Louis Picard presents a dispiriting picture. The South African state is, he suggests, a mainly colonial inheritance, alienating, hierarchical, centralised and patrimonial. In 1994, the new administration inherited authoritarian local level administration, widespread corruption and an institutionalised system of patronage. During Apartheid the state chiefly functioned to serve sectional interests, its officials becoming increasingly venal. These characteristics of the state system have been preserved. Why?

The modern South African state, Picard notes, is an imperial creation, constructed in the aftermath of the Anglo-Boer War. Between 1924 and 1994, a bloated bureaucracy evolved, filled with inept Afrikaner beneficiaries of racial patronage. Increasingly this became a bureaucracy acting mainly to promote its own interests. From the 1960s onwards, affected by national shortages of skilled manpower, it became inefficient and demoralised. Subsequent attempts at bureaucratic reform failed, mainly because of political resistance from conservative civil servants. In fact, the state expanded in the 1980s despite the contraction of the economy. In 1994, the ANC inherited a weak state.

The ANC had no detailed plans for public service transformation. The compromises that resulted from a negotiated accession to power included guarantees to civil servants that they would not lose their jobs or pensions. The top layer of ANC leaders were non-racialists, but racially assertive ‘Africanism’ prevailed among the movement's rank and file, and ANC cadres themselves expected senior jobs. A generous and very expensive voluntary severance package and the resulting mass resignations created the space for affirmative action. Leaders now had a policy choice: to quickly replace white managers with barely trained cadres and incompetent officials from the homelands, or replace whites more slowly while focusing on human resource development and capacity building and social changes. ANC chose the former option because affirmative action became increasingly driven by a preoccupation with expanding the African middle class. Meanwhile trade unions opposed any structural reorganisation of the public service. In effect the shortcomings of the apartheid bureaucracy were maintained. Picard argues that representation rather than human resource development, and political control rather than control of corruption, were the main concerns during the transition. Political expediency made it almost impossible to overcome the challenge of effective rationalisation. Throughout Mandela's administration, policy emphasised racial transition rather than organisational reform. The government relied on existing in-house training schemes that facilitated the adoption of reactionary core values of old public service and failed to exploit new public management initiatives. Race, not class or region, became the basis for hiring. The delivery of services declined, partly because of the exodus of experienced people. While the quality of the bureaucracy deteriorated, trade unions were successful in ensuring rising salaries for their members. Corruption increased, as did the scope of patronage.

These developments were partly a consequence of the absorption of homeland governments into the new provincial administrations. During late apartheid, homeland regimes resembled prestructural adjustment regimes in other African states: costly, parasitic and extremely inefficient. The construction of new provincial governments around old homeland bureaucracies simply expanded the range of patrimonial officialdom. After 1994 successive Mbeki administrations instituted only weak efforts to correct or reverse these trends despite initial commitments to reform. Today, ‘distorted institutional relationships’ are widespread within the South African bureaucracy as many civil servants ‘identify only with their own immediate economic and social interests’. Since 1994, patronage and corruption have curtailed the ‘development role that South Africa played not only at home but in the rest of the continent’. The South African state may well follow other African administrations in becoming a burden to society, impeding economic and social development.

This analysis is only partly persuasive. It is true that institutionalised segregation was entrenched by Lord Milner's imperial government, but was the South African government such a typical colonial state? Enfranchised white citizens ensured the construction of a welfare regime which to a degree included black South Africans within its provisions from the 1940s and which promoted industrialisation. Contrary to Picard's suggestions, parastatal corporations were often effective and creative: the munitions and petrochemical industries are cases in point. Discriminatory employment practices that favoured Afrikaners were usually not exercised through individualised patronage, and given the limits on state expansion since 1994, the scope for nepotism has remained restricted. Has public service delivery really declined? Picard supplies no hard evidence to this effect, and official statistics and academic surveys indicate that the state's delivery of benefits to its citizens has expanded considerably, mainly in favour of the rural poor. Picard himself admits there are exceptions to his generalisations. He concedes the Gauteng government's achievements in policy innovation and strategic management, partly a consequence of its use of new public service training facilities at nearby universities. Other provinces have enjoyed some success in reforming their social welfare departments, and public education has also improved. In these areas much of the corruption that characterised the late apartheid period has been checked. These developments are at odds with the patrimonial trajectory that Picard believes is so inevitable.