A surprising number of political pundits have questioned whether a breakaway from the African National Congress would be good for ‘democracy' in South Africa. The basic complaints are that the new movement lacks principle, is motivated purely by personal grievance, and is compromised by its association with the abuses of the Mbeki era.

In Business Day last week the UCT academic, Anthony Butler, put forward a more substantive argument for the dangers for democracy of a possible split. The questions he raises really concern whether ANC dominance has been good or bad for democracy in South Africa.

Butler's argued that the ruling party's dominance "has allowed the liberation movement to secure political stability, to defuse ethnic and racial tensions, and to offer a coherent and prudent programme of government. A strong ANC has also filled the vacuum left by weak liberal and state institutions to resolve otherwise intractable conflicts of interest and opinion in a divided society."

It would be better in the circumstances, Butler suggested, for a split to come later rather than sooner - after some two decades of ANC rule. Liberal institutions would, by this time, be "robust enough to copy with the immense internal strains that a more fluid political system would generate."

The danger of a split now would be that a breakaway would become a "quasi ethnic party" and that our institutions are not yet strong or legitimate enough to manage the tensions that are going to result. "In the absence of coherent ANC intervention, evils like ethnic entrepreneurship, provincial corruption and racial antagonism might spread unchecked. Democratic elections, moreover, are unlikely to lead to prudent economic policy choices."

There are a number of problems with this argument. For one thing Mbeki spent much effort stoking racial antagonism rather than trying to dampen it down. It is also not clear why Butler regards greater ethnic division as a problem but not the current racial divide in our politics. There was, initially, no institutional "vacuum." Indeed, in the mid-1990s many academics were in fact arguing that the worst effects of the ANC's looming electoral dominance would be mitigated by South Africa's strong liberal, state, and civil society institutions.

That many of our institutions are so weak today is because of determined efforts by the ANC, from around 1997 onwards, to bring all "levers of power" under party control. Independent institutions were softened up for takeover through racialist propaganda, and claims that they contained some kind of ‘counter-revolutionary' element. And then with this done party loyalists, such as Jackie Selebi, were deployed to senior positions - and tasked with implementing ‘African hegemony' whatever the cost.

The failure of much of civil society to oppose, at the time, the distinctly Soviet policy of cadre deployment can largely be ascribed to the immense moral authority the ANC then enjoyed due its ‘overwhelming' majority. As Maurice Duverger noted, if a party's electoral majority is great, its authority is great, "it is not embarrassed by the opposition; it can claim to represent the will of the country."

Equally, the ANC's determination to erase the boundaries between party and state cannot be separated from the sense of impunity generated by its dominant political position.

In democracies where there is a regular alternation in government, the ruling party could well end up in opposition after the next election. As such it has a pressing incentive to respect the limitations on its power and the autonomy of state institutions, and to refrain from abusing its office. This is because it knows that should the opposition come to power at the next election - as is entirely feasible - a new government could expose and punish any abuse of office and turn any accumulated powers (or precedents it may have set) against them.

James Madison observed that this ever present possibility of being (democratically) removed from office creates an "uncertainty of condition" which leads the stronger parties to accept "a government which will protect all parties, the weaker as well as the more powerful".  However, since the ANC in the mid-1990s thought it was going to be in power in perpetuity none of this applied. This problem was further compounded by the belief that only it knew the one truth path for society, and so should be allowed to rule "unfettered by constraints" (see article).

If the trajectory of the early years of Mbeki's rule had continued - uninterrupted by the Zuma rebellion - there would be very little left of South Africa's liberal or state institutions. The 2005 effort to bring the judiciary under greater direct control were opposed, and successfully resisted, precisely because of a waning of the moral and political authority of Mbeki and his ANC. Post Polokwane there seems to have been a widespread awakening to the extent to which the institutional rot had set in under Mbeki.

Rather than the benign scenario sketched out by Butler, one more likely result of two decades of monolithic ANC dominance would have been that of Zimbabwe - where the Zanu-PF party-state was able to thwart the MDC's electoral challenge through the plundering of private property and the manipulation of the levers of power. South Africa is perhaps fortunate that the early effects of centralisation - manifested in internal ANC dissension at the ‘abuse of state power' and conflicts over tenders and patronage - have counter-acted (for the moment) the probable longer term consequences.

It is certainly tempting to respond cynically to the breakaway ANC's claim to being defenders of the constitution, just as one could the National Party's belated embrace of liberal checks and balances in the early 1990s. But now that many of the losers at Polokwane are moving from government to opposition, they have an objective interest in trying to limit the power of the ruling party. One should not be too precious about this - this is how liberal democracy makes its converts.

This is not to say that a split, as Butler observes, could not have negative consequences for South Africa's political stability. The ANC's secure electoral dominance had a paradoxical effect on the ruling party, allowing it to be both more cautious and less restrained. While the lack of uncertainty allowed it to eviscerate the boundaries between party and state - and created the conditions for corruption to flourish - it also meant that it could pursue its programme gradually but purposefully. Since it faced no real electoral challenge there was no obvious need to biff the opposition.

For the first time since coming to power the ANC is facing a serious challenge to its own support base. To prevent violence the ANC leadership is not only going to have to refrain from killing talk of earlier this year (see article) but actively condemn any attacks by its supporters on opposition activists (see here). The split is certain a test of the ruling party's political tolerance and of South Africa's ability to sustain a more pluralistic democracy. But if the fears of Butler are realised this would be a reflection not of a weakening of democracy, but rather the exposure of pre-existing problems.

Click here to sign up to receive the Politicsweb headlines in your morning inbox