Kenya's political flaws highlight South Africa's strengths. The post-election political tragedy there both reflects some crucial limitations of Kenya's own governing order and offers new insight into South Africa's achievement since the first post-apartheid election in 1994 - as exemplified, in particular, in Jacob Zuma's election to the presidency of the African National Congress, thus making him the party's likely candidate for the state presidency in 2009.

Roger Southall is honorary research professor in the sociology of work unit, University of the Witwatersrand. Among his many books is (as co-editor) State of the Nation: South Africa 2007 (HSRC Press, 2007). He is editor of the Journal of Contemporary African Studies and contributes to the Review of African Political Economy

Much media analysis portrays the debacle in Kenya as a combustion of ancient ethnic enmities, and the 27 December 2007 elections as an ethnic census: in this imagining, Kenya is but a few steps away from a genocidal hell. It is true that ethnicity does play a major role in Kenyan politics (see Gérard Prunier, "Kenya: roots of crisis", 7 January 2008). However, we need to understand how the Kenyan post-colonial settlement has structured ethnicity and allows for its manipulation by actual and aspirant political power-holders.

Patterns of ethnic settlement, removal and migration (notably Kikuyu predominance within the richer central regions) play an important part in structuring differential ethnic access to land, wealth and opportunity. Yet they are far from determinate, not least because the multiplicity of ethnic groups in Kenya (which number over forty) demands that political leaders can only rule by constructing inter-ethnic coalitions.

These have changed significantly over the years. Crudely put, if the formerly ruling party, the Kenya African National Union (Kanu) primarily represented Kikuyu interests under Jomo Kenyatta and Kalenjin interests under Daniel arap Moi, then Mwai Kibaki's Party of National Unity (PNU) has attempted an uneasy alliance of both Kikuyu and Kalenjin (which is now in the process of unravelling). Likewise, the Orange Democratic Movement (ODM) of Raila Odinga represents a coalition of Luo and other outgroups seeking to dislodge the dominant groupings from power.

Also in openDemocracy on Kenya's crisis:

Peter Kimani, "A past of power more than tribe in Kenya's turmoil" (2 January 2008)

Michael Holman, "Kenya: chaos and responsibility" (3 January 2008)

Gérard Prunier, "Kenya: roots of crisis" (7 January 2008)

However, Kenyan politics revolves around much more than these simplicities.

After independence in 1963, the Kenyan economy remained dominated by settler and multinational interests. A decolonising land settlement saw settler land transferred to a Kenyan landholding elite, within which the Kikuyu - because they inhabited the most prosperous areas - became predominant. Yet the financial and commercial spheres of the economy were still under the control of whites, Indians and foreign companies. Inevitably, therefore, the rising African political class resorted to use of the state machinery to promote its entry into business: legitimately through demands for Africanisation, illegitimately through massive and systematic corruption. Politics thus became the competition of ethnic elites, as inter- and intra-party coalitions of elites sought to gain control of the state.

This dangerous game was facilitated by other features of the post-colonial settlement. Most notably, the British-style first-past-the-post (FPTP) electoral system provided for single-member parliamentary constituencies, which by emphasising the geography of ethnicity encouraged politicians to run as ethnic patrons of their constituents.

Moreover, the British-style prime ministership bequeathed Kenya at independence - unrestrained by the conventions of Westminster democracy - provided the foundations for a highly centralised executive presidency when Kenya became a republic. This tendency was then much enhanced by a constitutional amendment whereby the president became directly elected by the people, a move which quite deliberately strengthened the executive. From there it was a relatively short step to Kibaki rigging the presidential election of December 2007 and claiming state power even while the opposition won a convincing majority in the parliamentary election.

Polokwane's message

Also on South African politics in openDemocracy:

Gillian Slovo, "Making history: South Africa's Truth and Reconciliation Commission" ( 5 December 2002)

John Matshikiza, "Johannesburg: shanty city, instant city" (13 December 2002)

Paul Kingsnorth, "Apartheid: the sequel" (20 May 2003)

Nahla Valji, "South Africa: no justice without reparation" (2 July 2003)

David Mikhail, "South Africa and Iraq: the missing example" (15 December 2005)

Achille Mbembe, "South Africa's second coming: the Nongqawuse syndrome" (15 June 2006)

Achille Mbembe, "Whiteness without apartheid: the limits of racial freedom" (4 July 2007)

A very different form of civilian coup took place days before the Kenyan elections at the University of Limpopo in Polokwane, where the African National Congress (ANC) held its fifty-second conference on 16-20 December 2007. The main item on the agenda was the party's choice of leader (and probable presidential candidate after the 2009 election) - a contest dominated by the incumbent Thabo Mbeki and his chief rival Jacob Zuma.

In the event, Zuma won the vote of the conference delegates by the decisive margin of 2,329 to 1,505. Not so long ago, the possibility of such a "Zuma tsunami" was regarded by a wide array of commentators as heralding the end of South African civilisation as we know it. How things have changed! While the triumphalism of the Congress of South African Trade Unions (Cosatu) and the South African Communist Party (SACP) in the aftermath of Zuma's victory is to be expected, there has been a remarkably calm acceptance of the result elsewhere even by many who regard him as personally unsuited to occupy the highest office in the land. The reasons are fivefold.

First and foremost, the ANC has effectively shaken off the undesirable habits of a liberation movement, notably the claim of the victorious political elite to rule. From this perspective, the tsunami represents a reassertion of popular power over the established party leadership. The unambiguous message has gone out: South Africa is not Zimbabwe.

Second, the foolish attempt by President Mbeki to run as party leader for a third term has been contemptuously overturned by the party membership. Term limits on leadership have by implication been accepted, while the personalised Big Man politics of Africa have been rejected.

Third, the party has given notice that it requires its leadership to implement policies which serve the interests of the majority of the South African people. This leaves much ambiguity about the relation between state and party, yet suggests that that relationship will be one of healthy contestation.

Fourth, the ANC's national executive committee (NEC) will effectively instruct the ANC majority in parliament who to elect as state president after the next election - and in its first meeting since Polokwane, on 7 January 2008, the NEC confirmed that its candidate would indeed be Jacob Zuma. Nonetheless, the Zuma tsunami represents an assertion of parliamentary power over the executive - as Mbeki could find to his cost if in his remaining months in office he seeks to assert his state office as a wholly independent source of power.

Fifth, although many query the wisdom of the ANC having elected as its leader a man who now seems destined to appear in court on charges of corruption, this may have the consequence of consolidating constitutional democracy. Notwithstanding popular protestations that Zuma's prosecution is politically motivated, his appearance in the dock will enshrine the principle that the highest in the land are subject to the rule of law.

A politics of constraint

Not everything in the garden is rosy. In particular, there is justified concern that the prosecution of Zuma could allow various of his followers to ignite Zulu ethnicity. Yet the prospects of politicised ethnicity being contained are much better in South African than in Kenya. This is because of the nature of both the Zuma tsunami and the South African compromise. Three aspects stand out.

The first is that having achieved their party coup with Zuma at their head, it is very possible that his supporters will drain away if revelations about his alleged corruption in court depict him as a hopelessly compromised individual. The point is that the support for Zuma represents a popular programme, extending beyond the individual. It would seem that if events turned against Zuma, a majority of his supporters would quite happily line up behind the ANC's deputy president, Kgalema Motlanthe.

The second is that unlike Kenya, South Africa has abandoned first-past-the-post for proportional representation (PR) for the election of parliament. This has the virtue of providing for ethnic representation while discouraging ethnic mobilisation, for under a proportional system parties with aspirations to do more than represent a minority are compelled to eschew appeals to ethnicity. Certainly, PR leads to individual MPs being accountable to the party rather than to the voters, yet there is reasonable prospect that in time South Africa will move to a mixed electoral system with multi-member constituencies, as recommended in 2002 by the Frederick van zyl Slabbert-led electoral reform commission.

The third aspect is that South Africa does not have a directly elective presidency. Yes, this does mean that the president is effectively chosen by the ANC's national executive committee. Yet the NEC can only have its way through the instrumentality of the ANC's majority in parliament, and there can be no guarantee that in the future ANC parliamentarians will not flex their muscles.

In contrast, direct election of the president by popular vote - as in Kenya - would strengthen the writ of the executive at the expense of the legislature, with competing candidates for the presidency being selected by their party machineries anyway. What is needed in South Africa, rather, is for the ANC to open up its internal election procedures to wider political participation and public scrutiny. In retrospect, it may well turn out that this is very much what the Zuma tsunami has done, for it seems unlikely that any future leader of the ANC will be able to take his or her support for granted.

All this suggests that after its present convulsions have passed, Kenya - where successive attempts to implement wide- ranging constitutional reform have been stalled by presidential recalcitrance - has much to learn from South Africa: not least by looking to the virtues of proportional voting and of a presidency which is ultimately dependent upon parliament.