higher than 100%, a large turnout is not a prerequisite.

With the limited information available, these are the only voting stations at which an external party can identify a discrepancy because the total number of votes needs to be greater than the total number of people registered and total number of those who completed section 24 forms.

Third parties are unable to determine whether everyone who was registered did, in fact, cast their ballot, which means a 100% turnout of registered voters for the station in question is an assumption. In other words, while a voter turnout of over 100% does not imply fraud, a turnout of under 100% does not mean all votes were valid.

The IEC has since admitted to “negligible risks” of possible voter fraud after the statistician-general’s analysis of 1,020 voting stations identified 13 districts where section 24 votes were significantly higher than those in the ward.

Regardless of their significance, the IEC undertook to look into the unaccounted votes this week. While this might provide a final answer on the four voting stations Business Day queried, questions still remain over those the statistician-general identified, and potentially others not included in the sample.

These discrepancies may not have a material impact on the recent elections, especially where the determining factor is a couple of hundred thousand votes. But with every election that passes, the margins for victory are becoming smaller. It would serve the IEC well to ensure that the margins for error do the same.