Perhaps the most spectacular is president Thabo Mbeki?s fight with his Alliance partners in the Congress of South African Trade Unions (Cosatu) and SA Communist Party, which I?ll devote a future column to once the dust settles. The unions and communists backed an apparently corrupt ? at minimum sleazy ? deputy president, Jacob Zuma, ostensibly because he was the only ?friend of the workers? to be found at the highest levels of government.

Last month Zuma was fired by Mbeki, who replaced him with the pro-business mining minister Phumzile Mlambo-Nguka (whose husband was, as the state?s chief prosecutor, once Zuma?s main tormentor). Not only did the Alliance began to splinter, but Cosatu itself last week divided itself among three camps: the mainstream backing Zuma, the right backing Mbeki, and the left sickened by both.

Reflecting last week?s militant mood, Cosatu?s central committee also complained of ruling-party ?neoliberals (who) may use the state apparatus to turn the revolution in(to) an offensive against the true cadres of the revolution?.

Quickly denying the charges, Mbeki revealed how the charge of ?neoliberal? strikes fear into politicians? hearts here.

The key problem was the adoption by Mandela and Mbeki of a World Bank-inspired, market-oriented, ?willing-seller/willing-buyer? programme that limited the state function to providing a tiny once-off subsidy ($2,300 at today?s exchange rate), far too small to acquire a decent plot of land.

He now takes the argument further by focusing on South Africa?s liberal Bill of Rights. As recently as the early 1990s, the ANC opposed the post-apartheid recognition of colonial- and apartheid-era land grabs. However, the subsequent Bill of Rights in the Interim Constitution reflected the old apartheid National Party?s (NP?s) tough negotiating position. ANC representatives merely sought the power to redistribute, according to Ntsebeza, but ?without an excessive obligation to compensate owners?.

How is compensation determined? The Constitution refers only to a ?just and equitable dispensation?. As Ntsebeza notes, however, ?Land owners are inclined to inflate prices.? The Land Claims Court worked out a formula that emphasised the market value of the property, adjusted by subtracting an estimate of the present value of past state subsidies.

Ntsebeza poses this challenge: ?What alternatives did the ANC have when it came to power? We have to ask this question. Closely linked is the relative weakness of the land-based organisations, whether community-based or NGOs. Their hope was to influence the Department of Land Affairs. How this was to be achieved within the context of the neoliberal policy the ANC had adopted was not clearly articulated.? Hence when a group that came to be known as ?land reform liberals? moved from the National Land Committee to the state in the mid-1990s, they mainly lubricated the ANC?s neoliberal strategy.

Why 30%? The land market typically turned over 6% of land each year, so a five-year target of 30% would have been reasonable – in the event the state withdrew residual apartheid-era subsidies (ranging from irrigation to energy to cheap credit) from white commercial farmers and, in the interests of affirmative action, redirected those to black emerging farmers.

(Zimbabwe?s two-decade failure of the willing seller/willing buyer model, along with Robert Mugabe?s need to intimidate his citizenry after losing the February 2000 constitutional referendum, helps explain the destructive rural chaos misleadingly termed ?fast-track land reform?.)

Moreover, following South Africa?s ? and indeed the world?s – 1990-93 real estate market crash, the stage of the price cycle that prevailed in 1994 was favourable. Today, land prices in prime areas of South Africa mirror the global bubbling of speculative real estate markets. This makes redress using market mechanisms so expensive that even state officials ? including Mbeki – are looking skeptically at the hedonistic white-dominated golf estates and game farms that now pock once-productive agricultural soil.

By decommodifying land, specifically by removing it from the market once in state hands, the present trends that make life hard for small farmers ? historically high interest rates, ever-lower commodity prices, excessive agricultural mechanisation and export-orientation, and sophisticated cropping systems featuring genetic modification ? could have been more effectively contested.

Ntsebeza agrees, South Africa needs much stronger MST-style advocacy. ?The establishment of the Landless People?s Movement in 2001 brought to the fore all sorts of tensions within the National Land Committee, which led to the demise of this network.? As a result, today, ?there is very little effective pressure that comes from below.?

Achieving a decommodified model of thorough-going land reform requires more state resources, all these civil society networks argue. Hanekom had announced in mid-1996 that the bargain-basement price of $2.3 billion in state subsidies would pay for land reform over the subsequent five years. It was not clear how the figure was arrived at, because his staff reckoned that 1.7 million families required land (itself a conservative estimate). Given the standard grant of $2,300 (a figure chosen because it was equivalent to the housing subsidy) and correcting for inflation, fewer than a million families would be served.

In view of the weakness of civil society and the strength of patronage-based traditional leaders, Ntsebeza advocates a major rural power shift. Back home in the Eastern Cape community of Cala, Ntsebeza and his brother Dumisa (a judge who is one of South Africa?s premier public intellectuals) ran a famous Marxist bookshop for many years. Indeed, Ntsebeza has authored a brand new book about that community: Democracy Compromised: Chiefs and the Politics of the Land in South Africa (Leiden, Brill).

The ?cause? of land hunger is the combination of neoliberal fiscal austerity, the World Bank strategy, and the elite class pacting that characterise so many other developmental failures of post-apartheid South Africa.

But likewise, continue with the willing seller/willing buyer model, and the conditions for a Zimbabwe-style political breakdown only improve.