. 9
( 12)


labourers. Most were migrants, not because mineowners wished it “ they
thought migrant workers expensive and inef¬cient “ but because Africans
refused to exchange rural land rights for a lifetime amid the danger, disease,
and brutal conditions of deep-level mining “ ˜hell mechanized™, as a mission-
ary described it. Initially, therefore, mineowners had to pay wages suf¬cient
to attract Africans temporarily from their homes, but this changed as Africans
lost their independence. In 1895“7, especially, the Portuguese conquered the
Gaza kingdom in southern Mozambique and imposed taxation and compul-
sory labour, which were quickly followed by cattle plague and famine. By 1896“8
the goldmines drew three-¬fths of their ¬fty-four thousand African workers
from southern Mozambique, which supplied the largest single contingent of
mineworkers almost continuously until the 1970s. Many others came from the
Transkei and from Lesotho, 20 percent of whose able-bodied men were work-
ing in South Africa at any moment in 1911 and 47 percent in 1936. Instead of
working abroad once in early youth, men came to spend their lives oscillating
between homes and workplaces. Rural economies came to depend on their
remittances. Rural families adapted to survive absent fathers, often replacing
the patriarchal and polygynous homestead by a three-generation household
in which a wife lived with her parents and children until her husband retired,
perhaps bringing home tuberculosis, which by 1930 infected a large majority
of the Transkei™s adults.
For African cultivators, gold mining initially expanded the pro¬table urban
market already opened at Kimberley. Maize production increased among Zulu,
Sotho, and especially the peoples of the Orange Free State, the South African
Republic, and Natal, where African peasant farmers with ox-ploughs expe-
rienced a prosperity that their children remembered as a golden age, either
farming the minority of land remaining to them or cultivating as sharecrop-
pers on white farms. From the 1890s, however, white entrepreneurs competing
for urban markets and African labour sought to transform sharecroppers ¬rst
into labour-tenants and then into landless labourers. The Natives Land Act
Industrialisation and race in South Africa 275

14. Industrialisation and race in South Africa.

of 1913 had this objective, for it prohibited land transfers between races, ¬xed
the African share of South African land at 7 (later 14) percent, and restricted
the number of sharecroppers and cash-tenants who could reside on a white
farm outside the Cape Province. But legislation alone could not change the
countryside. Between the wars there was violent agrarian con¬‚ict as farmers
imposed more severe terms on resident Africans, who replied by burning crops,
slaughtering stock, and hearkening to the millennial promises of religious and
political prophets. As late as 1954, some 20 percent of ˜white™ farms had no
white resident, but by then the mechanisation of agriculture was ¬nally driving
sharecroppers and tenants from the land, as elsewhere in Africa. Meanwhile
the growing population on the limited and overexploited African reserves was
impoverished. Even in the 1920s, the reserves produced only half their food
needs and the proportion fell steadily thereafter. Only tiny privileged elites
clung to the freehold property needed to ¬nance education and professional
The commercialisation of agriculture in response to mining and urban
growth also transformed the rural white population. Their farms became
276 africans: the history of a continent

smaller and more numerous, but the bijwoners (squatters) who had grazed
their scrawny beasts on the fringes of nineteenth-century estates were driven
from the land to join the more than 300,000 white people (about one-sixth
of the entire white population) thought in 1930 to be ˜living in great poverty™,
often in the slums of industrial cities. Given South Africa™s hard and ancient
rocks, poor soils, and recurrent droughts, it cost the state £112 million in sub-
sidies to European agriculture between 1911 and 1936 to keep white men on
the land and to win their votes, chie¬‚y through a state-supported marketing
system, elaborate extension services, and transport geared to farmers™ interests.
Unlike white settlers elsewhere, most South African farmers did not become
producers of specialised export crops like wine or coffee. Despite low yields,
maize was 39 percent of their output by value in 1919 and 32 percent in 1976.
This white monopoly of the food market deprived Africans of bargaining
power to push their wages above bare subsistence. The number of African
and Coloured farm labourers rose gradually to a peak of about 1,500,000 dur-
ing the 1960s. Most were poorer than either African townsmen or the reserve
population, earning an average wage from all sources of £20 a year in the late
Although goldmining was vital to South African industrialisation, it did
not automatically cause it, for eighty years of copper mining did not industri-
alise the Congo, nor was South African industrialisation a sudden process. In
1891 the Cape Colony™s manufacturing output was already more valuable than
its diamond production. In the South African Republic, however, goldmin-
ing stimulated railway building, urbanisation, coal mining, and the coal-¬red
electricity that became the chief source of industrial power. By 1914 the Witwa-
tersrand possessed the world™s largest electrical power station, employing the
latest German technology. Manufacturing output expanded during the First
World War and almost doubled during the 1920s. One reason for industrial
growth, in contrast to tropical Africa, was that political independence allowed
white South Africans to be economic nationalists. General Smuts, as Prime
Minister from 1919 to 1924, made industrialisation a target of state policy. The
Afrikaner Nationalist government that replaced him in 1924 raised protective
tariffs and invested mining revenue in industry, notably the state-owned Iron
and Steel Corporation, which began production in 1934 and spearheaded the
transition to heavy industry that countries further north later found so dif¬cult.
From 1933, when South Africa abandoned the gold standard, greatly increased
gold prices stimulated even faster growth and enabled the economy to escape
the foreign-debt trap that was to check industrialisation in Algeria and else-
where during the 1980s. Between 1911 and 1945 the proportion of South Africa™s
foreign debt to total public debt fell from 91 to 3 percent.1 Another crucial
breakthrough took place during the Second World War, when manufacturing
employment rose by 60 percent and the engineering industry shifted from craft
Industrialisation and race in South Africa 277

production to mass manufacture, initially of war materials and subsequently
of consumer durables. The share of metal products and machinery in man-
ufacturing output rose between 1936 and 1951 from 4 to 19 percent, making
South Africa decisively a manufacturing country, with many characteristic fea-
tures of a late-industrialising economy: large enterprises, a major state sector,
heavy reliance on primary exports (of gold), and severe repression of labour.
Cheap labour, cheap energy, gold, government, and gradualism were distinctive
features of South African industrialisation.
Its most dramatic consequence was rapid urbanisation. In 1891 Cape Town,
with 51,000 inhabitants, was South Africa™s largest city, but by 1896 Johannes-
burg, only ten years old, already contained 100,000 people, half white and half
black, in an urban anarchy described as ˜a Monte Carlo superimposed upon a
Sodom and Gomorrah™.2 The country™s total urban population, some 1,225,000
in 1904, rose to 3,218,000 in 1936, including 68 percent of all white people and
19 percent of Africans. Municipal authorities tried to control urbanisation by
segregating Africans into locations, which became national policy under the
Natives (Urban Areas) Act of 1923. To impose this model on a swollen mining
town like Johannesburg, however, was beyond municipal capacity. When the
city centre™s notorious multiracial slums were demolished during the 1930s, for
example, their African inhabitants moved not to the distant, expensive, and
strictly controlled locations at Orlando (the nucleus of modern Soweto) but to
freehold land at Sophiatown and Alexandra on the edges of the white city. In
Cape Town, similarly, some 37 percent of the residential area was still racially
mixed in 1936, notably the largely Coloured working-class area called District
Six, close to the city centre.
Until the 1920s, the main threat to mineowners and the state came not from
Africans but from European workers. Initially most white miners were immi-
grant bachelors seeking quick earnings before tuberculosis killed them. They
vigorously defended their jobs and racial wage differentials against employers
anxious to replace them by equally competent but cheaper Africans. The white
miners™ tactics were militant unionism and racialism. In 1893 the ¬rst all-white
mineworkers™ union imposed a monopoly of blasting against the employers™
resistance. Twenty years later a strike obliged the employers and the state to
recognise the union. From 1911 to 1925, the Labour Party largely controlled
the Johannesburg City Council, whose employees brie¬‚y established a soviet
in the City Hall during 1918. A Communist Party came into being in 1921. A
year later, when the mineowners tried to break the union and reduce the ratio
of white workers, ˜Strike Commandos™ converted a strike into the Rand Revolt,
which brie¬‚y seized power in several mining towns until suppressed by the
army at a cost of between 150 and 220 lives. At this point, however, the state
used its victory to domesticate both capital and labour. The Mines and Works
Amendment Act of 1926 ¬xed the ratio of white to black workers, enabling
278 africans: the history of a continent

mineowners to mechanise the industry and miners to become the best-paid of
white workers.
For black workers, by contrast, early militancy brought little reward. Spo-
radic African dock strikes in Port Elizabeth and Cape Town can be traced
back to the mid-nineteenth century, but the ¬rst major African mine strike on
the Witwatersrand in 1913 was broken by troops with ¬xed bayonets. Rapid
industrialisation, urban growth, and in¬‚ation during the First World War then
radicalised white-collar workers as well as manual labourers, breeding sev-
eral unsuccessful strikes in 1917“20 and the ¬rst major African trade union,
the Industrial and Commercial Workers Union (ICU). This emerged in the
Cape Town docks during 1918, was led by a migrant clerk from Nyasaland
named Clements Kadalie, and expanded ¬rst into an urban general union and
then into near-millennial rural protest expressing the grievances of threatened
sharecroppers and labour-tenants on the highveld. At its peak in 1927, the ICU
claimed 100,000 members, but it then disintegrated in factionalism and disillu-
sionment. By 1933 there were only three African trade unions in South Africa, all
unrecognised by the state. During the next decade, Communist and Trotskyite
organisers gradually constructed a more substantial labour movement from the
shop-¬‚oor upwards. Wartime militancy culminated in 1946 in a major African
mine strike, but its violent suppression, with at least nine deaths and twelve
hundred injuries, demonstrated the continuing dominance of employers and
the state.

politics 1886“1948
Mining and industrialisation transformed South African politics. In 1899 the
British launched the costly Anglo-Boer War to protect their regional supremacy
against the South African Republic™s new wealth and power, but Afrikaner guer-
rillas surrendered only when the Peace of Vereeniging of May 1902 promised
that ˜the question of granting the Franchise to Natives will not be decided until
after the introduction of self-government™.3 British control of South Africa
depended on attracting enough English-speaking immigrants to the Transvaal
to outvote its Afrikaners. That meant restoring and expanding gold produc-
tion, which depended on recruiting suf¬cient nonwhite labour. Because not
enough Africans accepted work at the wages offered, some sixty thousand Chi-
nese contract labourers were imported. But this alienated English-speaking
white workers, who dashed imperial plans in 1907 by allying politically with
Afrikaner leaders.
These events were crucial to Afrikaner nationalism. Nineteenth-century
Afrikaners had been strongly aware of their difference from Britons and
Africans, but during the last thirty years of the century responsible government
at the Cape and the growth of more modern states in the northern republics had
Industrialisation and race in South Africa 279

encouraged sectional patriotisms. President Kruger had dismissed Hofmeyr,
the Cape Afrikaner leader, as ˜a traitor to the Africander cause™. The Anglo-Boer
War partly healed these divisions, for many Cape Afrikaners sympathised with
and sometimes aided the republics, but it also opened new con¬‚icts between
advocates of surrender and continued resistance. It was the uni¬cation of South
Africa and the creation of an electoral system that brought Afrikaners together
into political nationalism, as would happen later in West Africa. Uni¬cation
was pressed both by the British and by local white politicians, especially the for-
mer Afrikaner generals, Botha and Smuts, whose party, Het Volk, won the ¬rst
election in the Transvaal in 1907 with labour aid and thereby ensured Afrikaner
leadership of South Africa. The negotiations that led to independence under
the Act of Union in 1910 created a strong central government, entrenched the
legal equality of the English and Dutch (from 1925, Afrikaans) languages, and
left the franchise as it had been in each prewar province, so that Africans and
Coloured people effectively had the vote (on a quali¬ed franchise) only in the
The South Africa Party, led by Botha and Smuts, formed the Union™s ¬rst
government. It was dedicated to reconciliation between Afrikaners and British,
but this was undermined by the First World War, when South Africa™s participa-
tion on Britain™s side precipitated unsuccessful rebellion by Afrikaner extrem-
ists, and by disputes over the imperial relationship. In 1924 General Hertzog™s
National Party won power on the votes of most rural Afrikaners. Externally he
was satis¬ed with the dominion status recognised in 1926. Internally he pressed
forward the relief of poor whites and the segregation of Africans already embod-
ied in the Natives Land Act of 1913 and the Natives (Urban Areas) Act of 1923.
Almost all white South Africans favoured segregation, even missionaries and
liberals eager to protect Africans from deracination. For Hertzog, a key element
in segregation was to remove Cape African voters “ 10,628 of them in 1935“
from the common roll and give them separate representation and institutions.
To obtain the two-thirds majority necessary for this constitutional amend-
ment and to tackle the economic problems of the international depression,
Hertzog ˜fused™ his party with Smuts™s opposition in 1934. The new United
Party removed Africans from the common roll in 1936, but fusion alienated
Afrikaner extremists, who saw the new party as a capitalist coalition likely to
divide the Afrikaner nation along class lines. They broke away in 1934 under
D. F. Malan to form the Puri¬ed National Party. It became the chief exponent of
the ethnic separatism gaining support among Afrikaners during the later 1930s,
based on deliberate cultivation of the Afrikaans language, the v¨lkisch notions
of nationality fashionable in continental Europe, determination to win eco-
nomic equality with English-speakers, and historical symbolism popularised
by the Voortrekker Centenary of 1938. When South Africa™s entry into the
Second World War destroyed Hertzog™s government and left Smuts in power,
280 africans: the history of a continent

a bitter struggle for Afrikaner leadership ensued. Malan won. By 1945 he was
in a position to reunite the Volk.
For Africans and Coloured people, too, the Anglo“Boer War and Union
were key moments in political organisation. The Coloured people numbered
some 445,000 in 1904 and formed 9 percent of the population, mainly in the
Cape. Their ¬rst major association, confusingly named the African Political
Organisation (APO), was formed at the end of the war in 1902 by their small
professional elite led by Dr. Abdullah Abdurahman. It aimed to defend the
community™s distinctive identity and extend its rights, especially the right to
vote, into the newly conquered northern provinces. This aim con¬‚icted with
the Peace of Vereeniging. Instead the APO survived as the spokesman of the
Cape Coloured elite, whose aspiration to be accepted into white institutions
distanced them from the bulk of Coloured workers.
From the 1880s, the mission-educated African elite of the Cape Colony and
Natal “ clergymen, teachers, clerks, commercial farmers “ formed the ¬rst
small modern African political associations. The most articulate was the South
African Native Congress, founded in the Eastern Cape in 1898. After British
victory in 1902, these associations fostered similar bodies in the Transvaal and
Orange Free State and urged the extension of the Cape franchise to these
provinces. When the Act of Union denied this and the new white parliament
instead debated territorial segregation, elite leaders met at Bloemfontein in 1912
to form the South African Native National Congress (later the African National
Congress, ANC) ˜for the purpose of creating national unity and defending our
rights and privileges™. The ANC initially campaigned against the Land Act of
1913 by petitions and deputations. When this achieved nothing, its moderate
leaders were replaced in 1917 by more radical men from the Witwatersrand who
associated the organisation with postwar strikes and antipass protests. Alarmed,
the moderates regained leadership in 1920, lost it again in 1927 to the communist
Josiah Gumede, but ousted him once more in 1930. The following decade was
the least active in the ANC™s history. It did not effectively defend the Cape
African franchise in 1936. Its total funds four years later were ¬fteen shillings.4
The vitality of interwar African politics lay in two other directions. One was in
the countryside, where protest either took near-millennial forms (as in Zionist
churches or labour-tenants™ support for the ICU) or centred on resistance
to cattle dipping, soil conservation, and other of¬cial schemes to salvage the
overcrowded reserves. Political activity was also vigorous in the towns, where
a working-class lifestyle, often known as marabi culture, took shape around
the music, dance, sex, youth gangs, and illicit liquor of the shebeens in city-
centre slums and freehold townships like Sophiatown. Urban political action
often went no further than repelling police liquor raids, but it could embrace
anti-pass protests “ especially against attempts to make women carry passes “
and boycotts of municipal beerhalls, the most famous taking place in Durban
Industrialisation and race in South Africa 281

in 1929“30 and lasting eighteen months. Protests on this scale were organised
by grassroots politicians “ clerks or craftsmen threatened with unemployment,
taxi-drivers, shack-landlords, herbalists, Zionist preachers “ who drew ideas
and slogans from modern organisations like the ICU, ANC, and Communist
Party while also mobilising indigenous symbols and beliefs. After Clements
Kadalie had addressed a meeting in East London in 1930, for example, the next
speaker was ˜a kitchen girl, at the Strand Hotel . . . and a prophetess™ who
said she received a message from God that let all the natives listen to what
Kadalie tells them. God has revealed to her that Kadalie is the only leader who
is going to uplift Africa. She again has received a message from the Almighty
God that Kadalie should go to Gcalekaland in the Transkei and organise the
AmaXhosas at the Great Place of the Paramount Chief.5

This popular politics, often openly racialistic and tribalistic, was a world apart
from the staid multiracial resolutions of ANC conferences.
The Second World War did something to fuse the two political levels. On
the elite plane, in 1944 young men from the black University College at Fort
Hare, exasperated by the ˜gentlemen with clean hands™ who ran the ANC,
formed within it a Youth League as a ˜brains-trust and power-station™ to press
the state for full political equality. They were willing to associate Congress
with the popular protest mounting during the war as industrial growth bred
mass urbanisation that swamped housing and other facilities. In one series
of protests, residents of Alexandra boycotted buses and walked nine miles
each way to work rather than pay a fare increase of one penny. In another,
nearly 100,000 homeless people created illegal squatter camps on vacant plots
in and around Johannesburg. Urbanisation swamped segregation. ˜You might
as well try to sweep the ocean back with a broom™, Smuts complained in 1942.
Indian urbanisation, too, seemed to Durban™s white residents to ˜penetrate™
their suburbs, provoking shrill demands for restriction or repatriation. As the
war ended, all political tendencies agreed that South Africa needed a new racial
order and that only the central government could establish it.

the ascendancy of apartheid
In 1948 the largely white electorate faced a clear choice of racial policies. Malan™s
National Party offered apartheid, a newly coined word to describe a more rigid,
centrally enforced segregation, con¬ning each race to speci¬ed areas, allocat-
ing African labourers to farms or towns, but promising also to enable each
race to practise its own culture and manage its own affairs. Smuts™s United
Party, by contrast, claimed to defend South Africa™s traditional racial order,
where the state assisted communities to segregate themselves voluntarily but
saw African urbanisation as irreversible and gradual assimilation to Western
282 africans: the history of a continent

culture as desirable. Unexpectedly, the Nationalists won, with only 40 percent
of votes. Their ¬rst measures sought to attract further white support, ban-
ning mixed marriages, creating procedures for universal racial classi¬cation,
and setting up machinery for compulsory segregation under the Group Areas
Act of 1950. Subsequently, as support and con¬dence grew, their programme
expanded into ˜positive Apartheid™, including a separate Bantu education system
and ˜self-governing™ but dependent rural Homelands for the various African
˜tribes™. Whereas interwar governments had enacted segregatory legislation,
the Nationalist regime implemented it. Power, not policy, was the chief novelty
of apartheid. The power came from the growing wealth and administrative
capacity of the industrial state, the faith in state intervention and social engi-
neering common throughout the postwar world, and the racialism that enabled
Nationalists to justify their ruthlessness towards black people.
Until the mid-1970s, apartheid was remarkably successful. Its main achieve-
ment was the segregation of cities by moving their black inhabitants to suburban
townships isolated by ˜machine-gun belts™, a strategy made possible by elec-
tric trains and motor transport. In Johannesburg, Sophiatown was destroyed
between 1955 and 1963 and Africans were relocated into the 113,000 concrete
houses of Soweto, divided into tribal sections. The estimated 120,000 Africans in
Durban™s main freehold settlement, Cato Manor, were rehoused in two town-
ships in the neighbouring KwaZulu homeland. District Six in Cape Town was
razed during the 1970s; its Coloured inhabitants were resettled in an outlying
concrete wilderness where an incomplete survey in 1982 counted 280 juvenile
street gangs. Legislation decreed that only Africans born in a town or work-
ing there continuously for ¬fteen years (or ten years for one employer) had
permanent residential rights. The African urbanisation rate slowed from the
early 1950s, although the statistics probably underestimated it. Action under
the Group Areas Act also relocated 305,739 Coloured people, 153,230 Asians, and
5,898 whites by March 1976.
African resistance to this assault brought ANC elite politics and urban pop-
ular action closer together. In 1949 Youth League members gained leadership
of Congress. Three years later, in alliance with radical Indian politicians, they
launched the De¬ance Campaign of nonviolent resistance to unjust laws. The
most widespread protest the country had seen, it expanded ANC membership
to a claimed 100,000, or 1 percent of the African population, with the main
support in the industrial towns of the Witwatersrand and the Eastern Cape.
But the state broke the campaign by legislation punishing deliberate breach of
the law by whipping. As the 1950s proceeded, it became clear that mass non-
violent nationalism, so successful in India and tropical Africa, might heighten
political consciousness “ nobly expressed in the Freedom Charter of 1955 “
but scarcely threatened a regime prepared to shoot demonstrators. As poli-
tics grew more dangerous and African frustration increased, younger radicals
Industrialisation and race in South Africa 283

broke away from the ANC in 1958 and formed the Pan-African Congress (PAC),
rejecting alliance with non-African organisations. In March 1960, its anti-pass
campaign provoked the police to ¬re on a meeting at Sharpeville, killing sixty-
nine unarmed people. The government then banned both the ANC and the
PAC. Both turned to sabotage, but this too was crushed in 1964. Organised
African politics gave way for a decade to factional con¬‚icts in the Homelands.
Apartheid had again succeeded.
The economy also prospered. Between 1946 and 1973, real GDP increased
steadily at between 4 and 6 percent a year, not a rapid growth rate for a
middle-income developing country but substantial and sustained. Between
1950 and 1980, the volume of manufacturing output multiplied over six times.
Engineering and the metal industries became the largest manufacturing sec-
tor, supported by new technological industries like chemicals and plastics.
Rich new gold¬elds in the Orange Free State revitalised goldmining, while the
freeing of gold prices in the early 1970s raised them tenfold during the next
decade. Mechanisation transformed white farming, which no longer needed
the labour apartheid was designed to supply. Instead, an estimated 1,129,000
Africans were compulsorily moved from white farms between 1960 and 1983.
Most were relocated in the Homelands, along with perhaps another two mil-
lion people removed from towns and other ˜inappropriate™ areas. Many were
dumped in ˜closer settlements™ on the open veld with urban densities but no
urban industries or services “ areas like Qwaqwa on the Lesotho border, whose
population increased during the 1970s from 24,000 to 300,000.6 Between 1960
and 1985, the Homelands™ inhabitants increased from 39 to 59 percent of an
African population which itself doubled, giving South Africa an extreme form
of that uneven distribution of rural population that characterised the whole
continent. Agriculture collapsed in the Homelands, which relied instead on
remittances from migrant labourers and subsidies from Pretoria.
The ruthlessness of resettlement policy was one aspect of the regime™s gen-
eral brutalisation, especially during Vorster™s premiership from 1968 to 1978,
when fear of invasion led to the making of nuclear weapons and bred a secu-
rity apparatus using torture on a scale comparable to tropical Africa™s worst
tyrannies. The white electorate endorsed this. Whereas in 1948 the National
Party had won only 40 percent of votes, in 1977, at its peak, it won 65 percent,
including not only 85 percent of Afrikaners who voted but probably 33 per-
cent of English-speakers. The white race was consolidating under Afrikaner
leadership. For Nationalists, this was apartheid™s chief triumph.

the destruction of apartheid
Within little more than a decade of the electoral victory in 1977, apartheid lay in
ruins. Among the many forces destroying it, the most fundamental were those
284 africans: the history of a continent

also destabilising colonial and postcolonial regimes further north. At their root
was population growth. South Africa™s population trebled under apartheid,
from 12,671,000 in 1951 to 37,945,000 in 1991. In 1951, 21 percent of these people
were white, 12 percent Coloured or Asian, and 68 percent African. The white
population had long entered the second stage of demographic transition; the
decline of its birthrate to match the earlier fall in its deathrate had probably
begun as early as the 1890s. By the mid-1980s, whites were barely reproducing
themselves by natural increase. Asian fertility began to decline during the 1940s
and Coloured fertility during the 1960s. Less is known of African fertility; it
was certainly declining by the 1980s and probably during the 1960s, but it was
still higher than that of other races.7 Africans were consequently an ever larger
proportion of the population. By 1986, only an estimated 14 percent of South
Africans were white and 10 percent Coloured or Asian, while 76 percent were
African. The change took place despite white immigration and feverish of¬cial
attempts to encourage large families among whites and contraception among
Africans “ in 1991 South Africa had twice as many family-planning clinics as
health clinics.8 Moreover, the trend was likely to continue. One projection was
that by the year 2005 only 10 percent of the population would be white.9 Not
only did this shift the balance of racial power, but it made a modern economy
impossible to operate unless black people had a larger place in it as producers
and consumers, and it undermined apartheid plans, which assumed that the
African population at the end of the century would be only half what it actually
Given unequal distribution of land and wealth, rapid population growth
bred mass urbanisation on a scale that even Apartheid could not restrain.
Of¬cial ¬gures suggested that in 1970 some 33 percent of Africans were in
towns; estimates in 1985 suggested 58 percent, likely to rise to over 70 percent
by the year 2000.10 In tightly controlled Johannesburg, the immigrants swamped
Soweto, which was built for 600,000 but housed nearly twice as many by 2001.
In Durban they squatted around the city boundaries, where 1,400,000 Africans
were thought to live in 1984. Others huddled into the huge ˜closer settlements™
and, if lucky, travelled up to a hundred miles by bus to work each day. In 1990
one South African in ¬ve lived in ˜informal housing™. Not only were in¬‚ux
control and related measures ineffective, but at their peak they cost about 14
percent of the entire state budget.11 As in the Soviet Union, the state apparatus
designed to control the effects of an industrial revolution could regulate a more
advanced economy and society only by a scale of bureaucracy that was not only
ineffective but obstructed further development. Industrialisation had given the
state the power to impose apartheid; further industrialisation destroyed that
Moreover, the industrial economy was itself in crisis. Whereas the annual
growth of GDP was steadily over 4 percent until the 1970s, it fell thereafter to
Industrialisation and race in South Africa 285

only 1.6 percent during the 1980s, below population growth. The ¬rst symptom
was a decline in gold output, which peaked in 1970. This coincided with a major
structural change in the economy. Until the early 1970s South Africa had oper-
ated a low-wage economy for black people. Real African mining wages were
slightly less in 1969 than in 1911. In 1973, however, ostensibly leaderless strikes
began in the Durban docks and spread to manufacturing industry and the
goldmines. They challenged the low-wage structure at a moment when world
gold prices were soaring and mineowners feared that political independence
in Central Africa might rob them of long-distance migrants. Between 1972 and
1980, therefore, average real African mine wages trebled and the effect spread
throughout the economy, roughly doubling the real wages of full-time farm
labourers. South Africa suddenly became a relatively high-wage economy for
Africans in modern employment, by comparison with other developing coun-
tries. This added to the dif¬culty of exporting South African manufactures, long
hampered by the low skill levels of a hitherto ill-paid and ill-educated work-
force and by relatively small units of production geared to an internal market
restricted by African poverty. Falling investment rates, international sanctions,
East Asian competition, the oil price increase of 1973, and especially the global
depression of 1979“83 exacerbated the crisis. Many employers responded to
wage increases by replacing unskilled labour by machinery, creating more
skilled and permanent workforces at the cost of structural unemployment.
Between 1976 and 1985 the potential labour force grew by nearly 3,000,000 but
formal employment by less than 600,000. By the late 1980s, only about one in
every eight new entrants to the labour force found formal employment.
Structural unemployment was most devastating for the young townsmen
who formed an exceptionally large proportion of the population and domi-
nated the townships while their elders were away at work. Youth unemployment
coincided with major educational changes, for although Bantu Education pro-
vided mediocre schooling, it provided vastly more schooling than before, in
expectation of sustained economic growth. Between 1955 and 1987, African sec-
ondary school students increased from 35,000 to 1,474,300.12 Opinion studies
showed that education made Africans more radical, often exposing them to
the black consciousness ideas of racial self-reliance propagated by intellectuals
like Steve Biko, ideas that freed them from many inhibitions suffered by their
mission-educated parents, just as childhood in huge, all-black conurbations
like Soweto made the young streetwise to the political space that apartheid
provided. ˜We are not carbon copies of our fathers. Where they failed, we will
succeed™, one group proclaimed during the Soweto Uprising of 1976 in which
students protesting against educational policy struck the most important of all
blows against apartheid.13 Repressed early in 1977, insurrection revived in 1984
when students and unemployed youths came together as comrades to seize
control of many black townships, regarding themselves as freedom ¬ghters,
286 africans: the history of a continent

assaulting their opponents with a brutality that expressed the anger of dis-
honoured men from patriarchal societies, and proclaiming allegiance to the
banned African National Congress, which now for the ¬rst time gained the
force reserved in Africa to movements that attract the young.
The township revolt of 1984 was a protest against the National Party™s new
strategy for entrenching white supremacy, after Prime Minister P. W. Botha had
dismissed apartheid in 1979 as ˜a recipe for permanent con¬‚ict™. Recognising
the impossibility of controlling an advanced industrial society by police meth-
ods, Botha™s reform strategy sought instead to use market disciplines. It also
recognised the change that apartheid had wrought in the white population, for
between 1946 and 1977 the proportion of working Afrikaners who had white-
collar occupations had risen from 29 to 65 percent and education made whites
(unlike Africans) politically more moderate. Under Botha the National Party
became a bourgeois party concerned chie¬‚y with white security and appealing
to English as much as Afrikaners, thereby alienating many Afrikaners to right-
wing parties that would win 30 percent of white votes in 1989. Botha sought
to buttress white supremacy by associating with it the Coloured and Asian
people; they gained electoral representation in a tricameral parliament in 1983
and many moved towards the National Party as African power became a possi-
bility. The reform strategy also planned to divide Africans between a well-paid
permanent urban minority with commercial opportunities and trade union
rights, on the one hand, and a majority of impoverished Homeland residents,
on the other. South Africa, as ˜a nation of minorities™, could then become a
loose consociation in which whites and their allies would control the industrial
heartland and dominate a penumbra of black units.
Africans rejected and destroyed this strategy in the township revolt of 1984.
The brutality of the revolt and its suppression also alerted international opin-
ion and prompted economic sanctions, especially credit restrictions, which
damaged South Africa™s ability to restore economic growth. The crucial inter-
national development, however, was the collapse of Soviet communism dur-
ing the late 1980s. Not only did this permit Western powers to press South
Africa for reform, but it offered National Party leaders a unique opportunity to
negotiate an advantageous and lasting settlement before whites were swamped
demographically, while they still held real power, and at a time when African
nationalists appeared weak and isolated. As the new president, F. W. de Klerk,
a conservative National Party stalwart, explained in 1990,

The decline and collapse of communism in Eastern Europe and Russia put a
new complexion on things. The ANC was formerly an instrument of Russian
expansionism in Southern Africa; when that threat fell away, the carpet was
pulled from under the ANC; its base of ¬nancing, counselling and moral
support had crumbled.
Industrialisation and race in South Africa 287

It was as if God had taken a hand “ a new turn in world history. We had to
seize the opportunity.14
He legalised the ANC and released its imprisoned leader, Nelson Mandela, to
provide a negotiating partner.
De Klerk had underestimated the ANC. It proved more popular with
Africans, more united, and less easy to marginalise by encouraging rival African
organisations than National Party leaders had expected. Instead of dominating
the ANC, they became increasingly dependent upon it to achieve a settlement
suf¬ciently acceptable to Africans. Yet Mandela and his colleagues were equally
insecure. Bereft of military force and opposed by entrenched Homeland parties
headed by Inkatha in KwaZulu, they were also acutely and perhaps unreason-
ably frightened of the forces that had thrust them into power. ˜The youths in
the townships™, Mandela warned in 1992,
have had over the decades a visible enemy, the government. Now that enemy
is no longer visible, because of the transformation that is taking place. Their
enemy now is you and me, people who drive a car and have a house. It™s order,
anything that relates to order, and it is a grave situation.15
The election of April 1994 gave the ANC 63 percent of votes, the National Party
20 percent, and Inkatha 11 percent. As Mandela became president with a power-
sharing gevernment, it became clear that the confrontation between the races,
however important, was only the surface of politics. The deeper reality was
that two elites sought a settlement that would enable them to contain, and per-
haps in part relieve, the immense pressures from below bred by demographic
growth, mass poverty, urbanisation, education, and the demands of youth “
the strains arising from the impact of modern change on an ancient colonising
society. South Africa was reentering the mainstream of African history.

In the time of AIDS

as the twenty-¬rst century began, the african continent
was experiencing both crisis and renewal. Economic decline during the 1970s
had obliged governments to accept structural adjustment programmes that
exposed their peoples to two decades of acute hardship before signs of recovery
appeared. As impoverished governments had reduced their services, individ-
uals and groups had drawn upon their own ingenuity to survive. One-party
states had collapsed throughout tropical Africa, leaving behind both violence
and greater freedom, while in the north, Islamic fundamentalism threatened
surviving freedoms while giving purpose to many young lives. The rapid popu-
lation growth of the late twentieth century was slowing, facilitating stabilisation
and economic recovery. In its place, the AIDS epidemic had brought suffering
and new forms of social dislocation, but as the new century began, even this
most terrible of disasters showed the ¬rst signs of hope.

structural adjustment
During the late 1970s, as postwar growth gave way to global recession, indebted
African governments seeking loans from the International Monetary Fund
(IMF) invited the World Bank to examine their economic situation. The Bank™s
response, Accelerated development in sub-Saharan Africa (1981), reversed the
economic strategy of a generation. Written in the newly fashionable language of
monetary economics, the report condemned state-centred development poli-
cies that had exploited farmers and destroyed agricultural exports in the inter-
ests of inef¬cient, corrupt, and urban-based government enterprises. The state,
it proclaimed, was the obstacle rather than the agent of progress. Its role in the
economy must be reduced by privatising public enterprises, removing govern-
ment controls, abolishing subsidies and punitive taxes, charging realistic fees
for services, and allowing currencies to ¬‚oat freely, thereby enabling markets
to ˜get prices right™ so that economies could operate at maximum ef¬ciency.
Adopting this strategy, the IMF made its loans contingent on the adoption of
structural adjustment programmes, intended at this time to be short, sharp
shocks to put economies back on the right tracks. During the 1980s, thirty-six

In the time of AIDS 289

15. In the time of AIDS.
290 africans: the history of a continent

of the forty-seven countries in sub-Saharan Africa and nearly all of those in
the north formally adopted such programmes. Their effects varied widely.
Two of the most positive experiences were in Ghana and Uganda. Under
Nkrumah and his successors, Ghana had epitomised state-centred develop-
ment strategies and its economic growth had fallen far behind population
increase. By 1981 its industry was running at one-quarter of capacity, cocoa
output was only one-third of its earlier peak, and one-¬fth of all Ghanaians
were outside the country, ˜voting with their suitcases™. In that year young rad-
icals led by Flight-Lieutenant Jerry Rawlings seized power in a military coup
with socialist objectives. Within two years, they had switched to a structural
adjustment programme of austerity and market liberalisation that made Ghana
the IMF™s ¬‚agship. The reasons for their change of strategy included their good
sense, the sheer desperation of the country™s plight, the insistence of the inter-
national ¬nancial institutions, the lack of practicable alternative policies, the
bankruptcy of vested interests that might have opposed reform, and the absence
of electoral democracy during the regime™s ¬rst eleven years. Its initial step was
to devalue the currency by over 98 percent. By the late 1980s, liberalisation
of cocoa marketing had doubled producer prices and output. Civil service
employment was halved and tax revenue doubled. The annual growth of Gross
Domestic Product rose from 1.4 percent in 1965“80 to 3.0 percent in 1980“90,
4.2 percent in 1990“2001, and 4.8 percent in 2000“4.1 These were modest ¬gures
and not everything was successful. In 1998 real GDP per capita was still lower
than it had been in 1970. Ghana™s old and unproductive cocoa trees struggled
to compete with plantations outside Africa. Manufacturing employment fell
by nearly two-thirds between 1987 and 1993 as protection against imported
goods was withdrawn. The restoration of elections from 1992 led to ˜demo-
cratic demand in¬‚ation™ as governments expanded the money supply to win
votes, the currency depreciated, and debt rose by 1998 to three times its level
in 1981. The short, sharp shock became a seemingly permanent condition, but
economic decline was at least reversed.
Uganda™s experience was similar. General Amin™s regime and the subsequent
civil war had reduced per capita GDP by 42 percent between 1971 and 1986,
when President Museveni™s National Resistance Movement (NRM) took power.
After a brief experiment with radical policies had brought the economy close to
collapse, the NRM adopted a structural adjustment programme, which vested
interests were too weakened to resist. The currency was devalued by 76 percent,
the heavy taxation on coffee exports that had ¬nanced previous regimes was
abolished, export volume grew at 15 percent per year during the 1990s, producer
prices multiplied three or four times, the civil service was halved, the share of tax
revenue in GDP doubled, and the annual increase in real GDP between 1986
and 1999 averaged 6.3 percent. In 1996 the IMF declared Uganda™s economy
the most open in Africa. As in Ghana, the price was dependence. During the
In the time of AIDS 291

1990s, Uganda received nearly $5 billion in foreign aid, which ¬nanced between
one-third and two-thirds of public expenditure. Privatisation fostered rampant
corruption. Yet the recovery was real and GDP continued to grow during the
early 2000s at 5.8 percent per year.2
The most surprising convert to liberalisation was the new ANC govern-
ment in South Africa. When it took power in 1994, real per capita GDP had
declined during the previous 21 years by an average of 0.6 percent per year.3 To
reverse this and begin to rectify massive economic inequalities, the new regime
adopted a ¬ve-year Reconstruction and Development Programme (RDP)
designed to achieve growth through redistribution, with provision to nation-
alise commanding heights of the economy, redistribute 30 percent of agricul-
tural land, and build a million new houses. By 1996, however, there was little
sign of recovery and it was becoming clear that this ¬rst attempt to restructure
an industrial economy, without a prior social revolution and within a domi-
nant world capitalist system, must take much longer than radical nationalists
had dreamed. Instead, GDP was just keeping pace with population growth and
employment was falling rapidly as government pruned its payroll, goldmin-
ing contracted, commercial farmers reduced workforces, major ¬rms moved
their headquarters to Europe, and liberalisation of the exchange rate in 1996
led to a rapid fall in the value of the rand. South Africa was not yet at the
IMF™s mercy, but its leaders clearly resolved that stabilisation must temporar-
ily have priority over transformation. In 1996, without consultation, they
replaced the RDP by a new strategy, Growth, Employment, and Redistribution
(GEAR), which abandoned nationalisation, accelerated privatisation, restricted
public expenditure, reduced import tariffs to stimulate competitiveness,
encouraged foreign investment and labour market ¬‚exibility, promoted
exports, and prioritised growth over redistribution. When the eagerly mod-
ernising Thabo Mbeki became president in 1999, he insisted ˜that we aban-
don our embarrassment about the possibility of the emergence of successful
and prosperous black owners of productive property.™4 The chief bene¬cia-
ries of the new regime were indeed middle-class Africans, whose numbers
rose at an estimated 21 percent per year between 1993 and 2003. The chief
losers were the unskilled, for although GEAR promised to create 600,000 jobs
over ¬ve years, South Africa actually lost 500,000 during that period, chie¬‚y in
government, agriculture, and goldmining. The Confederation of South African
Trade Unions mounted three general strikes against the GEAR programme
during 2000“2. Only in the mid-2000s did unemployment show the ¬rst
signs of decline. By then, moreover, GDP growth at about 4 percent per
year was outrunning a nearly stable population, partly owing to a new surge
in gold prices. ANC leaders tacitly abandoned GEAR, shelved privatisation,
planned to expand public investment, and resumed their redistributive ambi-
tions, spending heavily on old-age pensions and child-support grants and
292 africans: the history of a continent

requiring that one-quarter of mining and industrial assets and 30 percent of
agricultural land should pass into African hands within a decade. While rad-
ical observers complained that the ANC had ˜squandered an opportunity of
world-historic proportions™, its leaders replied that ˜carefully measured actions
and studied moderation™ had ˜helped reassure skittish investors and interna-
tional markets™, ensuring ˜a decade of social peace underpinned by political
Elsewhere in the continent, structural adjustment provoked varied responses
and effects. In North Africa, economic liberalisation from above tended to
strengthen political authoritarianism, notably in Tunisia, which implemented
its own successful programme and enjoyed steady if unspectacular growth. In
Egypt, by contrast, policy ¬‚uctuated between Sadat™s hasty liberalisation of the
1970s, which brought rapid growth and massive foreign debt, and Mubarak™s
grudging response to international ¬nancial pressures. Both regimes were
especially reluctant to privatise state enterprises underlying their economic
and political power, a reluctance even stronger in Algeria, whose government
insisted on retaining ownership of the oil and gas industries, which in the mid
1990s provided 97 percent of foreign exchange earnings. North Africa™s strate-
gic importance curbed the IMF™s reforming zeal, but there was less restraint in
Central Africa, where the impact of structural adjustment was perhaps most
damaging. Zambia™s acute economic crisis, caused chie¬‚y by the collapse of
world copper prices during the 1970s, obliged it to negotiate an adjustment
programme in 1983, but unlike Ghana and Uganda its unpopular government
had been in power for twenty years, the programme did not provide foreign
exchange to support liberalisation, the currency sank by 90 percent in two years,
and the IMF insisted on the removal of food subsidies for the 45 percent of
Zambians living in towns. The result was strikes and urban riots, the govern-
ment™s collapse, and further economic decline to the end of the century. Even
more damaging was the programme devised in 1991 for Zimbabwe™s relatively
successful economy. By opening its industries to South African competition,
structural adjustment reduced manufacturing output by 21 percent in four
years, exports declined, unemployment rose, debt increased sharply, and when
world tobacco prices also fell at the end of the decade, the regime took refuge
in an expropriation of surviving European farms, which devastated agricul-
tural production. Between 2000 and 2004, Zimbabwe™s GDP probably declined
by about 30 percent, the currency lost 99 percent of its value, and perhaps
three-quarters of the population were reduced to poverty.6
It would be wrong to exaggerate the power that international ¬nancial insti-
tutions could exert over African governments. Most rulers recognised that
structural reform threatened their revenue, patronage, and freedom of action,
˜like telling the people to rise against us™, as President Stevens of Sierra Leone
complained. Some, like Senghor in Senegal and Nyerere in Tanzania, refused
In the time of AIDS 293

to implement adjustment programmes. Others obstructed them, as in Kenya,
or reduced them to cosmetic gestures, as in Cameroun. Everywhere certain
liberalising measures, such as ¬‚oating the currency, were easier to implement
than others, notably privatisation, which not only threatened vested interests
but often meant either transfering national assets to political favourites at bar-
gain prices or selling them to foreign investors, notably the South African
conglomerates that took the opportunity to buy most of sub-Saharan Africa™s
mining enterprises and many banking, brewing, retailing, electricity, and air-
line companies. One study of IMF programmes found that only about half were
completed during the loan period,7 but once the international institutions had
committed funds, they could do little but continue lending in the hope of
eventual success, for their own reputations were tied to the policy. Controversy
surrounding it gradually obliged them to shift their ground. Whereas Acceler-
ated development had insisted in 1981 on reducing the role of the state, by 1997 “
in the face of state collapse in countries like Somalia and Sierra Leone “ the
World Bank detected ˜a crisis of statehood™ in Africa and stressed the need to
extend state capacity,8 recognising that effective response to liberalising stimuli
often needed state backing and even state entrepreneurship. Two years later,
the international institutions replaced structural adjustment programmes by
Poverty Reduction Strategy Papers, which were three-year plans to be drawn
up by recipient governments in consultation with local business interests, trade
unions, and nongovernmental organisations before submission for the donors™
approval. The aim, according to the head of the IMF, was ˜to teach a society,
not just a government, how to live within its means™.9
By the beginning of the new millennium, no African state had yet emerged
from the structural adjustment process, while the IMF, the paragon of ¬nancial
prudence, was massively overlent to African governments. Yet there were signs
of recovery. Whereas sub-Saharan Africa™s annual growth rate of GDP had fallen
from its high level of 4.8 percent in 1965“80 to only 1.7 percent in 1980“90 and
2.6 percent in 1990“2001, between 2000 and 2004 it rose again to 3.9 percent,
well above the population growth rate of 2.2 percent.10 Whether this signi¬ed
lasting improvement, and whether structural adjustment was responsible for
it, was not yet clear.

state contraction and cultural change
As economic decline and structural adjustment cut into public revenue, state
services contracted and society, always the true strength of African civilisation,
adapted to new conditions as it had in the past adapted to the slave trade or colo-
nial rule. Education best illustrated this process. Most nationalist leaders had
owed their positions to schooling and had invested massively in it. Between 1960
and 1983, primary school enrolment in black Africa had roughly quadrupled,
294 africans: the history of a continent

secondary school places had multiplied sixfold, and the number of university
students had increased twenty-fold.11 It had been one of the great achievements
of independence, with important political implications. Thereafter, however,
education systems faltered as the ever-expanding child population pressed on
diminishing resources. University education, although often expensive and
declining in quality, was still coveted as a quali¬cation for employment. Between
1994 and 2000, Makerere University in Uganda expanded its enrolment from
seven thousand to twenty-two thousand students. Secondary school enrolment
also continued to rise. Primary schooling, by contrast, gave little advantage in
employment unless it led on to the secondary level. When primary education
was made free, as in Uganda in 1997 and Kenya in 2003, there was a massive
increase in enrolment, but in Tanzania, where in 1991 only 5 percent of primary
school leavers gained entry to public secondary schools, primary enrolment
fell between 1981 and 1997 from 94 to 67 percent of the age group.12 Yet many
parents, convinced that education was the chief means of advancement open to
their families, responded by providing their own schools, especially at the sec-
ondary level. By 1995 those in Tanzania outnumbered state secondary schools.
Many were run by churches, but some Islamic regions organised similar private
systems, even in Egypt. Kenya was one of many countries where pro¬t-making
schools proliferated.
Health care followed a similar pattern. Between the 1960s and the mid-1980s,
the increase in medical personnel, together with cheap drugs, immunisation
procedures, and general development, had reduced infant mortality by nearly
one-third and spared Africa any major epidemic. These medical interventions
were so powerful that their impact continued, less dramatically, through the
1980s on a continental scale, but the impact waned in areas of violence, famine,
and extreme economic decay. Ghana™s real per capita public expenditure on
health fell by 60 percent between 1974 and 1984; eight years later the country
had some ¬fty thousand cases of yaws, a disease of poverty supposedly eradi-
cated before independence, and its child mortality rate had risen. Tuberculosis,
cholera, and yellow fever became more prevalent, while at the end of the century
perhaps half a million Africans contracted sleeping sickness each year and one
in six suffered a clinical case of malaria. To these was added the AIDS epidemic.
Life expectancy at birth in sub-Saharan Africa had risen from 42 years in 1965 to
a peak of 53 years in 1996; by 2003 it had fallen again to 46 years.13 As state med-
ical systems were overwhelmed, many Western-trained doctors took refuge in
private practice. Popular responses included buying modern drugs from prolif-
erating retail outlets, consulting traditional practitioners, and adding modern
drugs to local pharmacopoeia in Africa™s characteristically eclectic manner.
Access to effective medicine depended increasingly on wealth, so that infant
mortality rates varied more widely with income in cities like Abidjan than they
had even in nineteenth-century Europe.
In the time of AIDS 295

This did not halt migration to towns, although it had slowed since the 1960s
and now focused on provincial centres rather than capitals. During the 1980s
and 1990s, sub-Saharan Africa™s townsmen increased twice as fast as its popula-
tion, forming 34 percent of the total in 1999 and nearly twice that proportion in
South Africa. Some of the most rapid migration was by those ¬‚eeing rural dislo-
cation, notably in Tanzania and Mozambique. The ANC government in South
Africa built or subsidised nearly two million new housing units during its ¬rst
twelve years in power, but elsewhere public housing provision lost all contact
with need, so that the poor clustered either into single rooms at extortionate
rents or into the self-built family shacks on city fringes known in Nouakchott
as ˜refuse dumps™. Whereas urban wages had far exceeded rural earnings during
the 1960s, they fell over 30 percent on average during the 1980s. Yet employ-
ment was still a privilege, because the poor, hitherto mainly the incapacitated,
now included the able-bodied unemployed. Statistics in the early 1980s gener-
ally showed them as 8 to 15 percent of the potential urban labour force, but
the proportion grew thereafter, reaching some 25 percent in Kenyan cities and
40 percent in South Africa in 1996, the latter ¬gure in¬‚ated by the lack of viable
peasant agriculture. Yet even these ¬gures were misleading because many of the
poor could not afford to be unemployed and instead undertook ˜occupations™
with minute earnings. The totally unemployed were mostly young people still
relying on family support. Over half of Algerians in their early twenties were
unemployed in the late 1980s as population growth and education outstripped
jobs. Street gangs like the Ninjas of Lusaka and Talibans of Nairobi ¬‚ourished.
In 1988 one of every ¬ve Nigerian prisoners was a teenager. An anthropologist
studying the Zambian Copperbelt at that time found ˜an overwhelming sense of
decline and despair™ as expectations of working lives in a modern environment
were dashed.14
Survival in decaying cities depended heavily on informal occupations, which
employed some 72 percent of Nigeria™s urban labour force in 1978, including its
innumerable women traders and youthful apprentices. Even in South Africa,
where the authorities had long repressed informal enterprise, it had expanded
by 2000 to generate an estimated 28 percent of GDP. Self-employed earnings
could be relatively high, but employees in the sector were severely exploited
and many young men began their working careers in unpaid jobs. Informal
occupations merged into the ˜second economy™ of black-marketeering, smug-
gling, corruption, and crime, which expanded as state power contracted. These
activities commonly relied on ethnic and family ties. Private schools, informal
enterprises, illicit trading diasporas, vigilante forces in place of nonexistent
police, and urban welfare associations in lieu of ineffective trade unions all
mobilised ethnic solidarities, as did the continental passion for football. Within
the family, elite women had generally enhanced their status since political
independence, except in North Africa where fundamentalism reversed earlier
296 africans: the history of a continent

gains. All adult women had commonly gained the vote and in some coun-
tries, including Uganda and South Africa, they exercised signi¬cant political
in¬‚uence. In 2006 Liberia elected the continent™s ¬rst woman president. Some
urban working women also gained greater equality within their families as
informal enterprise and female employment expanded while male earnings in
formal employment contracted. Peasant women, on the other hand, suffered
severely as a result of economic decline and the increasing pressure of popula-
tion on land. And all women, particularly the young, were especially vulnerable
to AIDS.
Religion was a further nexus of social solidarity, but it was also probably the
chief perspective through which Africans thought about their often threatening
world. Christianity and Islam spread widely as people hitherto excluded “ espe-
cially women and remote communities “ claimed places in the modern world.
˜Everybody had joined™, one woman explained, ˜and I was left behind like a fool™.
One estimate was that between 1950 and 1990 African Christians increased from
34 million towards 200 million.15 The most rapid expansion was in Kenya and
Zimbabwe, in the rapidly growing cities, and in the Sudanic region running
from Senegal to Ethiopia. Many southern Chadians and Sudanese accepted
Christianity during resistance to northern domination, while con¬‚ict between
expanding Christian and Muslim fundamentalists bred grave con¬‚ict in cen-
tral Nigeria. Church hierarchies were rapidly Africanised. In 1993 the Roman
Catholic Church had sixteen African cardinals. At the Lambeth Conference
of the Anglican Church ¬ve years later, African bishops outnumbered their
British counterparts by three to two. Missionaries and money continued to
¬‚ow into the continent, especially from North America, while African voca-
tions increased rapidly at the end of the century as competition for secular
employment intensi¬ed. Yet Christian numbers nevertheless outran pastoral
capacity, fostering a peasant Christianity in the Ethiopian manner with strong
village congregations, sparsely trained evangelists, little superstructure or in¬‚u-
ence on family life, much partly Africanised ritual, and much eclectic survival
of indigenous practices.
The expanding independent churches “ thought to number up to ten thou-
sand in the late 1980s “ often had similar structures. A few had become major
hierarchical institutions, notably the Kimbanguist Church in the Democratic
Republic of the Congo and the Zion Christian Church in South Africa, but
most were very small, providing supportive communities, spiritual protec-
tion and healing, and personal empowerment amidst economic austerity and
state contraction. Some of the newest churches also held strongly millenar-
ian beliefs, especially perhaps in western Uganda where the AIDS epidemic
was acute. ˜This chastisement He released . . . the world calls it AIDS dis-
ease . . . but from the Lord it is a punishment™, leaders of the Movement for
the Restoration of the Ten Commandments of God told their followers before
In the time of AIDS 297

killing over a thousand of them in March 2000 as they awaited their promised
˜new generation™.16
The need for certainty amidst the malaise and intellectual confusion sur-
rounding late twentieth-century Africans may explain the most remarkable
Christian phenomenon of the time: the explosive growth of pentecostal
churches, which had ¬rst reached South Africa early in the century but spread
throughout sub-Saharan Africa from the 1970s. By 2000, some 24 percent of
all Ghanaians claimed to be pentecostalists. These churches characteristically
stressed personal salvation through repentance and the empowering baptism of
the Holy Spirit. Unlike earlier independent churches, they rejected the African
past in favour of a globalising modernity. One impetus, especially in West Africa,
came from well-¬nanced American missionaries who taught the ˜prosperity
gospel™ that God would reward in this world those who ¬rst gave generously to
His work, a teaching congruent with indigenous expectations of this-worldly
bene¬ts from religion. Even pentecostal churches rejecting this teaching nev-
ertheless often appealed to upwardly mobile young urban Africans by display-
ing modernity and internationalism, stressing individualism and the nuclear
family, condemning corruption, and offering enticing prospects of success.
Rural pentecostalism, likewise, might champion an austere morality that
denounced polygyny, drink, patriarchal domination, witchcraft, and indige-
nous religious practices as works of the Devil.
This radical dualism, reminiscent of early North African Christianity, was
the distinctive feature of pentecostalism. It pictured the world as a battleground
between, on the one side, God and his born-again faithful, and, on the other,
Satan, the witches, the old gods, and those who had sold their souls for wealth
and political power. At one level, this bred a luxuriant demonology displayed
in video ¬lms and articulated by an of¬cial Kenyan enquiry into devil-worship,
which reported satanic cults with cannibalistic initiation rites, blamed them
for rail and car crashes, and recommended the screening of religious organisa-
tions, a ban on music in minibus taxis (notorious arenas for dissent), and the
censorship of televised wrestling programmes.17 At another level, pentecostal-
ism offered a critique of the state and social order as radically corrupted that
was psychologically satisfying to those who thought of politics in moral terms.
Dualism did not necessarily divide communities irremediably. As an anthro-
pologist noted, ˜People operate with black-and-white contrasts, but only to
create ever more complex new patterns.™18 Yet dualism did pervade other
attempts to comprehend the forces that Africans felt acted upon them. One
feature of the late twentieth century was the intense concern with which both
the powerful and the powerless regarded witchcraft. Nervous authorities in
many regions prosecuted suspected witches, relying upon diviners for ˜expert™
testimony. Villagers were more likely to kill their suspects by mob action.
Similarly, Islam not only shared Christianity™s numerical growth but displayed
298 africans: the history of a continent

a comparable dualism, most strikingly in fundamentalism “ as will be seen
later “ but also in millenarian protest such as that launched by the hetero-
dox teacher, Muhammadu Marwa, who created a ˜private republic™ of young
rural immigrants in Kano, ˜preaching that anyone wearing a watch, or riding
a bicycle, or driving a car, or sending his child to the normal State Schools was
an in¬del.™19 Its dispersal by the military in 1981 cost some four thousand lives.
That was the most violent protest in a postcolonial city. In normal circum-
stances, the urban poor, while resenting corruption and the gulf between ˜us™
and ˜them™, were too vulnerable, divided, dependent on patronage, committed
to rural values, and aware of recent social mobility to challenge their rulers
openly. Religious zealots might bring them into the streets. So occasionally
might organised trade unionists, as in the Three Glorious Days that destroyed
Brazzaville™s government in 1963. So might the breakdown of order during a
coup d™´ tat, as in the orgy of looting in Nairobi in 1982. The most common
urban disturbance was the ˜IMF riot™ against increased food prices, often due to
the removal of subsidies decreed by a structural adjustment programme. Such
riots brought down regimes in Liberia, Sudan, and Zambia and threatened
several others. Less often, but with greater brutality, urban crowds might turn
against the enemy within, as in Kinshasa in 1998, when townspeople responded
to Rwandan invasion by massacring available Tutsi, or in Abidjan in 2004,
when crowds revenged themselves on French people and property after ˜peace-
keepers™ had destroyed the Ivoirian airforce. Abidjan™s disturbances were the
work of ˜young patriots™, the militant youths whose counterparts had idolised
Nkrumah, liberated Zimbabwe and Uganda, and destroyed apartheid. In less
violent circumstances, most late twentieth-century African cities had distinct
youth cultures blending indigenous traditions with the coveted modernity of
global fashions. The boys Dakar of the 1990s, for example, cultivated a bul faale
(play it cool) ethos valorising hard work and urban sophistication, advocated
sopi (political change), spoke a Franco-Wolof argot peppered with American,
listened avidly to iconoclastic rap, and idolised a champion wrestler from their
milieu known as Tyson. Products of rapid population growth, educated for
unemployment, unable to gain married adulthood by traditional means, and
denied the rapid social advancement their fathers had enjoyed, they were avail-
able for change, whether creative or destructive.
The youth cultures were not the only groups that challenged the social
order established at independence. Philosopher-presidents from Senegal and
Tanzania had long been obliged to share the OAU with sergeants and street
politicians. Now, as government revenues dwindled, of¬cial salaries shrivelled,
and informal enterprises expanded, social preeminence often passed from the
´volu´ to the am´ricain who had made good as a migrant businessman in
e e e
the West, or even to the moodu moodu, the wealthy but uneducated trader.
Even Western-trained medical doctors, the ˜cream of the cream™ among the
In the time of AIDS 299

educated elite, stooped to seek a living wage by striking in de¬ance of profes-
sional ethics. In 2000 the entry quali¬cation to study medicine at Makerere
University was 15 percent lower than that for the more lucrative course in
pharmacy. Clergymen with doctorates in theology shrank before the spiritual
entrepreneurs of the prosperity gospel. Some members of the educated elite
took refuge in the private sphere, but many resisted loss of status. Often their
weapons were the nongovernmental organisations (NGOs) that proliferated
as contracting one-party states lost their capacity to monopolise public life. In
1980 Nigeria had 1,350 NGOs; in 2000, 4,028. A new style of African politics was
coming into being.

political change
When the Berlin Wall fell in 1989, forty-two of sub-Saharan Africa™s forty-seven
states had authoritarian regimes without seriously competitive elections. Five
years later, not one was of¬cially a one-party state, thirty-eight had held compet-
itive elections during the period, and sixteen formerly authoritarian countries
had installed new regimes through elections.20 Although the democratisation
movement of those years did not ful¬l its highest hopes, it was for many states
a new political beginning.
The movement originated in Benin (the former Dahomey) in January
1989, when students, unpaid civil servants, and urban crowds expressed their
economic grievances against President K´ r´ kou™s regime by protest marches
modelled on the oma, the demonstration of collective anger customary among
local vodun priests. K´ r´ kou™s army withdrew its support and France suspended
aid until he convened a national conference. It was designed to emasculate the
opposition, but instead its members declared themselves sovereign and ordered
elections in March 1991 in which a black African leader was for the ¬rst time
replaced through the ballot box. This model then reverberated through fran-
cophone Africa. National conferences and elections replaced leaders relatively
peacefully in Niger and Congo-Brazzaville. Elsewhere there were variations.
The astute Houphouet-Boigny in Cote d™Ivoire called a multiparty election so
quickly that his opponents could not organise themselves to prevent his reelec-
tion. President Biya in Cameroun refused a national conference, arranged a
meeting packed with his own supporters, and then won a multiparty election
thanks to the division among his rivals. In Togo, President Eyadema relied on an
army dominated by his fellow-tribesmen, allowed his political enemies to hold
a conference, dissolved it when it declared itself sovereign, and won an election
that the opposition boycotted. Mobutu, similarly, postponed a national confer-
ence four times, packed its 2,842 delegates with his supporters, let it meander
for eighteen months while over two hundred parties were formed, used it as
an excuse to avoid presidential elections, and eventually dismissed it.
300 africans: the history of a continent

Democratisation followed a similarly chequered course in anglophone
Africa, although without the echoes of 1789. The pioneer was Zambia, where
IMF riots mobilised the organised copper-miners and ruling party dissidents
to demand multiparty elections, which ousted President Kaunda in 1991 after
twenty-seven years in power. This inspired imitation elsewhere, especially in
neighbouring Malawi. Yet again there were variations. In Kenya, President
Moi resisted multiparty voting, insisting that it would tear the country apart.
When aid donors nevertheless forced him to hold elections in 1992, he spent
U.S.$100 million on the campaign, divided the opposition, and won with only
36 percent of the vote, repeating the trick in 1997 and witnessing his party™s
defeat only after his retirement in 2002. His neighbour in Uganda, President
Museveni, used the argument that party competition had nearly destroyed
his country during the 1960s to resist its reintroduction until 2006, when he
defeated his rivals. Tanzania™s powerful ruling party appointed a commission
to canvass public opinion, found that 77 percent of respondents preferred a
one-party state, but nevertheless instituted competitive elections in 1995 and
won them handsomely.
Many circumstances underlay this remarkable change in political behaviour.
The most fundamental was simply the passage of time, which robbed nationalist
parties of unique legitimacy either as liberators or agents of development.
One-party states had lost any claim to be modern, a major consideration for
many Africans. Structural adjustment had deprived regimes of patronage and
radicalised both urban crowds and many educated people. Moreover, state
contraction had fostered the civic associations and NGOs that characteristically
provided the delegates to national conferences. Every francophone conference
was chaired by a Catholic prelate, for this was in part a struggle by educated
elites to regain lost status and reassert the active associational life smothered by
one-party states. The relative strength of social groups was one determinant of
the success of democratisation movements, as in the role of Zambia™s copper-
miners, but these groups made no headway against Eyadema or Mobutu so long
as the soldiers remained loyal, whereas armies in Benin, Niger, Mali, Congo-
Brazzaville, and Malawi all refused to suppress demands for democracy. ˜As went
the military, so went the transition™, one analysis concluded.21 External factors
were also important. The movement in Benin preceded the fall of communism
in Eastern Europe, whose direct impact on events in Africa was probably quite
small, although a visit to postcommunist East Germany apparently convinced
Nyerere that Tanzania™s one-party system was doomed. Indirectly, however, the
end of the Cold War freed Western donors to coerce Moi in Kenya and Banda
in Malawi to accept liberalisation, although no such pressure was attempted in
Cameroun, Togo, or in North Africa where strategic considerations remained
In the time of AIDS 301

The changes brought about by democratisation were limited but impor-
tant. Only sixteen authoritarian regimes were changed through elections by
1994. In ¬fteen others the single parties in power in 1989 were still in power in
2002 after winning multiparty elections. Even where regimes changed, politi-
cians often remained the same: ˜born-again politicians™, as they were known in
Malawi, who had been ousted from one-party governments, regained power
through democratisation, and still saw the state as chie¬‚y a source of personal
advancement. During the ¬rst seven years of democratisation, the number of
parliamentarians in sub-Saharan Africa increased by 22 percent.22 The process
was about people rather than policies or, for the most part, political structures.
Not a single left-wing party emerged. Moreover, democratisation was largely
an urban phenomenon. National conference delegates came overwhelmingly
from the towns, especially the capitals. In the countryside, the vulgarity and
divisiveness of party competition were often resented, so that old-style auto-
crats like Mugabe in Zimbabwe could frustrate urban pluralism by mobilising
rural support. Multiparty voting was chie¬‚y on regional lines, often exacerbat-
ing ethnic and religious tensions, and money changed hands even more freely
than before. Africa™s economic and social structures had not altered suf¬ciently
over half a century to make political democracy decisively easier to entrench
than it had been at independence.
In 1997 three leading political analysts declared ¬‚atly that democratisation
had failed.23 They had in mind not only the limitations just described but also
the reversions to authoritarianism that had taken place. The ¬rst multiparty
democracy to collapse was in Burundi, where an election in 1993 returned a
largely Hutu government overthrown after four months by the Tutsi army. A
year later Rwanda™s experiment in liberalisation ended in genocide. The presi-
dent dethroned by Congo-Brazzaville™s national conference regained power in
1997 by civil war. During the twelve years after multiple parties were legalised
in Niger in 1991, the country experienced three republics, eleven elections or
referenda, and two coups d™´ tat. Even Cote d™Ivoire, once the most stable of
West African states, was reduced to partition and near civil war in 2002 by the
folly of rival politicians. Experience in other continents showed that although
a country™s poverty did not prevent the establishment of democracy, poverty
made it very dif¬cult to sustain.
Yet this was too negative an assessment. Although no democratic regime in
Africa was yet ˜the only game in town™, nevertheless by the end of 2000 some
thirty-¬ve countries had held second elections, and ten had held a third.24 Many
old-style politicians had survived, but younger activists had also secured elec-
tion, as had members of social groups excluded from in¬‚uence at independence.
Candidates for election had to pay far more attention to their constituents than
had hitherto been normal. Tanzania™s eight television channels, ten or more
302 africans: the history of a continent

radio stations, and forty newspapers in 2000 illustrated an extraordinary and
widespread liberation of expression, just as transparent ballot boxes exempli-
¬ed a new concern with democratic procedures. With the possible exception
of Botswana, this was presidential rather than parliamentary democracy, but
presidential terms now had limits that could be transgressed only by sacri¬c-
ing legitimacy. Democracy had not supplanted the practices of the one-party
state, but it had blended with them. The most characteristic outcome was
the dominant party state, where one party “ often the old single party “ won
elections, exercised patronage, and embodied national unity, while especially
discontented groups and regions were free to create a penumbra of minor par-
ties. The Indian National Congress had demonstrated that this was a workable
transitional form of government for large and heterogeneous countries. In the
Africa of the early 2000s, it was the pattern in Botswana, Tanzania, Ethiopia,
Egypt, Tunisia, Sierra Leone, Mali, Cameroun, Gabon, Namibia, and South
South Africa™s presence in the list further demonstrated the continuity char-
acterising its transition to majority rule, although here the transition was from
a dominant white party to a dominant black one. The transitional election of
1994 was in effect an ethnic plebiscite: only 3 percent of whites voted for the
ANC and 3 percent of Africans for the National Party. The ANC won 63 per-
cent of votes and increased that proportion slightly at elections in 1999 and
2004, although its parliamantary representation was notably multiracial and
relatively balanced between men and women. The National Party ceased to
exist at the national level in 2005, leaving fourteen minor parties sharing 107 of
the 400 parliamentary seats, the most important representing Africans in rural
KwaZulu-Natal and the Coloured and white populations of the Western Cape.
The new constitution took still further the gradual centralisation of South
Africa during the twentieth century, making the provinces dependent on the
central government for over 95 percent of their revenue. From 1998 provincial
premiers were no longer elected but appointed by the president. Three years
later, the central government exercised its constitutional power to take tem-
porary control of the Eastern Cape when the provincial authorities ran out of
money to pay social grants. In the meantime, ANC leaders returning from exile,
with the authoritarian tendencies of a liberation movement, had established
their dominance over the innumerable civic organisations that had conducted
the internal struggle against apartheid during the 1980s. The government had
also broken free from formal consultation with organised labour and asserted
its sole prerogative to make policy. While preserving a degree of internal democ-
racy, the ANC had become chie¬‚y an electoral party. A Constitutional Court
was an important check on executive power and parliament remained more
effective than in most African states, but South Africa under President Mbeki
moved signi¬cantly towards the presidential form of democracy.
In the time of AIDS 303

Nigeria, by contrast, struggled towards a dominant party state. The possibil-
ity seemed to exist after 1979, when the military government victorious in the
civil war restored a civilian political structure in which nineteen states com-
peted for the power and revenue controlled by the oil-rich federal government.
The northern-based National Party of Nigeria (NPN) secured a narrow victory
for its presidential candidate in 1979, with only 34 percent of the total vote, but
during the next four years its control of patronage attracted minority groups
in all regions, although not the core Yoruba and Igbo peoples. In the 1983 elec-
tion, the NPN claimed 264 of the 450 National Assembly seats and 47 percent
of the presidential votes and was perhaps poised to become a dominant party,
but the election was so corrupt and the economic situation so parlous that the
military again seized power. For the next sixteen years, they frustrated political
development, especially by annulling in 1993 a presidential election, designed
to restore civilian democracy, at the moment when it became clear that a south-
ern candidate (a Yoruba) had won. This reopened the regional divisions that
had caused the civil war of 1967“70. After six further years of military dicta-
torship, democratic elections in 1999 returned the People™s Democratic Party
(PDP) “ modelled on the ANC of South Africa “ and a president, Olusegun
Obasanjo, who won support in all regions except his own embittered Yoruba-
land. During the following years, the PDP gathered Yoruba support, but its bid
for political dominance was threatened by the growth of religious fundamen-
talism, by violent con¬‚ict for control of cities in the Middle Belt between north
and south, and by the determination of northern zealots to enforce sharia law
in Muslim areas. In 2003 Obasanjo secured reelection with 62 percent of the
votes and the PDP won twenty-eight of thirty-six state governorships, but now
the core Muslim states of the north were in opposition. With Obasanjo barred
from standing again in 2007 and no outstanding alternative candidate, many
Nigerians feared for the political future.
The growth of Islamic fundamentalism was a phenomenon comparable
in signi¬cance to democratisation. As originated in Egypt by the Muslim
Brotherhood from 1928, fundamentalism (or Islamism, as its adherents termed
it) emphasised core beliefs: that God, not the people, was sovereign; that His
will was solely and completely revealed in the Koran; that His law was not
subject to human variation; that a secular state was by de¬nition anathema;
that the nation-state, likewise, must give way to the Muslim community; and
that innovations (including religious brotherhoods) were forbidden. To these
principles the most in¬‚uential Egyptian fundamentalist, Sayyid Qutb, added
in the 1960s the idea that all existing states, even those professing Islam, were
in reality pagan, comparable to the jahiliyya preceeding the coming of Islam,
and must therefore be targets of jihad “ a demonisation similar to that by
Christian pentecostalists. This radicalisation was a response to persecution by
Nasser™s regime, which executed Sayyid Qutb, and it was intensi¬ed by Israel™s
304 africans: the history of a continent

defeat of Egypt in the Six Days War of 1967, the success of the Iranian Revolu-
tion of 1978“9, and refusal by Egypt™s rulers to allow the Muslim Brotherhood
to function as a political party. In reaction, younger militants, mainly from
peasant backgrounds in neglected southern Egypt, launched terrorist organ-
isations during the 1970s. Repressed by the police, these took refuge either
in Cairo™s informal housing areas or in exile, where one leader, Ayman al-
Zawahiri, became Osama bin Laden™s deputy. In the meantime, however, the
Muslim Brotherhood™s moderate fundamentalism gained increasing in¬‚uence
over Egyptian law and culture. The election of eighty-eight of its members to
the legislature as independents in 2005 made it the chief opposition to President
Mubarak™s ailing regime.
Elsewhere in North Africa, fundamentalists found equal dif¬culty in secur-
ing political power against the opposition of rival Muslim groups and
national governments. They were most successful in Sudan, where the Muslim
Brotherhood, which had arrived from Egypt during the 1940s, found that it
could not win election against entrenched political parties and instead allied
from 1979 with successive military regimes in which its gifted leader, Hassan
al-Turabi, exercised great in¬‚uence until his ambition led the military to oust
him in 2000. In Tunisia, by contrast, the Movement of the Islamic Way, formed
in 1981 and later renamed en-Nahda, faced an unusually disciplined and secu-
lar dominant party that allowed fundamentalists to contest election with some
success as independents in 1989 but suppressed the movement three years later
when its radical wing adopted terrorist tactics. The Libyan government, which
asserted its own Islamic character, banned the local Muslim Brotherhood. The
Moroccan monarchy also claimed religious authority, through descent from
the Prophet, and exercised an often ruthless authority over a society where
brotherhoods and other Islamic traditions were strong. Several fundamental-
ist groups operated here with limited freedom. The most important, led from
1981 by Abd al-Salam Yasin, focused chie¬‚y on spiritual revival rather than state
power but emerged after 1999 to lead popular opposition to the liberalising pro-
gramme of the young King Mohammed VI, while a moderate fundamentalist
party won forty-two parliamentary seats in 2002.
The major contest between fundamentalists and the state took place in
Algeria. In the early 1980s, the temporary collapse of oil prices threw the
country™s heavy industrialisation programme into disorder, intensifying unem-
ployment and distress among the young immigrants who ¬‚ocked into the
bidonvilles of Algiers and other coastal cities from a countryside where pop-
ulation was growing rapidly and only 3 percent of land was cultivable. At the
same time, the Front de Lib´ ration Nationale™s (FLN™s) single-party regime
became increasingly corrupt and discredited as the liberation war receded into
the past. Following major urban riots in October 1988, President Chadli sought
to restore his authority by taking the risk “ rejected in Egypt and Tunisia “ of
In the time of AIDS 305

instituting multiparty democracy, perhaps hoping thereby to distance him-
self from the FLN™s unpopularity and to transfer to politicians the power that
military commanders had effectively held since independence. Among the ¬rst
parties formed, in February 1989, was the Front Islamique du Salut (FIS), organ-
ised by fundamentalists who had ¬rst appeared in Algeria during the 1960s and
had gained increasing in¬‚uence during the 1980s. Of their main leaders, one,
Abbasi Madani, had participated in the initial attack on French rule in 1954,
had then taken a doctorate at the Sorbonne before teaching at the university in
Algiers, and typi¬ed the sophisticated intellectuals heading most major fun-
damentalist movements, men whose aim was (in Yasin™s phrase) ˜to Islamise
modernity™ and whose appeal was especially to students seeking an ideological
rationale for traditional behaviour. The other main leader, Ali Belhadj “ also
from the desert edge where Donatism and Kharijism had ¬‚ourished “ was a
younger and less educated man whose position as imam of a Friday Mosque
in a working-class quarter of Algiers gave him in¬‚uence among the young
unemployed of the bidonvilles who provided fundamentalism with much of its
popular support.
Although FIS demonstrated its popularity in local government elections in
June 1990, Chadli apparently calculated that it could not win a parliamentary
election because other parties would unite against it in the second ballot. In
the event, when the election took place in December 1991, FIS did so well in the
¬rst ballot, winning over 40 percent of the votes, that the military intervened,
cancelled the election on the grounds that fundamentalists were seeking ˜to
use democracy in order to destroy democracy™, and banned the party. Militants
took to the hills, following a long Algerian tradition, and two ¬ercely hostile
armed movements emerged, the Arm´ e Islamique du Salut (AIS), which fought
a limited war for a return to constitutional processes, and the very brutal
Groupe Islamique Arm´ (GIA), which held that jihad was the only way to create
an Islamic state. Against them, hardline professional of¬cers insisted on total
victory while a succession of civilian presidents sought to negotiate a return to
normality. The AIS abandoned violence in 1999 and moderate fundamentalist
parties contested multiparty elections, although FIS itself was not relegalised.
Remnants of the GIA were still active in 2005. By then the war had cost an
estimated 150,000 lives.
One further upheaval matched democratisation and fundamentalism in his-
torical signi¬cance: the Ethiopian Revolution, which began in 1974 and took
a course distinctive to late twentieth-century Africa. Like the continent™s two
other genuine revolutions, in Rwanda and Zanzibar, Ethiopia™s both overthrew
and inherited the state structure of an ancien r´ gime divided ethnically from
most of its subjects “ in Ethiopia from the outlying non-Amharic peoples con-
quered during Menelik™s southward expansion in the late nineteenth century.
When Haile Selassie™s army mutinied in January 1974 over military grievances,
306 africans: the history of a continent

the old aristocracy had been too weakened by his autocracy to resist effectively,
while attempts to create a liberal constitution failed because Ethiopia™s unde-
veloped capitalism provided no bourgeoisie. Instead the initiative passed ¬rst
to students and trade unionists in Addis Ababa and then, when Ethiopia™s ter-
ritorial integrity seemed threatened, to a unitarist military faction, headed by
Major Mengistu Haile Mariam, which adopted Marxian language and relied
on communist aid. Mengistu™s regime failed to conquer or conciliate those
seeking secession for Eritrea and greater autonomy for Tigray and other outly-
ing regions. Legislation in 1975 destroyed landlordism and empowered peasant
associations to redistribute land, but actual redistribution was uneven and the
regime subsequently lost rural support by attempting to extract agricultural
surplus by techniques that alienated peasants everywhere in Africa: state farms,
of¬cial marketing, producer cooperatives, compulsory villagisation, and forced
resettlement in outlying regions, all amidst recurrent famine.
Weakened by the Soviet Union™s collapse, Mengistu was overthrown in
May 1991 by a coalition of regional guerrilla forces, the Ethiopian People™s
Revolutionary Democratic Front (EPRDF), led by former revolutionary
students. Eritrea seceded in 1993, leaving a disputed border over which the two
countries fought a pointless and costly war in 1998“2000. In Ethiopia a new con-
stitution recognised the strength of ethnic feeling by creating a federation with
much provincial autonomy and ˜an unconditional right to self-determination,
including the right to secession™, although in practice regionalism was used
as a means of incorporation and the EPRDF ruthlessly frustrated secessionist
movements. The federation™s long-term survival was one outstanding question.
The other was control of land, for the course of the revolution had destroyed
both the old landlord class and subsequent socialist experiments, thereby open-
ing the way to marketable freehold and the capitalism that some Ethiopians
thought essential to economic development. The EPRDF thought otherwise
and favoured a relatively egalitarian tenure by peasants as leaseholders from
the state. Scanty evidence suggested that most peasants agreed, but whether
they could defend it in practice remained unclear.
The EPRDF was distinguished from most other guerrilla movements in late
twentieth-century Africa by capturing rather than destroying a functioning
state. In Uganda and Rwanda, also, there were at least strong state traditions
that assisted victorious guerrillas to establish order. But three other long-drawn
civil wars of the period “ starting in Somalia in the late 1980s, Liberia in 1989, and
Sierra Leone in 1991 “ were almost entirely destructive. This was partly because
all began as rebellions against regimes that were sectional, oppressive, and yet
gravely weakened by economic decline. Once launched with political aims,
moreover, the rebellions fell into the hands of warlords who recruited youths
excluded from education and employment, armed them with weapons cheaply
available from Eastern Europe after communism collapsed, and ¬nanced their
In the time of AIDS 307

campaigns not with the Cold War subventions of the past but by exploiting
the natural resources within their base areas, notably Sierra Leone™s diamonds.
There and in Liberia foreign peacekeeping troops eventually restored state
control, but from 1991 Somalia returned to something approaching its pre-
colonial statelessness.
In this pattern of con¬‚ict, events in Rwanda and Burundi were unique, for in
each country two self-conscious ethnic groups, Hutu and Tutsi, occupied and
laid claim to a single territory, whereas elsewhere in Africa each ethnic group
possessed a separate rural location. In Rwanda, the Hutu revolution at the time
of independence in 1959“62 had destroyed the monarchy and driven thousands
of Tutsi into exile, many of them in Uganda. This made much land available
to the rapidly increasing Hutu population. In 1973 power was seized by a Hutu
faction, led by Juvenal Habyarimana, based in the northwestern provinces that
had been incorporated into the state only in the colonial period and were
especially hostile to the former Tutsi rulers. During the later 1980s and early
1990s, four developments destabilised this situation. First, population pressure
created new land scarcity and the economy suffered in the general depression of
the period, leaving many young men landless and unemployed. Second, Tutsi
exiles of the Rwanda Patriotic Front (RPF), many with military experience in
Uganda™s civil war of the 1980s, reinvaded Rwanda in 1990 and, although initially
unsuccessful, gradually penetrated the north. Third, Habyarimana was pressed
by Western donors and rival Hutu groups to democratise his regime, eventually
agreeing, by the Arusha Accords of 1993, to incorporate other Hutu and Tutsi
representatives and to give the RPF 40 percent representation in the army,
which meant discharging over twenty thousand Hutu soldiers and alienating
his hardline supporters.
The fourth destabilising in¬‚uence on Rwanda was neighbouring Burundi,
which had gained independence simultaneously but under a regime led by
moderate Tutsi. In 1965, inspired by the example of Hutu power in Rwanda,
Burundi™s Hutu extremists attempted a coup d™´ tat but were defeated by Tutsi
troops. Seven years later, a Hutu rebellion led the army to massacre perhaps
200,000 Hutu with the deliberate intention of destroying their potential leaders.
From the late 1980s, however, Burundi too came under pressure to democratise.
Its military leader, Pierre Buyoya, created a power-sharing government and
held multiparty elections in 1993, only to be defeated by a Hutu party that won
65 percent of the votes. Four months later, the Tutsi army murdered the new
Hutu prime minister. Many Hutu ¬‚ed to Rwanda.
There the refugees added to the growing paranoia and instability. Threatened
by RPF invasion and power-sharing, Hutu extremists in Habyarimana™s party
had begun in 1991 to convert the party™s youth wing into a militia known
as Interahamwe. By late 1993, some extremists contemplated a general mas-
sacre of Tutsi, but what precipitated genocide in April 1994 was the shooting
308 africans: the history of a continent

down of a plane carrying Habyarimana, probably by the RPF. Within hours,
Habyarimana™s presidential guard began the systematic killing of Tutsi and
moderate Hutu leaders in the capital. As news of the president™s death reached
the countryside, local military commanders and party leaders organised sim-
ilar massacres, mobilising the Interahamwe and urging or sometimes com-
pelling peasants to implicate themselves by murdering neighbouring Tutsi and
suspected Hutu sympathisers. During the next three months, perhaps 800,000
people were killed, often with deliberate cruelty, in an organised genocide
that drew on generations of inequality and exploitation, an ethnic animosity
crystallised by European racial theories and by extremist propaganda, a deep
popular fear of renewed Tutsi rule, and the desperate need of the poor for
Tutsi land and property. Once the RPF army grasped the situation it advanced
swiftly southwards, driving the Hutu regime and perhaps two million refugees
across the borders, opening Rwanda to over 600,000 returning Tutsi exiles, and
establishing a regime determined that the Tutsi people should never again face
Inevitably, the Rwandan genocide reacted back on Burundi, but the impact
was surprising and perhaps still uncertain a decade later. Following the Tutsi
military coup of 1993, Hutu forces launched a guerrilla war, Buyoya regained
power, and international mediators pressed both sides to reach a new settle-
ment. This was achieved in 2005 when over 90 percent of voters approved
a constitution creating a national assembly and government with 60 percent
Hutu and 40 percent Tutsi representation and an army with equal representa-
tion of each group. Six months later the leader of the main Hutu guerrilla force
was elected president. The relationship between this settlement and the totally
different regime in Rwanda left the future uncertain for both.
In the meantime the Rwandan genocide had also triggered the ¬rst war
involving numerous independent African states. In 1994 retreating Hutu troops
and Interahamwe from Rwanda established themselves across the border in
the Democratic Republic of the Congo, whose distant government in Kinshasa
had no capacity and perhaps little wish to control them. In 1996 RPF forces
entered the Congo to disperse and destroy them, very brutally, followed by
Ugandan units anxious to secure their own border. The invasion expanded
into an attempt to replace Mobutu™s failing regime in Kinshasa by a more


. 9
( 12)