What’s weather got to do with voting?

The honest answer: nobody is totally sure. There has been little research into this subject in South Africa, but international studies have found that the weather can affect voter turnout.

One 2017 study found that in the US, for every 10°C increase in temperature, voter turnout rose by 1.4%.

A general truism of election turnout is that the higher the cost voters incur to vote, the lower the likelihood that they turn up to the polls. “Cost” doesn’t mean financial expenditure: it refers to anything that makes the act of voting difficult or inconvenient. Rainy or cold weather constitutes a potential cost.

Complicating the matter is the fact that the findings in the US have not been replicated in other western democracies.

Research undertaken in Sweden has failed to find any meaningful impact of weather on voter turnout, but this is speculated to be because the costs of voting in Sweden are very low: Citizens do not have to register ahead of time and voting takes place on a Sunday. (One might also suspect that it has something to do with Swedes being stoically accustomed to adverse weather.)

The costs of voting in South Africa are higher, with the requirement of pre-registration, a valid bar-coded ID document and the fact that even on a public holiday some citizens still need to work. So voting is already a bit of a schlep: Throw in bad weather, and it may be enough to discourage people from heading to the polls.

In 2014, the IEC estimated that the average South African voter had to wait 16 minutes before casting their ballot – a relatively low time investment, but one made less appealing in bad weather.

In the last general election, the IEC also reported that 4% of South African voters had to travel for 60 minutes or more to get to a voting station. These were voters in deep rural areas, who were also more likely to be poor. Combine the financial outlay of an hour’s travel with bad weather, and you have a strong incentive not to make the trip.

But South Africans love voting, don’t they?

It’s true that South Africa has had a much higher voter turnout since 1994 than most western democracies – probably for the obvious reason that the majority of the population has only been able to vote for the past 25 years.

The general turnout trajectory has been on a slow downward slide: from 77% of registered voters turning up in 2009 to 73% in 2014. This is normal for maturing democracies.

Ahead of the 2019 polls, 74.5% of the eligible voter-age population is registered to vote, down from the 80.5% registration rate for 2014.

Of the registered voters, it is impossible to know in advance how many will turn up. A recent Ipsos poll suggested that as many as a third might stay away on the day.

International electoral research suggests that other factors that play a role in turnout, beyond weather, include the perceived closeness of the electoral race and the level of the perceived difference between electoral contenders. If voters perceive leading parties as being very different choices, they are more likely to vote. Where there is a sense of ideological similarity between contenders, turnout tends to drop.

From this perspective, if turnout on 8 May is relatively high in the end, South Africa may have the EFF to thank – since the radicalism of the Fighters’ policies strongly differentiates them from the two biggest parties, the ANC and the DA.

Which parties stand to benefit in South Africa if voters stay away?

Hard to say. In the final Ipsos poll, released on Tuesday, the polling company suggested that the ANC stood to benefit both from a high voter turnout scenario and a low voter turnout scenario.

It said: “In a low voter turnout scenario, the ANC voters seem to be more committed to vote, and therefore the party support increases notably. In a high voter turnout scenario, the ANC succeeds in convincing voters, who would otherwise not have voted, to come out and participate. Thus, both a low turnout and high turnout will be to the singular advantage of the ANC.”

Clear as mud. Ipsos said that for the smaller parties, turnout level was unlikely to have a major effect on results.

Other political analysts, however, have suggested that a lower voter turnout is likely to cost the ANC most acutely.

So what’s the weather going to be like on elections day?

Accuweather reports that showers are predicted in parts of the Western Cape and the Free State in the morning. The Eastern Cape faces thunderstorms in the afternoon, as does the Northern Cape.

In Durban and Johannesburg, however, sunny and warm conditions are predicted – doubtless music to the ears of politicians given that Gauteng and KwaZulu-Natal are two of the most hotly contested provinces.

Could one argue that the weather might already have had an impact on the elections?

Indeed one could! Rain has already tangibly affected voting in KwaZulu-Natal, in terms of the fact that one effect of the recent floods was that hundreds of people reportedly lost their ID documents. Home Affairs and the IEC have been scrambling to make a plan for these people to enable them to vote.

And if you’d like to take things a step further, one could say that a lack of rain in the past might affect the results in the Western Cape – that is, if voters decide to punish the DA for the handling of Day Zero and the hike in water tariffs. DM