Projections aren’t foolproof


A day after the US presidential election, The New York Times on November 9 last year ran an editorial headlined “Donald Trump’s Revolt”.

President Donald Trump, the editor wrote, are “three words that were unthinkable to tens of millions of Americans – and much of the rest of the world – have now become the future of the US”.

An editorial published in the Los Angeles Times on the same day cried out in bewilderment “President Trump? How did that happen?”

The triumph of Trump in the US presidential race was an outcome that was seen as mathematically and morally implausible. The presidential win by the highly ridiculed Trump flummoxed the world’s most credible media and political gurus who miscalculated the quantum of his electoral support. These pundits failed to accurately interpret pre-election statistics, mood and sentiment.

A severe discounting of the might of the protest vote that was, in the final analysis, the real fuel of the Trump win, added to the monumental blunder of media and political analysts, not only in the US, but across the globe.

In SA in 2012, the ANC presidential race in which President Jacob Zuma came to power, also saw many seasoned political analysts humbled by an outcome that did not subscribe to the numerical formula promoted in the popular press.

Currently, a stream of analyses on the outcomes of the ANC presidential race have followed close on the heels of the final tally of nominations from the nine provinces, with most pundits confident that Cyril Ramaphosa will emerge as the next president of the ANC, well ahead of Dr Nkosazana Dlamini Zuma.

These analyses have been based principally on provincial nomination numbers, without necessary consideration of the weight and distribution of delegate numbers, not only by province but by branch.

The abundance of analyses showing a win for Ramaphosa, have been done on the count calculator of provincial nominations alone and as such may be highly distorted.

The grand visuals of these provincial statistics, which at a glance of support for Ramaphosa, may, calculatedly, or inadvertently, be creating a defective portraiture of the pre-election landscape.

While a breakdown of the provincial statistics show that Ramaphosa has secured about 55% of the provincial nominations, what analysts are failing to capture in their scrutiny of the patterns of support is that the regions and branches that nominated Dlamini Zuma are more capacious than those supporting Ramaphosa.

It is therefore highly probable that Dlamini Zuma could step into next week’s conference with a far more powerful voting powerful voting constituency, numerically, than Ramaphosa.

Until we have a final tally of branches and the size of branch delegations, alongside the choices of league and leadership structures, mathematics will mislead us and manufacture faulty and factional formulae.

As history and mathematics has taught, even the most durable statistically sound calculation can falter in the face of unaccounted for or unforeseen factors.

The X-factor in the election could be a floor nomination. An emotive, highly charged floor nomination is capable of disrupting and dislodging a premeditated outcome, as can the solidarity of a protest vote by delegates who enter the conference under the camouflage cloak of imposed “unity”.

The rather uniform forecasts that have become commonplace in mainstream media over the past few days are largely unschooled and show scant study or comprehension of the electoral guidelines, protocols and procedures of the ANC electoral conference.

There seems, too, to be a little grasp of the makeup and character of the composite voter collective that will determine the outcome of the conference.

Numbers often lose their neutrality in the counting game of election forecasting when applied for calculated partisan purposes, rather than to serve, as they should, as the starting point for unsullied statistical analysis or scholarly interrogation.

When trickery trumps truth, numbers become factional, rather than fact. It is difficult to compute whether media commentators and political experts lack numerical or political literacy or if they punt a decidedly partisan narrative. Both are equally problematic.

The easy distribution of statistics that are unsteady and misleading, by media practitioners and political commentators, as well as by the spin doctors within both campaign teams, may be rendered as propaganda, but nonetheless are potentially very damaging, even dangerous, for it fosters a fake narrative, rooted on a spine of mathematically unsound and incoherent assumptions.

This can so easily create a false belly of expectations on election outcomes, which if not realised, could trigger considerable and irreparable rupture, not only within the ambit of the electoral conference but across the breadth of the nation and beyond.

It would be a tragic indictment on the integrity of opinion leaders, both in the media and political spheres, as well as political leaders within the ANC, if this was indeed their agenda, for such will disrupt and diminish democracy, within and outside of the ANC.

At this late stage its looks like the race is so close it could end in a photo-finish. It is difficult to predict what the headlines on December 21 will be.

Whatever the outcome, the ANC leadership should fix their sights on a post-December paradigm of principled unity across pivotal policy imperatives and implementation. This, rather than the presidential race, will be the real win for ANC.

Kim Heller is a social commentator