Third World Quarterly, Vol. 32, No.

10, 2011, pp 1843–1869

Inclusive Elite Bargains and the

Dilemma of Unproductive Peace:
a Zambian case study
STEFAN LINDEMANN

ABSTRACT This article seeks to contribute to recent debates on the link

between political settlements and state building. It proposes a theoretical
framework that centres on the alternative concept of ‘elite bargain’ and suggests
that inclusive elite bargains can be expected to facilitate both peace and
economic development. Yet a detailed case study of elite bargains in Zambia
shows that all good things do not always go together. While inclusive elite
bargains have indeed helped to avoid civil war, they have often constrained
economic development—a dilemma of unproductive peace.

One of the main insights from the state-building literature is that political
and developmental outcomes are not predetermined by adequate institu-
tional design but instead depend on the underlying configuration of political
power. Against this backdrop the notion of ‘political settlement’ has become
a fashionable topic in the state-building debate. In the policy community
Alan Whaites has defined political settlements as ‘the deeper, often
unarticulated, understandings between elites’ that not only bring about the
conditions to end conflict but also facilitate more ‘responsive state-building’.1
Similarly the British Department for International Development (DFID)
considers political settlements as the ‘expression of a common understanding,
usually forged between elites, about how power is organised and exercised’,
which is key to establishing peaceful and viable states.2 In the academic
literature the idea of political settlement as the balance of power between
contending groups and classes is deeply rooted in the historical political
economy approach, evident in the work of Karl Marx and Barrington
Moore, among others.3 More recently North et al have argued that the
establishment of stable political order requires the formation of a ‘dominant
coalition’ at the level of elites (that is, a political settlement), which limits
access to economic rents to members and thereby creates credible incentives
to co-operate rather than fight.4

Stefan Lindemann is at the Max Planck Institute for the Study of Religious and Ethnic Diversity, Hermann-
Föge-Weg 11, 37073 Göttingen, Germany. Email: [email protected].

ISSN 0143-6597 print/ISSN 1360-2241 online/11/101843–27

Ó 2011 Southseries Inc., www.thirdworldquarterly.com
http://dx.doi.org/10.1080/01436597.2011.610585 1843
 STEFAN LINDEMANN

While the lens of political settlement holds great potential, its precise
meaning remains somewhat elusive. This is especially true for one particular
dimension of political settlements, namely their inclusiveness. Significantly
most contributions suggest that the inclusiveness of political settlements
matters a great deal for state building, yet remain rather unspecific about the
‘how and why’. DFID, for instance, maintains that ‘the inclusiveness of a
settlement, and public perceptions of its fairness, is critical to state legitimacy
and the sustainability of the settlement in the long term’.5 However, there is
little reflection on who needs to be included and with what effects for peace
and economic development. Similarly North et al’s otherwise theoretically
sophisticated work does not elaborate much on the inclusiveness of the
‘dominant coalition’. Even though they mention that the exclusion of
powerful elites from access to privileges may result in violent conflict,6 it does
not become clear whether the systematic inclusion of all relevant social
groups and their leaders is considered a prerequisite for the creation of
peaceful states. The notion of a ‘dominant coalition’, where access to rents is
limited to members, implies that others remain outside the coalition, which
then raises the question of why the latter would endure their marginalisation.
Even more problematically there is no reflection on how the inclusiveness of
the ‘dominant coalition’ may affect the process of economic development.
To better account for the role of inclusiveness and its effects on political
and developmental outcomes, this article first introduces a theoretical
framework that centres on the alternative concept of ‘elite bargain’, defined
as the distribution of positions of state power between representatives of
contending social groups. Building on existing research, the framework
suggests that inclusive elite bargains can be expected to facilitate both peace
and economic development. To further probe the plausibility of this
argument, the article then investigates the case of postcolonial Zambia.
The analysis shows that all good things do not always go together. While
inclusive elite bargains have indeed helped to avoid civil war, they have often
constrained economic development—a dilemma of unproductive peace. The
final section discusses the wider relevance of the findings and identifies areas
for future research.

Elite bargains and implications for peace and economic development

The term ‘elite bargain’ has long been used in a rather broad sense to describe
any kind of settlement between political elites, especially between government
and opposition.7 More recently scholars at the Crisis States Research Centre
(CSRC) proposed a deeper concept of elite bargain, defined as the ‘distribution
of rights and entitlements’ across groups and classes in society, on which any
state is based.8 This is a significant advance in that it recognises the
fundamental importance of elite bargaining in the process of state formation.
Drawing on this work, I adopt a slightly different approach and
conceptualise elite bargains as the distribution of positions of state power
between representatives of contending social groups, not least since those
who are in positions of state power ultimately determine the distribution of
1844
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

rights and entitlements. I suggest that elite bargains can be captured in two
ideal types that describe the extent to which rulers have used the distribution
of positions of state power to accommodate the dominant cleavages in society
(ethnicity, language, region, religion, class, etc). Inclusive elite bargains exist
where rulers provide representatives of contending social groups with
balanced access to positions of state power, while exclusionary elite bargains
involve a biased inter-group distribution of public appointments.
To measure the inclusiveness of the elite bargain, I focus on the extent to
which positions of political, military and economic power are shared between
members of competing social groups. Political power sharing is important in
that it provides group leaders with visible recognition, a ‘say’ in decision
making and control over resources. A first obvious indicator is the
composition of the government, which can be measured by the inter-group
distribution of the cabinet; deputy ministers; and the ‘inner core’ of political
power. The concept of the ‘inner core’ is useful to determine whether power
sharing extends to the positions of real power and influence. Other relevant
indicators are the composition of the ruling party, measured by the
distribution of positions in the top party organ, and the composition of
the civil service, measured by the distribution of permanent secretaries.
Military power sharing is crucial in that it shapes their feelings of physical
security and survival. While balanced recruitment at the level of the rank and
file may be considered important, it is especially representation at the upper
levels of the army that gives group leaders a real stake in the security sector.
Key, therefore, in terms of military power sharing is the composition of the
officer corps, measured by the inter-group distribution of the top command
positions and the higher ranks.
Economic power sharing is of immediate material interest for members of
competing groups. A first useful indicator may be control over key state-
owned enterprises, which are among the most lucrative public institutions in
patronage-based political systems. Over the past few decades the parastatal
sector has come under pressure, however, with far-reaching privatisation
programmes since the 1990s. A second relevant indicator may therefore be
control over key privatised companies. Both indicators can be measured by
the inter-group distribution of board directors and senior management
positions.
What are the consequences of inclusive versus exclusionary elite bargains?
I focus on the link between elite power sharing and two goals that are at the
very heart of the state-building agenda, namely peace and economic
development.
The relationship between elite bargains and peace seems to be fairly
straightforward. Wimmer et al’s recent work on ethno-power relations (EPR),
for example, shows that ethnic exclusion from executive-level state power is
strongly and robustly associated with civil war.9 Further evidence for the
peace-enhancing role of elite power sharing can be found in the literatures on
consociationalism,10 elite settlements,11 inclusive political coalitions,12 and
horizontal inequalities.13 In the post-conflict literature the empirical evidence
is more ambiguous. First, large-N studies come to contradictory conclusions
1845
 STEFAN LINDEMANN

on whether post-conflict power-sharing leads to lasting peace.14 Second, case

study evidence shows that experiments with post-conflict power sharing have
so far had mixed outcomes.15 While this indicates that power sharing may be
more difficult in post-war situations, most evidence suggests that the
prospects of peace and stability depend on the inclusiveness of the elite
bargain. Against this background I expect that inclusive elite bargains
accommodate the dominant social cleavages, stabilise the inter-group
competition over the control of the state and thereby favour civil war
avoidance. Exclusionary elite bargains, by contrast, should fail to accom-
modate dominant social cleavages, intensify inter-group struggles over the
distribution of state power and ultimately favour the onset of civil war.
The relationship between elite bargains and economic development is less
well understood. Even though there is abundant research on how ethnic
fractionalisation affects economic development,16 this literature relies on
demographic measures of ethnic diversity that tell us nothing about whether
power is shared between representatives of competing groups. In an attempt
to remedy this shortcoming, a recent quantitative study by Min et al uses the
above-mentioned EPR dataset to show that ethnic exclusion decreases long-
run economic growth rates.17 In line with this finding I expect that inclusive
elite bargains will favour economic development, whereas the opposite
should be true for more exclusionary elite bargains. The causal mechanisms
behind this assumed relationship can be both indirect and direct. First, a lack
of elite power sharing is likely to lead to violent conflict, which should in turn
undermine economic activity.18 By contrast, inclusive elite bargains favour
peace and stability and can therefore be expected to be more conducive to
economic development. Second, there is also reason to assume a more direct
relationship. Regimes based on an exclusionary elite bargain lack an
‘encompassing interest’ in society and therefore resemble Mancur Olson’s
‘roving bandit’,19 who prefers predation and consumption over public goods
provision and productive investment—a constellation that is prone to
economic decline. More broad-based regimes, by contrast, have a more
‘encompassing interest’ in society and thus assume the role of Olson’s
‘stationary bandit’, who engages in universal public goods provision and the
promotion of productive economic activities.
Altogether pre-existing research suggests that inclusive elite bargains will
be conducive to both peace and economic development. To further explore
the plausibility of this argument, this article investigates the (partially)
‘deviant’ case of Zambia—a country where inclusive elite bargains have been
accompanied by enduring peace (in line with previous findings on cross-case
relationships) but also by persistent economic stagnation (in contrast to
previous findings on cross-case relationships). The ‘deviant case method’ is
useful to understand theoretical and empirical anomalies and probe for new
hypotheses.20 As such, it helps to shed new light on the often complex
relationship between elite power-sharing and political and developmental
outcomes.
The analysis is based on a comprehensive set of original data on the
distribution of political, military and economic posts, which was put together
1846
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

during fieldwork in Zambia in 2008. To collect the data, I first compiled lists
of all ministers, deputy ministers, party leaders, permanent secretaries,
parastatal directors and army officers since independence. I then identified
the linguistic affiliation of every individual, since language is arguably the
most salient cleavage in Zambia (see below for details). In doing so, I relied
on help from many Zambians. For each name I made sure to get at least five
different opinions, preferably from people who knew the individual
personally, which produced very congruent answers. My dataset is
complemented by evidence from 53 interviews with a great variety of
stakeholders.

Zambia and the persistence of inclusive elite bargains

In pre-colonial times the territory of present-day Zambia contained many
ethnic groups, the largest ones being the Bemba in the North, the Ngoni and
Chewa in the East, the Tonga in the South and the Lozi in the West (see
Figure 1). Politically a few groups were organised in centralised kingdoms
(especially the Lozi), whereas most lived in decentralised societies.21 The
advent of colonial rule from the 1890s—first through the British South Africa

FIGURE 1. Ethnic, linguistic and provincial cleavages in Zambia.

1847
 STEFAN LINDEMANN

Company (BSAC) and later through the British colonial office—did little to
integrate these different groups. To ensure efficient taxation, the colonialists
ruled ‘indirectly’ through local chiefs whose position was greatly strength-
ened—a process that resulted in the ‘tribalisation’ of local authority
structures.22 Moreover, missionary activity, colonial education policies and
labour migration favoured the progressive emergence of four language
groups, including Bemba-, Nyanja-, Tonga-, and Barotse-speakers. A fifth,
albeit looser language group came to exist in the northwest, where no single
dominant regional lingua franca emerged but Lunda, Kaonde and Luvale
enjoyed first-among-equals status. As a result, national conflicts were
henceforth predominantly seen in linguistic terms, while local interactions
continued to be ethnically framed. The picture was further complicated by
the fact that the different groups were clustered into eight (later nine)
administrative provinces.
In addition, the growth of the mining industry on the Copperbelt from the
1930s involved rapid urbanisation and class formation. The salience of class
cleavages first became evident in the 1935 mineworkers strike on the
Copperbelt, which displayed a consciousness of common interest and the
capability to organise concerted resistance within the urban environment.23
Such class consciousness increased from the late 1940s, when a growing
emphasis on mechanisation forced the mining companies to stabilise African
mineworkers so that they could acquire and retain certain skills. Accordingly,
the African population on the Copperbelt grew from 175 747 in 1951 to 458
654 in 1963.24
The multi-ethnic Copperbelt gave rise to a broad-based nationalist
movement.25 African resistance gained strength from the mid-1940s and
came to be led by Kenneth Kaunda’s United National Independence Party
(UNIP). UNIP disposed of a powerful, mass-party organisation and
functioned as an important agent of integration among Africans. Never-
theless, its impact was uneven across the country. While the liberation
struggle was deeply rooted among Bemba-speakers in Copperbelt, Northern
and Luapula Provinces, other parts of the country remained largely
untouched. In Western Province, the Barotse Agreement with the British
had granted the Lozi Kingdom the status of a ‘protectorate within the
protectorate’ and thereby enabled the traditional government to maintain a
high degree of autonomy throughout the colonial period.26 The growth of
African nationalism was perceived as a threat by traditional leaders who
made repeated, albeit unsuccessful secessionist bids from the late 1950s.
Moreover, Harry Nkumbula’s African National Congress (ANC)—from
which UNIP had split in 1958—retained mass support among Tonga-
speakers in Southern and Central Provinces.27 Finally, UNIP faced
competition from the powerful African Mineworkers Union (AMWU),
which controlled the urban workforce on the Copperbelt and resisted its
subordination under the party. Altogether there was considerable potential
for conflict when UNIP led Zambia to independence in October 1964,
evident in pronounced ethnic, linguistic and—albeit to a lesser extent—class
cleavages.
1848
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

The UNIP elite bargain (1964–91)

Under the multiparty regime of the First Republic (1964–2) UNIP initially
ruled the country with a comfortable majority, albeit challenged by the
strength of the ANC in Southern Province and the emergence of the Lozi-
based United Party (UP) in Western Province. In 1968 UNIP won the general
elections but had to accept defeat to the ANC not only in the South but also—
after the banning of the UP—in the West. An even more significant challenge
to UNIP emerged in 1971 when the Bemba-based United Progressive Party
(UPP) challenged its authority in the North. As UNIP found itself on the verge
of becoming a regional party, it imposed a one-party state in December 1972.
During the Second Republic (1972–91), UNIP ‘won’ the general elections in
1973, 1978, 1983 and 1988, with considerable support for Kaunda even at the
height of economic crisis. However, the ruling party’s fortunes changed in the
early 1990s with growing pressure for a return to multipartyism and with a
humiliating electoral defeat to the Movement for Multiparty Democracy
(MMD) in 1991.
Between 1964 and 1991 UNIP made a sustained attempt to forge an
inclusive elite bargain. ‘One Zambia, One Nation’ being the overarching
motto, President Kaunda relied on a practice called ‘tribal balancing’,
whereby access to positions of state power was to be distributed equitably
among representatives of competing groups. This practixe was implemented
at all levels of the public sector, evident in high degrees of political, economic
and military power sharing.

Political power sharing. The UNIP governments during the First and Second
Republic were on average carefully balanced between members of the
country’s main language groups (see Table 1). The early postcolonial
governments were characterised by some imbalances (especially in favour of
Barotse-speakers), which were, however, soon redressed. From the late 1960s
the distribution closely matched the language groups’ respective population
share, with none of them being significantly underrepresented. Significantly
this proportional balance was achieved not only for ministers and deputies
but also for the more consequential positions in the ‘inner core’ of political
power. During the Second Republic government appointments continued to
reflect the country’s linguistic demography. Moreover, appointments were
strategically used to anchor the one-party state in former opposition areas,
most evident in the prominent representation of Tonga- and especially
Barotse-speakers.28
The quest for political power-sharing was mirrored in UNIP’s Central
Committee (CC)—the party’s top organ, which became the key decision-
making body under one-party rule.29 After independence Bemba-speakers
were somewhat underrepresented (see Table 2) and felt neglected relative to
their major contribution to the freedom struggle.30 These tensions culminated
in the upheavals surrounding the August 1967 UNIP elections, when Bemba-
speaking leaders used this occasion to join up with the weak Tonga-speaking
faction to unseat leading Lozi- and Nyanja-speaking leaders.31 To redress
1849
 TABLE 1. Distribution of government positions between language groups, 1964–90 (%)

1850
Language Group Population 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1990 Average

Ministers (Cabinet)
Bemba 40,9 25,0 33,3 27,8 33,3 36,0 41,7 38,5 35,7 40,0 40,9 41,7 43,5 45,0 37,1
Nyanja 22,0 18,8 20,0 16,7 23,8 16,0 12,5 15,4 21,4 20,0 18,2 16,7 13,0 25,0 18,3
Tonga 14,2 18,8 20,0 22,2 14,3 16,0 12,5 11,5 14,3 15,0 13,6 16,7 17,4 20,0 16,3
North-Western 9,1 6,3 6,7 5,6 14,3 16,0 12,5 15,4 14,3 15,0 13,6 12,5 8,7 5,0 11,2
Barotse 8,6 25,0 20,0 22,2 9,5 16,0 16,7 15,4 14,3 10,0 13,6 12,5 17,4 5,0 15,2
Other 5,2 6,3 0,0 5,6 4,8 0,0 4,2 3,8 0,0 0,0 0,0 0,0 0,0 0,0 1,9
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0
Deputy Ministers
Bemba 40,9 36,4 35,7 41,9 41,4 43,3 25,0 26,3 23,1 34,8 34,8 36,4 42,9 32,4 34,9
Nyanja 22,0 18,2 14,3 19,4 24,1 20,0 33,3 21,1 15,4 13,0 8,7 13,6 14,3 29,4 18,8
Tonga 14,2 13,6 14,3 9,7 10,3 10,0 25,0 26,3 38,5 17,4 17,4 22,7 23,8 14,7 18,7
North-Western 9,1 13,6 10,7 12,9 10,3 10,0 8,3 15,8 15,4 13,0 17,4 13,6 9,5 5,9 12,0
Barotse 8,6 9,1 14,3 9,7 10,3 10,0 8,3 10,5 7,7 21,7 21,7 13,6 9,5 17,6 12,6
Other 5,2 9,1 10,7 6,5 3,4 6,7 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 2,8
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0
‘Inner Core’
Bemba 40,9 30,0 30,0 36,4 42,9 27,3 45,5 36,4 38,5 45,5 38,5 42,9 38,5 50,0 38,6
STEFAN LINDEMANN

Nyanja 22,0 10,0 20,0 27,3 28,6 27,3 9,1 18,2 23,1 27,3 30,8 28,6 23,1 25,0 22,9
Tonga 14,2 20,0 20,0 9,1 14,3 27,3 27,3 27,3 15,4 9,1 7,7 7,1 15,4 8,3 16,0
North-Western 9,1 0,0 0,0 0,0 14,3 18,2 9,1 9,1 7,7 0,0 0,0 0,0 0,0 8,3 5,1
Barotse 8,6 30,0 30,0 18,2 0,0 0,0 0,0 0,0 15,4 18,2 23,1 21,4 23,1 8,3 14,4
Other 5,2 10,0 0,0 9,1 0,0 0,0 9,1 9,1 0,0 0,0 0,0 0,0 0,0 0,0 2,9
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0

Notes: Data for 1988 are missing. The ‘Inner core’ is defined as including the President, the Vice-President (Prime Minister) and the Ministers of Defence, Foreign
Affairs, Home Affairs, Finance, Development Planning, Mines, Agriculture, Works & Supply, Commerce & Industry, Local Government and Legal Affairs. The
category ‘Other’ refers to Zambians of White and Asian background.
Source: Author’s own data compiled and calculated based on Central Statistical Office (CSO), Census of Population and Housing 1969: Final Report, Vol I, Total
Zambia, Lusaka: CSO, 1973; CSO, 1980 Population and Housing Census of Zambia: Analytical Report, Vol III, Lusaka: CSO, 1985; and Government of Zambia (GOZ),
Official Verbatim Report of the Parliamentary Debates, Lusaka: GOZ, various years.
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

these imbalances, Kaunda disbanded the UNIP CC in 1969 and replaced it

with a more balanced Interim Executive Committee. Afterwards he made
sure that the UNIP CC remained remarkably reflective of the country’s main
language groups throughout the Second Republic (see Table 2). Further, the
UNIP CC was used to co-opt former opposition stalwarts. Key appointments
in this respect included those of Edward Liso (Nkumbula’s lieutenant) and
Nalumino Mundia (the former UP leader).
Top civil service appointments, finally, were also ‘tribally balanced’ (see
Table 3). Even though some groups were at times slightly better or worse
off, the overall picture was of balancing the claimants. Insiders confirm that
this was a deliberate policy meant to ensure that all groups would feel that
they had a substantial stake in the administrative running of the country.32
Moreover, the available evidence indicates that this balance at the level

TABLE 2. Distribution of the UNIP Central Committee positions between language

groups, 1964–88 (%)
Language
Group Population 1964 1967 1969 1973 1976 1979 1982 1985 1988 Average

Bemba 40,9 28,6 45,5 41,7 44,0 43,4 37,5 40,0 36,0 28,7 38,4
Nyanja 22,0 28,6 9,1 16,7 16,0 17,4 20,8 20,0 20,0 19,7 18,7
Tonga 14,2 14,3 27,3 25,0 16,0 17,4 16,7 16,0 20,0 22,7 19,5
North- 9,1 0,0 0,0 8,3 12,0 13,0 12,5 12,0 12,0 18,2 9,8
Western
Barotse 8,6 21,4 18,2 8,3 12,0 8,7 12,5 12,0 12,0 10,6 12,9
Other 5,2 7,1 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,0 0,8
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0

Source: Author’s own data compiled and calculated based on CSO, Census of Population and Housing
1969; CSO, 1980 Population and Housing census of Zambia; United National Independence Party (UNIP),
Annual Report, Lusaka: UNIP, various years; RL Rotberg, ‘Tribalism and politics in Zambia’, Africa
Report, 12(9), 1967, pp 29–35; and ‘KK scraps top UNIP committee’, Times of Zambia (Lusaka), 26 August
1969.

TABLE 3. Distribution of permanent secretary positions between language groups,

1964–85 (%)
Language Group Population 1964 1968 1972 1976 1980 1985 Average

Bemba 40,9 0,0 31,8 28,0 13,8 40,0 30,8 24,1

Nyanja 22,0 0,0 22,7 36,0 37,9 22,9 33,3 25,5
Tonga 14,2 0,0 9,1 12,0 13,8 14,3 17,9 11,2
North-Western 9,1 0,0 4,5 4,0 13,8 8,6 5,1 6,0
Barotse 8,6 6,3 18,2 20,0 17,2 11,4 12,8 14,3
Other 5,2 93,8 13,6 0,0 3,4 2,9 0,0 18,9
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0

Source: Author’s own data compiled and calculated based on CSO, Census of Population and Housing 1969;
CSO, 1980 Population and Housing Census of Zambia; and Government of Zambia (GOZ), Zambian
Government Gazette, Lusaka: GOZ, various years.

1851
 STEFAN LINDEMANN

of permanent secretaries was mirrored at the lower levels of the civil

service.33

Economic power sharing. Attempts at economic power sharing were initially

hampered by the fact that Zambia had very little control over its own
economy. As a consequence, Kaunda announced the ‘Mulungushi economic
reforms’ in April 1968, whereby the government took a 51 per cent
controlling share in 26 large industrial and commercial firms and used its
power to restrict certain economic opportunities, especially retail trading, to
Zambian citizens.34 The Industrial Development Corporation (INDECO) was
given vast new responsibilities and thus became, almost overnight, an
industrial corporate giant. In August 1969 the government took a 51 per cent
ownership share of the two copper-mining giants (Anglo American
Corporation and Roan Selection Trust), which were placed under the
Mining Development Corporation (MINDECO). INDECO and MINDECO were
made subsidiaries of a huge conglomerate, the Zambia Industrial and Mining
Corporation (ZIMCO). The political economy dimension of the nationalisa-
tion measures was all too obvious in that the reforms extended patronage
opportunities into the productive sectors. As shown in Table 4, these
opportunities were distributed in a remarkably equitable manner as members
of all language groups—with the partial exception of northwesterners—
received proportional shares of appointments to the boards of directors and
the senior management of the country’s three major parastatals.

Military power sharing. UNIP also promoted military power sharing.

Accordingly recruitment into the Zambian Defence Forces (ZDF) was based
on the principle of ‘tribal balancing’, not least since the ruling party
recognised that ethnic imbalances had played a major role in military coups
across Africa.35 At the level of the rank and file a newly introduced quota
system prescribed that army units were to be composed of soldiers from all
provinces and districts—a system that ensured that all ethnic groups had a
stake in the army.36 Similarly the officer corps was deliberately drawn from
all parts of the country. Accordingly members of all five language groups
enjoyed substantial representation at the very top of the military hierarchy
during the Second Republic (see Table 5).

The decline of the UNIP elite bargain. The Achilles’ heel of UNIP’s inclusive
elite bargain was clearly the failure to effectively accommodate class
cleavages, evident in the insufficient incorporation of the Copperbelt. This
problem went back to the breakaway of the UPP in the early 1970s that had
destroyed much of the ruling party’s support in what had been its stronghold
since the nationalist struggle. Even though UNIP had officially attempted to
‘reconcile’ with former UPP supporters, the party on the Copperbelt remained
deeply affected by suspicion.37 As a result, the Copperbelt failed to retrieve its
old prominence in party and government, which meant that the urban
population there was left without recognisable leadership of national stature.
In the context of the escalating economic crisis from the early 1980s, the trade
1852
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

TABLE 4. Distribution of parastatal appointments between language groups, 1970–

91
ZIMCO

Language Group Population 1974 1978 1982 1986 1990 Average

Bemba 40,9 50,0 47,1 42,1 39,4 41,9 44,1

Nyanja 22,0 27,3 11,8 10,5 24,2 29,0 20,6
Tonga 14,2 4,5 17,6 15,8 18,2 6,5 12,5
North-Western 9,1 9,1 0,0 0,0 0,0 6,5 3,1
Barotse 8,6 9,1 17,6 26,3 15,2 9,7 15,6
Other 5,2 0,0 5,9 5,3 3,0 6,5 4,1
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0
INDECO

Language Group Population 1975 1979 1983 1987 1991 Average

Bemba 40,9 28,6 54,5 33,3 33,3 25,0 35,0

Nyanja 22,0 35,7 9,1 33,3 25,0 25,0 25,6
Tonga 14,2 0,0 0,0 8,3 16,7 25,0 10,0
North-Western 9,1 7,1 0,0 0,0 0,0 16,7 4,8
Barotse 8,6 14,3 18,2 8,3 8,3 0,0 9,8
Other 5,2 14,3 18,2 16,7 16,7 8,3 14,8
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0
MINDECO ZCCM

Language Group Population 1970 1974 1982 1986 1990 Average

Bemba 40,9 50,0 36,4 28,6 28,6 42,9 37,3

Nyanja 22,0 12,5 45,5 14,3 14,3 0,0 17,3
Tonga 14,2 12,5 0,0 28,6 42,9 42,9 25,4
North-Western 9,1 0,0 0,0 0,0 14,3 0,0 2,9
Barotse 8,6 0,0 9,1 28,6 0,0 0,0 7,5
Other 5,2 25,0 9,1 0,0 0,0 14,3 9,7
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0

Source: Author’s own data compiled and calculated based on CSO, Census of Population and Housing 1969;
CSO, 1980 Population and Housing Census of Zambia; Zambia Industrial and Mining Corporation
(ZIMCO), Annual Report, Lusaka: ZIMCO, various years; Industrial Development Corporation (INDECO),
Annual Report, Lusaka: INDECO, various years; Mining Development Corporation (MINDECO), MINDECO
Mining Yearbook, Lusaka: MINDECO, various years; and Zambia Consolidated Copper Mines (ZCCM),
Annual Report, Lusaka: ZCCM, various years.

unions began to fill the political void on the Copperbelt and ultimately
became the main driver behind the opposition’s rise to power.38 The unions
received support from businessmen, churchmen and intellectuals who were
disillusioned by the combined effects of one-party dominance and economic
crisis.39 Altogether UNIP came to face an opposition movement that was very
similar to itself—an inclusive coalition that drew together various organised
interests and representatives from all parts of the country. The emergence of
broad-based opposition centred on the trade unions was only the logical
consequence of an elite bargain that was non-sectarian in nature but had only
shallow control over the urban areas.
1853
 STEFAN LINDEMANN

TABLE 5. Distribution of army commanders between language groups, 1976–91

SECOND REPUBLIC
Unified command structure (1976)
Zambia National Defence Force
Gen. G. K. Chinkuli (Tonga)
Lt.-Gen. P. D. Zuze (Nyanja)
Lt.-Gen. B. N. Mibenge (Bemba)
De-unified command structure (1980)

Air Force Army National Service

Maj.-Gen. C. Kabwe (Bemba) Gen. M. N. Masheke (Barotse) Brig.-Gen. C. J. Nyirenda (Nyanja)

Lt. Gen. A. Lungu (Nyanja) Lt.-Gen. C. S. Tembo (Nyanja) Brig.-Gen. F. S. Mulenga (Bemba)
Maj.-Gen. Simbule (Bemba) Lt.-Gen. G. M. Kalenge Maj.-Gen. T. Fara (Nyanja)
(North-Western)
Lt.-Gen. Simutowe (Bemba) Lt.-Gen. F. G. Sibamba
(Barotse)

Source: Author’s own data based on P Wele, Zambia’s Most Famous Dissidents: From Mushala to
Luchembe, Solwezi: PMW, p 158; interview with General M Masheke, former army commander and UNIP
prime minister, Lusaka, 7 August 2008; interview with G Haantolobo, deputy chief research officer,
National Assembly, Lusaka, 31 July 2008.

The MMD elite bargain (1991–2008)

The Third Republic did not revolutionise Zambia’s political culture. In 1991
the former trade unionist Frederick Chiluba became president and his MMD
won a large parliamentary majority, losing only in Eastern Province, which
remained a UNIP stronghold.40 In 1996—benefitting from UNIP’s electoral
boycott—MMD even managed to win a majority throughout the entire
country, whereby multiparty politics remained more formal than real.41 The
situation changed in 2001 when the new MMD candidate, Levy Mwanawasa,
hampered by Chiluba’s unconstitutional bid for a ‘third term’, won the
presidency by a narrow margin and even failed to gain a majority in
Parliament.42 However, MMD’s position was again strengthened in 2006 when
Mwanawasa won by a larger margin and recaptured a solid parliamentary
majority.43 Mwanawasa’s second term was abruptly ended by his death in
July 2008. He was replaced by Vice-President Rupiah Banda, who won the
presidential poll in October 2008.44
Over time MMD was consistently the only political party with considerable
political support in all parts of the country. Significantly both Chiluba and
Mwanawasa continued to make a UNIP-style attempt to build support by
forming a ‘maximum coalition’,45 which came to be reflected in high degrees
of political, economic and military power sharing.

Political power sharing. Officially the Chiluba government rejected the

convention of ‘tribal balancing’, arguing that such an approach would be
undemocratic and economically harmful.46 Moreover, it was commonly
accused of a ‘Bemba bias’ in appointments, typically related to the fact that
1854
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

Chiluba himself is a Bemba-speaking Lunda from Luapula Province.47 Yet

my data suggest that ‘tribal balancing’ was informally retained, albeit to a
lesser extent than under Kaunda (see Table 6). While Bemba-speakers were
only marginally overrepresented between 1992 and 1998, all other language
groups received relatively proportional representation (even though some of
them were at times slightly better or worse off). Only Nyanja-speakers
were underrepresented, which reflected the fact that UNIP had won all
parliamentary seats in Eastern Province in 1991. Nonetheless, Chiluba used
his right of appointment to bring Nyanja-speakers on board, which ensured
that they were always prominently represented in the ‘inner core’ of political
power.48
After Mwanawasa took over in 2001, the need for a balanced government
continued to be recognised (see Table 6). Initially this was difficult since MMD
had almost no MPs outside its strongholds in the north. As a consequence, the

TABLE 6. Distribution of government positions between language groups, 1992–2008

(%)
Language Average Average
Group Population 1992 1994 1996 1998 Chiluba 2002 2004 2006 2008 Mwanawasa

Ministers (Cabinet)
Bemba 41,7 40,0 40,0 37,5 42,3 40,0 54,5 55,0 33,3 32,0 43,7
Nyanja 23,8 16,0 12,0 12,5 11,5 13,0 9,1 5,0 16,7 12,0 10,7
Tonga 13,9 16,0 16,0 12,5 23,1 16,9 13,6 25,0 20,8 20,0 19,9
North-Western 7,7 8,0 8,0 8,3 7,7 8,0 13,6 5,0 20,8 20,0 14,9
Barotse 6,9 12,0 16,0 16,7 11,5 14,1 9,1 5,0 8,3 12,0 8,6
Other 6,0 8,0 8,0 12,5 3,8 8,1 0,0 5,0 0,0 4,0 2,3
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0
Deputy Ministers
Bemba 41,7 54,8 41,5 46,5 53,7 49,1 65,7 53,7 30,6 29,7 44,9
Nyanja 23,8 3,2 12,2 11,6 7,3 8,6 5,7 12,2 25,0 29,7 18,2
Tonga 13,9 22,6 24,4 25,6 22,0 23,6 5,7 9,8 8,3 5,4 7,3
North-Western 7,7 3,2 9,8 7,0 9,8 7,4 17,1 17,1 13,9 16,2 16,1
Barotse 6,9 6,5 9,8 9,3 7,3 8,2 5,7 7,3 22,2 16,2 12,9
Other 6,0 9,7 2,4 0,0 0,0 3,0 0,0 0,0 0,0 2,7 0,7
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0
‘Inner Core’
Bemba 41,7 46,2 41,7 45,5 50,0 45,8 40,0 40,0 36,4 36,4 38,2
Nyanja 23,8 23,1 8,3 18,2 8,3 14,5 0,0 0,0 9,1 9,1 4,5
Tonga 13,9 15,4 16,7 9,1 25,0 16,5 30,0 40,0 36,4 36,4 35,7
North-Western 7,7 7,7 16,7 0,0 8,3 8,2 10,0 0,0 9,1 18,2 9,3
Barotse 6,9 0,0 0,0 9,1 8,3 4,4 20,0 10,0 9,1 0,0 9,8
Other 6,0 7,7 16,7 18,2 0,0 10,6 0,0 10,0 0,0 0,0 2,5
Total 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0

Notes: Data for 2000 are missing. The ‘Inner core’ is defined as including the president, the vice-president,
the minister without portfolio and the ministers of defence, foreign affairs, home affairs, finance, trade &
industry, mines & mineral development, agriculture, local government and legal affairs.
Source: Author’s own data compiled and calculated based on CSO, Zambia 2000 Census Report, Vol 10,
Lusaka: CSO; and GOZ, Official Verbatim Report of the Parliamentary Debates.

1855
 STEFAN LINDEMANN

president chose to co-opt several prominent opposition MPs and appointed

them as ministers or deputy ministers.49 Moreover, the president used his
appointment prerogatives to nominate prominent individuals from opposi-
tion areas as ministers. This conciliatory attitude ensured that members of
the Tonga, Barotse and northwestern groups were prominently represented.
Nyanja-speakers, by contrast, remained underrepresented. To rebuild
support in the east, the president appointed Rupiah Banda, a Chewa, who
successfully campaigned for MMD before the 2006 elections and was
subsequently made vice-president. The re-emergence of MMD as a ‘national’
party in 2006 facilitated the establishment of a more balanced government.
A broadly national outlook was also retained in the ruling party and in the
civil service. While MMD as a ruling party was initially dominated by Bemba-
speakers,50 this bias has clearly eased over time (see Table 7). In the civil
service, permanent secretary appointments continued to be distributed more
or less proportionally (see Table 8).

TABLE 7. Distribution of the MMD National Executive Committee positions between

language groups, 2001–08 (%)
Language Group Population 2001 2008 Average

Bemba 41,7 50,0 42,1 46,1

Nyanja 23,8 12,0 14,0 13,0
Tonga 13,9 16,0 21,1 18,6
North-Western 7,7 12,0 12,3 12,2
Barotse 6,9 9,0 8,8 8,9
Other 6,0 1,0 1,8 1,4
Total 100,0 100,0 100,0 100,0

Source: Compiled and calculated based on CSO, Zambia 2000 Census Report; K Hulterstrom, In Pursuit of
Ethnic Politics: Voters, Parties and Policies in Kenya and Zambia, Uppsala: Uppsala University, 2004,
p 103; and Movement for Multiparty Democracy (MMD), ‘National Executive Committee Members’,
mimeo, Lusaka, 2008.

TABLE 8. Distribution of permanent secretary positions between language groups,

1992–08 (%)
Language Group Population 1992 1998 2002 2008 Average

Bemba 41,7 41,0 45,0 47,5 22,9 39,1

Nyanja 23,8 20,6 15,0 19,0 22,9 19,4
Tonga 13,9 23,1 20,0 12,2 22,9 19,5
North-Western 7,7 7,7 7,5 9,5 11,4 9,0
Barotse 6,9 7,7 10,0 11,9 20,0 12,4
Other 6,0 0,0 2,5 0,0 0,0 0,6
Total 100,0 100,0 100,0 100,0 100,0 100,0

Source: Author’s own data based on CSO, Zambia 2000 Census Report; GOZ, Zambian Government Gazette,
various years; and GOZ, ‘List of Permanent Secretaries’, mimeo, Lusaka, various years.

1856
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

Economic power sharing. Zambia has undergone comprehensive liberal-

isation since 1991.51 A key aspect of economic reform was the privatisation of
the bloated parastatal sector, in which 262 companies out of a total working
portfolio of 285 were divested between 1993 and 2006.52 Interestingly
privatisation opened new avenues for rent-seeking and increased rather than
decreased the scale of corruption.53 Accordingly officeholders in politics and
administration used the opaque privatisation process to acquire public assets
cheaply, especially small-scale companies. The main beneficiaries were
Chiluba and his friends and relatives, many of whom were later targeted
during Mwanawasa’s anti-corruption drive.54 Yet, while some claim that
Bemba-speakers benefited disproportionately, my data indicate that the
spoils of privatisation were distributed relatively equitably (see Table 9). In
fact, Bemba-speakers were even underrepresented in terms of their
population share. The only clearly discernible bias was in favour of
Zambians of Asian background (represented under ‘Other’). This confirms
public perceptions that privatisation really benefited foreign investors and
‘non-Zambian’ citizens.

Military power sharing. In the army MMD retained UNIP’s tradition of

inclusive recruitment and appointment policies. According to Mbita Chitala,
member of the Defence Council under Chiluba, ‘tribal balancing’ was
continued at all levels of the armed forces to ensure ‘that the whole country is
represented’.55 Such claims are confirmed by the fact that the top command
positions in the army were still distributed fairly equitably among members
of the country’s major language groups (see Table 10).

Cracks in the MMD elite bargain. Signs of continuity notwithstanding, the

MMD elite bargain over time came to suffer from a number of cracks. First,
efforts at ‘tribal balancing’ were at least partially contradicted by the
existence of the so-called ‘family tree’, evident in the prominence of
Mwanawasa’s Lamba and especially Lenje relatives in government and the

TABLE 9. Distribution of privatised companies between language groups, 1993–06

(%)
Companies sold
Language Companies sold to Companies sold to to Zambian
Group Population political leaders Zambian individuals corporate bodies Average

Bemba 41,7 16,7 52,6 32,1 33,8

Nyanja 23,8 5,6 21,1 7,1 11,3
Tonga 13,9 27,8 0,0 10,7 12,8
North-Western 7,7 22,2 0,0 3,6 8,6
Barotse 6,9 22,2 10,5 10,7 14,5
Other 6,0 5,6 15,8 35,7 19,0
Total 100,0 100,0 100,0 100,0 100,0

Source: Author’s own data based on CSO, Zambia 2000 Census Report; and Zambia Privatisation Agency
(ZPA), Privatisation Transaction Summary Sheets, 1992–2005, Lusaka: ZPA, 2006.

1857
 STEFAN LINDEMANN

TABLE 10. Distribution of army commanders between language groups, 1991–2008

THIRD REPUBLIC

Air Force Army National Service

Lt. –Gen. R. Shikapwasha (Tonga) Gen. N. M. Simbeye (Bemba) Lt.-Gen. W. G. Funjika

(North-Western)
Lt.-Gen. S. Kayumba Lt.-Gen. S. L. Mumbi (Nyanja) Maj.-Gen. M. Mbao
(North-Western) (Nyanja)
Lt.-Gen. Ch. Singogo (Bemba) Lt.-Gen. G. R. Musengule (Bemba)
Lt. Gen. S. Mapala (Nyanja) Lt.-Gen. I. Chisuzi (Tonga) Maj.-Gen. R. Chisheta
(Bemba)

Source: Author’s own data based on Wele, Zambia’s Most Famous Dissidents, p 158; and interviews with
Gen M Masheke and G Haantolobo.

administration.56 Second, MMD failed to rebuild political support in the

South, which meant that the United Party for National Development (UPND)
became entrenched as a regional party. Third, Mwanawasa did not only
progressively fall out with the Bemba-speaking bloc but also lost political
control over the urban areas. This has favoured the rise of Michael Sata’s
Patriotic Front (PF)—the strongest opposition party in 2006, which expresses
both Bemba grievances and urban discontent on the Copperbelt and in
Lusaka.57

Zambia’s inclusive elite bargains and implications for peace

and economic development
Existing evidence suggests that inclusive elite bargains favour both peace and
economic development. Yet, as I show below, the Zambian case draws
attention to the fact that elite power sharing may at times have contradictory
effects. While inclusive elite bargains have indeed fostered enduring peace,
they have often conflicted with economic development.

Inclusive elite bargains and peace

Zambia has always remained an oasis of peace and stability, with no instance
of civil war since independence in 1964.58 I argue that civil war avoidance in
Zambia can be traced back to the persistence of inclusive elite bargains. This
can be shown when looking at several ‘moments of crisis’ in the country’s
postcolonial history.
A first crisis occurred after the 1967 UNIP CC elections when the victory of
the Bemba-speaking bloc and the ensuing imbalances in party and
government produced widespread fears of Bemba domination throughout
much of the country. This was arguably the most serious political crisis in
Zambia’s history and threatened to tear apart the still fragile unity of ‘One
Zambia, One Nation’. These intense tensions were, however, defused by the
subsequent appointment of a more balanced Interim Executive Committee
1858
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

and reinforced attempts to increase the national outlook of government and

administration. If the controversial 1967 UNIP CC had not been dissolved, this
would almost certainly have caused lasting alienation among the Nyanja-
and Barotse-speaking factions that might ultimately have turned violent.
A second moment of crisis was the imposition of the one-party state in
1972, which involved the banning of all opposition parties—most notably of
the UPP—and therefore caused considerable frustration and disaffection. If
the introduction of one-party rule had been followed by systematic
discrimination against the core constituencies of the banned opposition
parties, the likely outcome would have been violent conflict. However, this
was never the case. While the one-party state was undoubtedly autocratic in
nature and did not hesitate to victimise those individuals who challenged its
authority, there is no evidence that representatives of particular groups were
ever systematically discriminated against. As detailed above, Bemba-, Tonga-
and Barotse-speakers, who had previously voted for opposition parties,
continued to be extremely prominently represented at all levels of the public
sector and therefore did not develop cohesive group grievances.
A third moment of crisis took place in the mid-1970s, when Adamson
Mushala and his followers tried to launch an armed rebellion in North-
western Province, claiming to fight the one-party state and the margin-
alisation of the province.59 Even though the Mushala rebellion temporarily
destabilised parts of the province, it ultimately failed to gain substantial
support and never came close to qualifying as a civil war.60 The main reason
for this failure was arguably that the mainstream of the northwestern
leadership was fully integrated in the country’s elite bargain throughout the
Second Republic. This was not only true for northwesterners as a whole but
also for members of all three ethnic groups in the province, including the
Kaonde, Lunda and Luvale. Of particular interest in this context is the
position of the Lunda leadership—Mushala and most of his followers were
Lunda—who were prominently represented not only in government but also
in other parts of the public sector.61 If the Lunda leadership had been
marginalised, the insurgency might have been more successful. Yet as this
was not the case, Mushala remained a relatively isolated figure who proved
unable to mobilise significant support until he was killed in 1982.
A fourth serious crisis occurred during the 1980s, when Zambia was hit by
extreme economic decline and mounting social unrest. This situation
ultimately culminated in a series of urban riots, which reflected UNIP’s
above-discussed failure to effectively accommodate class cleavages. At the
same time, however, the worsening economic crisis was not—as happened in
other African countries—accompanied by a biased distribution of the
dwindling ‘national cake’, which could have caused a dangerous, destabilis-
ing ‘ethnicisation’ of political competition. Instead, the country witnessed an
equitable sharing of losses, evident in the fact that elite politics remained
inclusive even at the height of economic crisis. According to the political
veteran Vernon Mwaanga, ‘even at the time when resources were extremely
limited, an attempt was made to equitably share jobs and resources among

1859
 STEFAN LINDEMANN

the various competing interests’.62 This arguably prevented the country’s

disintegration into violent group conflict.
A fifth crisis took place during and after the transition to multiparty
politics in 1991. Protagonists generally agree that this transition was peaceful
mainly because the MMD was a broad-based movement whose leadership
cultivated a national image through the trade unions.63 If MMD had come to
be dominated by members of one group, alienation and conflict would have
been a plausible scenario. This can be shown with the example of the Barotse-
speaking leaders. The latter had originally played an important role within
MMD, evident in the prominence of Arthur Wina or Aka Lewanika.
However, after Chiluba captured power, Barotse-speaking leaders were
temporarily sidelined in party and government, which culminated not only in
the creation of the Lozi-based National Party (NP) but also in renewed claims
for the restoration of regional autonomy under the Barotseland Agreement.64
In mid-1993 representatives of the Barotse Royal Establishment even made
open secessionist demands, which led to a dangerous standoff with the
Chiluba government.65 Two years later secessionist demands in Barotseland
resurfaced over the passage in parliament of a bill that withdrew the power to
allocate land from traditional rulers and transferred it to the president.66
Tellingly protests eased with the improved representation of Barotse-
speakers from the mid-1990s and the non-implementation of the land bill.
A sixth moment of crisis was Chiluba’s unsuccessful attempt to change the
Constitution in order to gain a third term, which caused considerable conflict
within the ruling party and ultimately resulted in the MMD’s loss of a
parliamentary majority during the 2001 elections.67 Even though communal
sentiments were generally of rather minor importance during the third term
controversy, the broader debate over Chiluba’s succession did involve
sectarian connotations in that there was a widespread perception of Bemba
dominance in the southwestern parts of the country, which found expression
in the growing strength of the UPND. If the president had picked a Bemba-
speaker as his successor, the succession crisis might have led to lasting
alienation in other parts of the country. Instead, the choice of Mwanawasa—
a Tonga-speaker who subsequently tried to rebuild MMD support in the non-
Bemba-speaking areas—was another shrewd choice of ‘tribal balancing’,
which once again undermined the consolidation of cohesive group
grievances.
A final moment of crisis occurred after Mwanawasa’s unexpected death in
mid-2008 when the ruling party came to be characterised by intense factional
struggles. According to political insiders, there were three main factions
within MMD at the time.68 One faction was led by Katele Kalumba, the MMD
National Secretary, who not only enjoyed strong support among the lower
party cadres but also maintained close links with the followers of the former
President Chiluba. A second faction included influential members of the
‘family tree’ who supported the candidature of Ng’andu Magande, the
minister of finance, but had limited support within the party. A third faction
grouped around Rupiah Banda, the vice-president, who was, however, widely
regarded as an ‘outsider’ within the party. The choice of Kalumba as the
1860
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

MMD presidential candidate would have angered the non-Bemba, whereas a

‘family tree’ candidate might have further alienated the Bemba bloc. As a
result, the former UNIP member and party outsider Banda was elected as
MMD candidate and subsequently replaced Mwanawasa. This can be
interpreted as another example of ‘tribal balancing’. On the one hand, the
election of Banda ended the relative underrepresentation of Nyanja-speakers
within the MMD. On the other hand, Banda is said—through Kalumba as his
‘kingmaker’—to enjoy close connections with Chiluba’s followers which may
help to reintegrate parts of the Bemba-speaking bloc, albeit at the expense of
Mwanawasa’s anti-corruption drive.69
Altogether it seems safe to argue that inclusive elite bargains have
facilitated Zambia’s lasting peace. This assessment is generally shared by
political insiders. When asked about the drivers of civil war avoidance, the
overwhelming majority of my interviewees made immediate reference to the
practice of ‘tribal balancing’. In the words of Dipak Patel, formerly a key
MMD minister:

It is not possible to govern this country if you don’t deal with balancing. The
day you remove balancing, the country will disintegrate into civil war.70

Admittedly such statements may be somewhat exaggerated. Decade-long

efforts to promote national unity have undoubtedly left deep traces in the
political culture of a country where open ethnic campaigning is widely
considered a ‘no go’. Accordingly temporary imbalances in public appoint-
ments are unlikely to become an immediate trigger of violent conflict.
Nevertheless, the above-mentioned cracks in the MMD elite bargain—in
particular the strength of PF and UPND—show that ethnicity remains a salient
social cleavage. Against this backdrop it should be kept in mind that the
political culture of ‘One Zambia, One Nation’ is a result of deliberate
political action and can be undone if such an approach is no longer deemed
necessary.

Inclusive elite bargains and economic development

Upon independence Zambia was a middle-income country believed to have
considerable economic potential. Yet, after a brief boom in the late 1960s, the
Zambian economy was hit by a number of external shocks.71 Oil prices
increased threefold, while the reduced demand for copper resulted in a price
drop of 40 per cent in 1975. The shocks led to rising trade deficits, escalating
inflation and mounting foreign debt. In 1978 Zambia entered into a conflict-
ridden relationship with the IMF, but subsequent rounds of structural
adjustment could not halt the economic decline. As a consequence, the
country experienced below African average growth rates and one of the worst
declines in per capita income in the history of postcolonial Africa (see
Figure 2). After 1991 the MMD government engaged in a new round of
structural adjustment. Yet economic performance continued to be dismal,
evident in low economic growth and declining GDP per capita throughout the
1861
 STEFAN LINDEMANN

1990s. It is only since the early 2000s that Zambia’s economy has showed
signs of slow recovery.
There are many well known reasons for Zambia’s poor economic
performance, the most important one being low and erratic world market
prices for copper, which is the backbone of the country’s economy.72
However, I argue that another crucial, yet often overlooked reason for
economic decline has been the political imperative of forging and
maintaining inclusive elite bargains, which has resulted in a bloated,
economically burdensome public sector.
The trade-off between political and economic objectives became evident
immediately after independence. As political power sharing between
competing interests required a sufficient reservoir of patronage, the number
of posts available for distribution had to be increased. Accordingly the size of
government increased considerably throughout the First Republic, with the
total number of ministers and deputies growing from 38 in 1964 to 55 in 1972
(see Figure 3). This inflated government was clearly beyond any functional
necessity and involved significant economic costs. Even more costly was
Kaunda’s practice of frequent cabinet reshuffles, the so-called ‘ministerial
game of musical chairs’, which was typically used to accommodate political
challengers and was often in contradiction to economic objectives.73 A case in
point would be Simon Kapwepwe’s appointment as vice-president with
responsibility for development planning—an appointment that followed the
electoral victory of Bemba-speakers during the 1967 UNIP CC elections. The
fact the Kapwepwe, the leader of the Bemba-speaking bloc, lacked country-
wide acceptance had a negative effect on his developmental role and
hampered the implementation of the First National Development Plan.74
More generally sectarian struggles within UNIP during the late 1960s and
early 1970s meant that the implementation of the Development Plan became
subordinate to the political need to maintain party unity.
During the Second Republic the country continued to suffer from the
burden of a bloated and inefficient government—a situation that still
reflected the political imperative of maintaining an inclusive elite bargain.
Even though the size of government was initially reduced to 36 posts in 1974,
it subsequently crept back up to 54 in 1990 (see Figure 3). In addition, the
frequent reshuffling of portfolios between ministers continued to disrupt
policy making.75 At the same time, the coexistence of the UNIP CC and
cabinet involved an economically wasteful duplication of positions. A Report
of the Special Parliamentary Select Committee (1977) placed the responsi-
bility for Zambia’s worsening economic crisis squarely on the country’s
‘political superstructures’, which were seen as a ‘serious constraint on
efficiency’.76 Similarly the need to accommodate ‘the fluctuating pressures of
strategic provincial and other interests’ undermined the performance of the
country’s civil service.77 As permanent secretaries and other top personnel
were frequently reshuffled, the development of expertise and effective
working hierarchies were negatively affected. At the same time, the civil
service continued to grow in size and was politically protected even during
the economic crisis of the late 1980s, when a report by the UNIP CC blamed
1862
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

FIGURE 2. Zambia’s economic performance, 1964–2008. Source: Compiled and

calculated based on data by Angus Maddison, at http://www.ggdc.net/maddison/.

the government for having failed to decisively cut the ineffective and
inefficient state bureaucracy.78
Perhaps most importantly the negative economic implications of Zambia’s
inclusive elite bargain also became evident in the parastatal sector. As
discussed above, economic power sharing was only made possible through
the comprehensive nationalisation measures of the late 1960s, which
extended patronage opportunities into the productive sector and allowed
for the distribution of rewards outside state and party jobs. While these
reforms clearly helped to contain the factional struggles of the time, they also
created a bloated parastatal sector that turned out to be highly inefficient and
1863
 STEFAN LINDEMANN

FIGURE 3. Size of government in Zambia, 1964–2008. Source: Author’s own data,

based on Government of Zambia (GOZ), Official Verbatim Report of the
Parliamentary Debates, Lusaka: GOZ, various years.

ultimately imposed huge economic costs on the country.79 Particularly

damaging in economic terms was the nationalisation of the copper industry
from the late 1960s.80 As the newly created mining parastatals suffered from
a lack of skilled domestic workers and managers, they came to be
characterised by high levels of economic mismanagement. In the end the
politically motivated state interventionism in the mining sector became a key
driver behind the dramatic decline of copper exports. As such, it seems
plausible to argue that the decline of copper revenue was not only the result
of low world market prices but at least partially endogenous to the political
imperative of elite power sharing.
The primacy of political over economic concerns remained in place after
the political transition in 1991. Significantly the size of the MMD governments
even exceeded the size of the Kaunda governments, especially under Chiluba.
The total number of ministers and deputies grew from 54 in 1990 to 67 in
1998 and was still at 61 in 2008 (see Figure 3). This growth ran counter to
functional necessity and economic rationality, in particular the high number
of deputy ministers, and was solely a response to the political need to
preserve a ‘tribally balanced’ government. Moreover, the MMD governments
struggled hard to avoid cuts in the size of the civil service.81 This indicates
that sufficient scope for patronage was still privileged at the expense of
economic considerations. While the civil service as an important bastion of
the elite bargain was protected, expenditure cuts were implemented in public
investment and social spending, with obvious implications for economic
development.
In the parastatal sector rapid progress in privatisation since the mid-1990s
has reduced the scope for patronage distribution—a change that could have
1864
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

opened the way to increased economic efficiency. Yet the process of economic
reform has again been subverted by the political imperative of elite
accommodation, which is reminiscent of Nicolas van de Walle’s argument
about the ‘partial reform’ syndrome in Africa.82 As discussed above,
privatisation and deregulation were used to create new opportunities for
accumulation and corruption. While this new form of economic power
sharing has helped to pay off and co-opt potential challengers, the envisaged
positive economic effects of the reforms hardly materialised. Instead of
leading to productive primitive accumulation, privatised companies were in
many cases simply stripped of their assets rather than recapitalised or fully
relocated to other countries in the Southern African region, which caused an
enormous loss of wealth.83
On the whole there is reason to argue that Zambia’s inclusive elite bargains
have often been detrimental to developmental outcomes. Elite power sharing
has followed a logic of politics, not of economic efficiency—a situation that
has put serious constraints on the country’s economic performance. Does this
in turn mean that a more exclusionary elite bargain would have led to greater
economic development? Yes and no. On the one hand, less attention to the
political imperative of inclusiveness might have prevented the emergence of a
bloated, inefficient public sector and thereby favoured economic develop-
ment. On the other hand, a less inclusive elite bargain in a context of high
social fragmentation would also have increased the likelihood of civil war
with obvious negative implications for Zambia’s economic record. In the end,
there was probably no alternative to sacrificing economic development for
the sake of political peace, especially in the early decades of postcolonial rule
when the unity of the country was still extremely fragile.

Conclusions
Pre-existing research suggests that inclusive elite bargains should facilitate
both peace and economic development. This article has sought to contribute
to these debates by studying elite bargains in postcolonial Zambia. The
findings show that all good things do not always go together. While inclusive
elite bargains have indeed helped to avoid civil war, they have often
constrained the country’s economic development—a dilemma of unproduc-
tive peace.
Is this argument just specific to Zambia? The above-cited large-N research
by Min et al finds that ethnic inclusion increases long-term economic growth
rates, which seems to indicate that the Zambian case is little more than an
‘outlier’. Nonetheless, there is evidence that the Zambian dilemma of
unproductive peace also affects other countries in sub-Saharan Africa and
beyond. Tanzania, for instance, has long exhibited trade-offs between peace
and economic development that are strikingly similar to those observed in
Zambia.84 More generally, Englebert’s research on state legitimacy demon-
strates how ethnic diversity in Africa requires co-opting leaders of different
ethnic groups, often to the detriment of economic development.85 Similarly,
LeVan recently argued that postconflict power sharing may entail serious
1865
 STEFAN LINDEMANN

economic costs.86 All this indicates that there is a need for more detailed
research on the maybe not so obvious relationship between inclusive elite
bargains and economic development. At the same time it may be of interest
also to investigate under what conditions exclusionary elite bargains and the
related violent conflict may lead to developmental outcomes. This is
especially the case since processes of economic transformation have
historically been extremely violent in nature.87
To be sure, this article is not a challenge to promoting inclusive power
structures, which are indispensable for preserving peace and stability.
Instead, it is merely a warning to systematically consider potential trade-
offs with economic development, not least with a view to being able to
minimise such contradictions.

Notes
1 A Whaites, States in Development: Understanding State-building, Working Paper, London: DFID, 2008,
p 7.
2 DFID, ‘Building peaceful states and societies’, Practice Paper, London, 2009, p 22.
3 For a literature review, see J DiJohn & J Putzel, Political Settlements, GSDRC Issues Paper,
Birmingham: University of Birmingham, 2009.
4 DC North, J Wallis & BR Weingast, Violence and Social Orders: A Conceptual Framework for
Interpreting Recorded Human History, Cambridge: Cambridge University Press, 2009, pp 18–21.
5 DFID, ‘Building peaceful states and societies’, p 23.
6 North, Wallis & Weingast, Violence and Social Orders, p 21.
7 See, for example, M Mufti, ‘Elite bargains and the onset of political liberalization in Jordan’,
Comparative Political Studies, 32(1), 1999, pp 100–129.
8 DiJohn & Putzel, Political Settlements, p 5.
9 A Wimmer, L-E Cederman & B Min, ‘Ethnic politics and armed conflict: a configurational analysis of
a new global data set’, American Sociological Review, 74(2), 2009, pp 316–337.
10 A Lijphart, Democracy in Plural Societies, New Haven, CT: Yale University Press, 1977; and Lijphart,
Thinking about Democracy: Power-Sharing and Majority Rule in Theory and Practice, London:
Routledge, 2008.
11 MG Burton & J Higley, ‘Elite settlements’, American Sociological Review, 52(3), 1987, pp 295–307.
12 D Rothchild & MW Foley, ‘African states and the politics of inclusive coalitions’, in D Rothchild & N
Chazan (eds), The Precarious Balance: State & Society in Africa, Boulder, CO: Westview Press, 1988,
pp 233–264.
13 F Stewart (ed), Horizontal Inequalities and Conflict: Understanding Group Violence in Multiethnic
Societies, New York: Palgrave Macmillan, 2008.
14 See, among others, BF Walter, Committing to Peace: The Successful Settlement of Civil Wars,
Princeton, NJ: Princeton University Press, 2002; C Hartzell & M Hoddie, ‘Institutionalizing peace:
power sharing and post-civil war conflict management’, American Journal of Political Science, 47(2),
2003, pp 318–332; and AK Jarstad & D Nilsson, ‘From words to deeds: the implementation of power-
sharing pacts in peace accords’, Conflict Management and Peace Science, 25(3), 2008, pp 206–223.
15 See, for example, IS Spears, ‘Africa: the limits of power-sharing’, Journal of Democracy, 13(3), 2002,
pp 123–136; D Rothchild, ‘Reassuring weaker parties after civil wars: the benefits and costs of
executive power-sharing systems in Africa’, Ethnopolitics, 4(3), 2005, pp 247–267; and A Mehler, ‘Peace
and power sharing in Africa: a not so obvious relationship’, African Affairs, 108(432), 2009, pp 453–
473.
16 See W Easterly & R Levine, ‘Africa’s growth tragedy: policies and ethnic divisions’, Quarterly Journal
of Economics, 112(4), 1997, pp 1203–1250; and A Alesina & E La Ferrara, ‘Ethnic diversity and
economic performance’, Journal of Economic Literature, 43(3), 2005, pp 762–800.
17 B Min, L-E Cederman & A Wimmer, ‘Ethnic exclusion, economic growth, and civil war’, mimeo,
University of California Los Angeles, 2010.
18 On ‘civil war as development in reverse’ see P Collier, L Elliott, H Hegre, A Hoeffler, M Reynal-Querol
& N Sambanis, Breaking the Conflict Trap: Civil War and Development Policy, Washington, DC:
World Bank, 2003.
19 M Olson, ‘Dictatorship, democracy, and development’, American Political Science Review, 87(3), 1993,
pp 567–576.

1866
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

20 J Gerring, Case Study Research: Principles and Practices, Cambridge: Cambridge University Press,
2007, pp 105–108.
21 AD Roberts, A History of Zambia, London: Heinemann, 1976.
22 DN Posner, Institutions and Ethnic Politics in Africa, Cambridge: Cambridge University Press, 2005,
pp 26–56.
23 Roberts, A History of Zambia, p 188.
24 C Gertzel & M Szeftel, ‘Politics in an African urban setting: the role of the Copperbelt in the transition
to the one-party state 1964–1973’, in C Gertzel, C Baylies & M Szeftel (eds), The Dynamics of the One-
Party State in Zambia, Manchester: Manchester University Press, 1984, p 119.
25 Roberts, A History of Zambia; and R Hall, Zambia 1890–1964: The Colonial Period, London:
Longman, 1976.
26 C Gertzel, ‘Western Province: tradition, economic deprivation and political alienation’, in Gertzel et al,
The Dynamics of the One-Party State in Zambia, pp 206–236; and O Sichone & N Simutanyi, ‘The
ethnic and regional questions: ethnic nationalism and the state in Zambia: the case of Barotseland
1964–1994’, in O Sichone & BC Chikulo (eds), Democracy in Zambia: Challenges for the Third
Republic, Harare: Sapes Books, 1996, pp 173–196.
27 G Macola, ‘Harry Mwaanga Nkumbula, UNIP and the roots of authoritarianism in nationalist
Zambia’, in J-B Gewald, M Hinfelaar & G Macola (eds), One Zambia, Many Histories: Towards a
History of Post-colonial Zambia, Leiden: Brill, 2008, pp 17–44.
28 There was even an informal agreement that the prime minister would always be either a Tonga- or
Barotse-speaker, Sichone & Simutanyi, ‘The ethnic and regional questions’, p 186. Accordingly three
out of the six prime ministers during the Second Republic were Tonga-speakers, while the remaining
three were from the Barotse-speaking bloc.
29 W Tordoff, ‘Introduction’, in Tordoff (ed), Administration in Zambia, Manchester: Manchester
University Press, 1980, p 17.
30 R Molteno, ‘Cleavage and conflict in Zambian politics: a study in sectionalism’, in W Tordoff (ed),
Politics in Zambia, Manchester: Manchester University Press, 1974, p 66.
31 RL Rotberg, ‘Tribalism and politics in Zambia’, Africa Report, 12(9), 1967, pp 29–35.
32 Interview, V Mwaanga, Former UNIP and MMD minister, Lusaka, 15 July 2008.
33 D Dresang, ‘Ethnic politics, representative bureaucracy and development administration: the Zambian
case’, American Political Science Review, 68(4), 1974, p 1612.
34 S Johns, ‘The parastatal sector’, in Tordoff, Administration in Zambia, pp 104–129.
35 G Haantobolo, ‘Civil control of the military in Zambia’, PhD dissertation, University of the
Witwatersrand, 2008, p 136.
36 Interview, Gen M Masheke, former army commander and UNIP prime minister, Lusaka, 7 August
2008.
37 M Larmer, ‘Enemies within? Opposition to the Zambian one-party state, 1972–1980’, in Gewald et al,
One Zambia, Many Histories, p 108.
38 C Gertzel, ‘Dissent and authority in the Zambian one-party state’, in Gertzel et al, The Dynamics of the
One-Party State in Zambia, pp 79–118; and M Larmer, ‘The hour has come at the pit: the
Mineworkers’ Union of Zambia and the Movement for Multi-Party Democracy, 1982–1991’, Journal
of Southern African Studies, 32(2), 2006, pp 293–312.
39 DM Bartlett, ‘Civil society and democracy: a Zambian case study’, Journal of Southern African Studies,
26(3), 2000, pp 429–446.
40 C Baylies & M Szeftel, ‘The fall and rise of multi-party politics in Zambia’, Review of African Political
Economy, 19(54), 2002, pp 75–91.
41 C Baylies & M Szeftel, ‘The 1996 Zambian elections: still awaiting democratic consolidation’, Review
of African Political Economy, 24(71), 1997, pp 113–128.
42 P Burnell, ‘The tripartite elections in Zambia, December 2001’, Electoral Studies, 22(2), 2003, pp 388–
395.
43 M Larmer & A Fraser, ‘Of cabbages and king cobra: populist politics and Zambia’s 2006 election’,
African Affairs, 106(425), 2007, pp 611–637.
44 Note that I limit my analysis to the Chiluba and Mwanawasa administrations.
45 JK Van Donge, ‘Zambia: Kaunda and Chiluba—enduring patterns of political culture’, in JA
Wiseman (ed), Democracy and Political Change in Sub-Saharan Africa, London: Routledge, 1995,
p 214.
46 B Chikulo, ‘Presidential and parliamentary elections in the Third Republic, 1991–1994’, in Sichone &
Chikulo, Democracy in Zambia, 1996, p 33.
47 B Osei-Hwedie, ‘The role of ethnicity in multi-party politics in Malawi and Zambia’, Journal of
Contemporary African Studies, 16(2), 1998, p 235; and various interviews.
48 The key appointment was clearly that of Brig-Gen Miyanda, who first served as the powerful minister
without portfolio (third in the hierarchy) and later became vice-president (second in the hierarchy).

1867
 STEFAN LINDEMANN

After the 1996 elections Miyanda was replaced by Gen Tembo—another Nyanja-speaker from the
Tumbuka ethnic group.
49 G Erdmann & N Simutanyi, Transition in Zambia: The Hybridisation of the Third Republic, Konrad
Adenauer Foundation Occasional Papers, Lilongwe: Konrad Adenauer Foundation, 2003, p 31.
50 Bartlett, ‘Civil society and democracy’, p 433.
51 See L Rakner, Political and Economic Liberalisation in Zambia, 1991–2001, Uppsala: Nordic Africa
Institute, 2003.
52 Zambia Privatisation Agency (ZPA), ‘Zambia Privatisation Agency status report’, Lusaka, 2006.
53 BC Chikulo, ‘Corruption and accumulation in Zambia’, in HR Kempe & BC Chikulo (eds), Corruption
and Development in Africa: Lessons from Country Case-Studies, London: Palgrave Macmillan, 2000,
pp 161–182; and M Szeftel, ‘Eat with us: managing corruption and patronage under Zambia’s three
Republics, 1964–99’, Journal of Contemporary African Studies, 18(2), 2002, pp 207–224.
54 JK Van Donge, ‘The plundering of Zambian resources by Frederick Chiluba and his friends: a case
study of the interaction between national politics and the international drive towards good
governance’, African Affairs, 108(430), 2009, pp 69–90.
55 Interview, Mbita Chitala, former MMD deputy minister, Lusaka, 29 July 2008.
56 ‘Mwanawasa hacks at State House ‘‘family tree’’’, Times of Zambia (Lusaka), 17 June 2004. Note that
Mwanawasa’s father is a Lenje from Central Province (Tonga-speakers), while his mother is a Lamba
from Copperbelt Province (Bemba-speakers).
57 Larmer & Fraser, ‘Of cabbages and king cobra’; and J Gould, ‘Zambia’s 2006 elections: the
ethnicization of politics?’, News from the Nordic Africa Institute, Uppsala: Nordic Africa Institute,
2007.
58 I define armed conflict as a civil war if: 1) the parties are politically and militarily organised and have
publicly stated objectives; 2) the government is a principal combatant; 3) the insurgents are locally
represented and recruit locally; 4) the conflict causes at least 500 to 1000 deaths during the first year or
at least 1000 cumulative deaths in the next three years; and 5) the weaker party is able to mount
effective resistance, measured by at least 100 deaths inflicted on the stronger party. See N Sambanis,
‘What is civil war? Conceptual and empirical complexities of an operational definition’, Journal of
Conflict Resolution, 48(6), 2004, pp 829–831.
59 M Larmer & G Macola, ‘The origins, context and political significance of the Mushala rebellion
against the Zambian one-party state’, International Journal of African Historical Studies, 40(3), 2007,
pp 471–496.
60 The Mushala rebellion does not qualify as a civil war since war-related casualties were in the dozens
rather than in the hundreds and there was no effective resistance. See P Wele, Kaunda and Mushala
Rebellion: The Untold Story, Lusaka: Multimedia Publications, 1987; and Larmer & Macola, ‘The
origins, context and political significance of the Mushala rebellion’.
61 I have used my own data based on Central Statistical Office (CSO), Census of Population and Housing
1969, Lusaka: CSO, 1973; CSO, 1980 Population and Housing Census of Zambia, Lusaka: CSO, 1985; and
Government of Zambia (GOZ), Official Verbatim Report of the Parliamentary Debates, Lusaka: GOZ,
various years.
62 Interview, V Mwaanga.
63 Various interviews, Lusaka, July–October 2008.
64 Sichone & Simutanyi, ‘The ethnic and regional questions’, p 188.
65 ‘Zambia’s second-biggest province seeks secession’, Agence France Presse, 22 July 1993; and ‘Zambia:
government warns against secession’, Inter-Press Service, 20 July 1993.
66 P Englebert, ‘Compliance and defiance to national integration in Barotseland and Casamance’, Africa
Spectrum, 39(1), 2005, pp 29–59.
67 Burnell, ‘The tripartite elections in Zambia, December 2001’.
68 Various interviews, Lusaka, July–October 2008.
69 Evidence of this is that corruption charges against Chiluba were suddenly dropped after years of
investigation. ‘Chiluba acquitted’, Times of Zambia, 18 August 2009.
70 Interview, D Patel, former MMD minister, Lusaka, 9 September 2008.
71 V Chisala, A Geda, H Dagdeviren, T McKinley, A Saad-Filho, C Oya & J Weeks, Economic Policies
for Growth, Employment and Power Reduction: Case Study of Zambia, Lusaka: UNDP, 2006.
72 Ibid.
73 W Tordoff & R Molteno, ‘Government and administration’, in Tordoff, Politics in Zambia, p 281.
74 Ibid.
75 Tordoff, ‘Introduction’, p 16.
76 Cited in Gertzel, ‘Dissent and authority in the Zambian one-party state’, p 88.
77 D Dresang & R Young, ‘The public service’, in Tordoff, Administration in Zambia, p 86.
78 C Legum, Africa Contemporary Record 1987–88: Annual Survey and Documents, New York: Holmes &
Meier, 1990, p B828.

1868
 INCLUSIVE ELITE BARGAINS AND THE DILEMMA OF UNPRODUCTIVE PEACE

79 JR Craig, ‘State enterprise and privatisation in Zambia, 1968–1998’, PhD Dissertation, University of
Leeds, 1999.
80 Ibid; and J Thurlow & P Wobst, ‘The road to pro-poor growth in Zambia: past lessons and future
challenges’, Development Strategy and Governance Division Discussion Paper 16, Washington, DC:
International Food Policy Research Institute, 2004.
81 Erdmann & Simutanyi, ‘Transition in Zambia’, p 63.
82 N van de Walle, African Economies and the Politics of Permanent Crisis, 1979–1999, Cambridge:
Cambridge University Press, 2001.
83 Szeftel, ‘Eat with us’, p 219; and interview, R Lifuka, Transparency International Zambia, Lusaka, 1
August 2008.
84 M Khan & H Gray, State Weakness in Developing Countries and Strategies of Institutional Reform—
Operational Implications for Anti-corruption Policy and a Case Study of Tanzania, London: DFID, 2006.
85 P Englebert, State Legitimacy and Development in Africa, Boulder, CO: Lynne Rienner, 2000.
86 C LeVan, ‘Power sharing and inclusive politics in Africa’s uncertain democracies’, Governance, 24(1),
2011, pp 31–53.
87 C Cramer, Civil War is Not a Stupid Thing: Accounting for Violence in Developing Countries, London:
Hurst, 2006.

Notes on contributor
Stefan Lindemann is a research fellow at the Max-Planck-Institute for the
Study of Religious and Ethnic Diversity, and is also associate lecturer at the
Department of Political Science, Georg-August-Universität Göttingen. He
received his PhD from the London School of Economics and Political
Science (LSE). His current main research interests include ethnic conflict,
power sharing, civil war, and military coups. His most recent publication is
‘Just another change of guard? Broad-based politics and civil war Museveni’s
Uganda’, African Affairs, 110(440), 2011, pp 387–416.

1869
Copyright of Third World Quarterly is the property of Routledge and its content may not be copied or emailed
to multiple sites or posted to a listserv without the copyright holder's express written permission. However,
users may print, download, or email articles for individual use.