When — or rather if — the ANC meets in Gauteng on December 16 to elect a new party leadership, many delegates will have the 2019 elections on their minds.

Across two decades of one-party dominance, the movement seemed largely immune to serious electoral challenges, at least outside the Western Cape. This allowed ANC governments to act quite decisively, and sometimes to take unpopular decisions, without the fear of a fatal backlash from voters.

But it also resulted in a significant degree of inattention to the demands of competitive electoral politics in a representative democracy.

In 2016, however, the ANC’s leadership suffered a serious blow to its confidence in the local government elections. The growing disarray of Jacob Zuma’s administration and the sudden prominence of significant opposition parties with black leaders further increased the pressure on ANC election planners.

The electorate itself has been undergoing a quiet transformation. In the first decade of democracy, as many as nine out of 10 voters identified closely with a particular party.

But in recent years, a growing number of citizens have become “floating voters”. This suggests an increasing willingness to use evaluations of party performance to inform electoral choices.

In such circumstances, Nkosazana Dlamini-Zuma would be a curious choice for party president. Indeed, her election would perpetuate the ANC’s dangerous habit of taking its voters for granted.

She is linked by her name, and by her networks of associates, to the incumbent. She is a campaign manager’s worst nightmare.

(Indeed, her apparent campaign manager, Carl Niehaus, is himself a campaign manager’s worst nightmare.)

Activists may try to comfort themselves that there are two ways in which the ANC can survive the self-inflicted wound of a Dlamini-Zuma victory.

First, the winning faction could reach an accommodation with the “external faction”, known as the EFF.

If ANC support falls to 45%, this argument runs, the EFF’s 10% would be sufficient to retain control at national level and in the province of Gauteng.

The basic arithmetic cannot be faulted. However, the red berets will have to pretend to campaign strongly against the ANC if they are to attract voters. A semisecret coalition deal is a high-risk strategy that depends on an unstable combination of deeply cynical EFF leaders and utterly credulous EFF voters.

A second strategy would be to rig the 2019 elections. Institutional obstacles to accumulation, such as the Hawks, the National Prosecuting Authority and the South African Revenue Service, have been dismantled and swept aside in recent years.

Destabilising the ramshackle “independent” Electoral Commission of SA would be a relatively simple challenge for our expert masters of institutional destruction.

The key, however, is that it will be impossible to rig elections without citizens knowing they are being rigged.

While one recent Afrobarometer survey suggested that six out of 10 citizens might hypothetically forgo elections if an unelected government could guarantee housing, jobs and the rule of law, substantial majorities nonetheless continue to reject autocracy, military rule and one-party rule.

While dubious internal elections are largely seen as a party’s own business, there will be very little popular tolerance for a political party that is caught rigging national and provincial elections.

Most ANC delegates will probably decide that advance coalition deals and crude election-fixing efforts are high-risk gambles that are best avoided altogether.

They may feel that a simpler path to follow is to choose an electable leader at the conference in December.

• Butler teaches public policy at the University of Cape Town.