Publications
Partners
Covid-19 is crippling SA, and we need the best minds trying to predict its path
South Africa is reeling from the impact of Covid-19. We’ve had – at the time of writing – almost 6 000 deaths, almost 400 000 infections and seen the recovery of a little under 230 000 South Africans.
The toll on society, and the economy, has been enormous. Not only has the pandemic, which made landfall in South Africa less than five months ago, altered our way of living, it has decimated an already fragile economy and exposed a state which has been in severe decline for more than a decade.
There are no signs that the pandemic is slowing down – to the contrary, the velocity of infection is only increasing, with Gauteng and the Eastern Cape emerging as the biggest hotspots nationally.
The only way to deal with this still-unfolding national crisis is to ensure that the state is enabled to act on the best information available. Over the last few weeks, criticism has been levelled at a consortium of modellers whose projections and epidemiological models the national department of health and cabinet have relied upon to chart their course of action.
In today’s edition of the Friday Briefing the consortium – consisting of some the country’s best minds – respond to some of that criticism. They argue that in constructing a model as responsive and reliable as possible, you must acknowledge that there is much that is unknown and a lot that is simply unknowable. The purpose of modelling, however, is using the best available information to try and guide government to enable it to use the vailable resources in the most efficient way.
Quoting a recently published paper on modelling, they say "modellers must not be permitted to project more certainty than their models deserve; and politicians must not be allowed to offload accountibility to models of their choosing".
We’re also publishing a contribution of their biggest critics, Pandemic Data and Analysis (PANDA), an ad hoc grouping of (among others) actuaries who believe the consortium’s modelling is out of kilter.
Today’s Friday Briefing isn’t light reading, but it will broaden your understanding. The virus is having a profound impact on our country – and we need the best minds using the best available tools attempting to predict its path. Whether government acts on it is a different matter entirely.
Best,
Pieter du Toit
Asisstant Editor
Modelling Covid-19 on scarce data and unknowns
Modelling is no perfect science with all the knowns, unknowns and unknowable variables.
By Gesine Meyer-Rath, Sheetal Silal, Juliet Pulliam and Harry Moultrie from the SA Covid-19 Modelling team.
Don’t play around with Covid-19 deaths and statistics
There should not be any manipulation of death statistics during the Covid-19 pandemic.
By Pandemic Data and Analytics (PANDA)
