Parties, Platforms, and Political Mobilization: The Zambian Presidential Election of 2008

Abstract

The death of President Levy Mwanawasa in August 2008 plunged Zambian politics into a state of flux. This article argues that the way the main parties responded to the challenge of the resulting presidential by-election has three lessons to teach the emerging literature on political parties. First, Rupiah Banda’s rise to power within the MMD demonstrates the extent to which intra-party machinations can leave a party saddled with an unpopular leader, and hence illustrates the great significance of succession struggles within dominant-party systems. Second, the main parties’ continual repositioning of their electoral platforms reveals that not all African elections take place in an ideological vacuum, and shows that the platforms parties adopt can only be fully understood in the context of the wider party system and the way in which parties interact over time. Finally, the ability of controversial opposition leader Michael Sata to mobilize a diverse support base - by employing a ‘populist’ message in urban areas at the same time as receiving the support of his ethno-regional community in rural areas - lays bare the complexity of party strategies and the limits of the ‘ethnic census’ model of party support. Taken together, these findings suggest that the tendency to divorce the study of elections from the study of how parties function and interact impoverishes our understanding of African politics.

Full text

PARTIES, PLATFORMS, AND POLITICAL
MOBILIZATION: THE ZAMBIAN
PRESIDENTIAL ELECTION OF 2008
NIC CHEESEMAN AND MARJA HINFELAAR
*
ABSTRACT
The death of President Levy Mwanawasa in August 2008 plunged Zambian
politics into a state of flux. This article argues that the way the main parties
responded to the challenge of the resulting presidential by-election has three
lessons to teach the emerging literature on political parties. First, Rupiah
Banda’s rise to power within the MMD demonstrates the extent to which
intra-party machinations can leave a party saddled with an unpopular
leader, and hence illustrates the great significance of succession struggles
within dominant-party systems. Second, the main parties’continual repo-
sitioning of their electoral platforms reveals that not all African elections
take place in an ideological vacuum, and shows that the platforms parties
adopt can only be fully understood in the context of the wider party sys-
tem and the way in which parties interact over time. Finally, the ability of
controversial opposition leader Michael Sata to mobilize a diverse support
base –by employing a ‘populist’message in urban areas at the same time as
receiving the support of his ethno-regional community in rural areas –lays
bare the complexity of party strategies and the limits of the ‘ethnic census’
model of party support. Taken together, these findings suggest that the ten-
dency to divorce the study of elections from the study of how parties
function and interact impoverishes our understanding of African politics.
THE UNTIMELY DEATH OF PRESIDENT Levy Mwanawasa on 19 August
2008 plunged Zambian politics into a state of flux. In contrast to the vast
majority of presidential systems, which allow for automatic internal succes-
sion within the ruling party, the Zambian constitution stipulates that
following the death of the executive a ‘by-election’must be held within 90
days. Consequently, Mwanawasa’s death was followed by a mixture of na-
tional mourning, factional realignment, and political uncertainty that was
only partially resolved by the narrow victory of the ruling Movement for
*Nic Cheeseman (nicholas.cheeseman@politics.ox.ac.uk) is University Lecturer in African
Studies and Hugh Price Fellow, Jesus College, Oxford University. Marja Hinfelaar (marja.
hinfelaar@gmail.com) is an independent scholar based in Lusaka, Zambia. The research for
this article was supported by a grant from Jesus College, Oxford University. The authors
would like to thank the editors and anonymous reviewers for their thoughtful and insightful
suggestions.
51
African Affairs, 109/434, 51–76 doi: 10.1093/afraf/adp070
© The Author 2009. Published by Oxford University Press on behalf of Royal African Society. All rights reserved
Advance Access Publication 24 November 2009
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

Multi-Party Democracy’s Rupiah Banda. The unexpected presidential elec-
tion represented a major test for both the government and the opposition.
The central challenge for the ruling Movement for Multi-Party Democracy
(MMD) stemmed from the lack of an obvious replacement for Mwanawasa;
the government was thus forced to manage a difficult period of succession
politics while simultaneously rolling out an election campaign. The main
problem for the opposition was the short timeframe within which to secure
election funds and mobilize support, and the need to respect the 21 days of
national mourning.
The way in which the rival parties dealt with the snap election has
much to tell us about how intra-party politics, coalition formation, and
political mobilization are developing in Africa’s multi-party systems.
How parties function is critical to the performance of the wider political
system; as Thomas Carothers has written recently, parties are frequently
the ‘weakest link’in new democracies.
1
Despite this, parties have typically
been ignored as units of study in their own right, in part because they are
often seen as being little more than the playthings of their leaders, on the
one hand, and a sad reflection of the societies which give rise to them on
the other.
2
Although many parties do conform to these negative stereo-
types, a significant number buck the trend. Most obviously, in South
Africa Thabo Mbeki’s ruling clique was defeated in party elections by
the challenge of Jacob Zuma, paving the way for an internal transfer of
power within the ruling African National Congress.
3
In Ghana, a durable
two-party system looks to be consolidating on the basis of deep political
traditions that can be traced back to the late colonial era, supporting one of
Africa’s most remarkable processes of democratic transformation –including
a largely peaceful, if controversial, second transfer of power in early 2009.
4
If
we are to understand the current state of multi-party politics in countries such
as Zambia, we must research the conditions under which more mature and
institutionalized parties and party systems emerge, and the impact on the pro-
spects for democratic consolidation if they do not.
This article begins by reviewing the ‘three waves’of literature on African
political parties that have emerged over the past seventy years, then uses
interviews with MPs, first-hand experience of the campaign, a question-
naire administered to election observers, and interviews conducted by a
number of research assistants in a variety of locations to demonstrate three
1. Thomas Carothers, Confronting the Weakest Link: Aiding political parties in new democracies
(Carnegie Endowment for International Peace, Washington, DC, 2007).
2. Important exceptions are discussed below.
3. Alexander Beresford, ‘Comrades “back on track”? The durability of the tripartite alliance
in South Africa’,African Affairs 108, 432 (2009), pp. 391–412.
4. Lindsay Whitfield, ‘“Change for a better Ghana”: party competition, institutionalization
and alternation in Ghana’s 2008 elections’,African Affairs 108, 433 (2009), pp. 621–41.
AFRICAN AFFAIRS52
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

important lessons that the Zambian presidential by-election of 2008 can
contribute to this body of work. First, Rupiah Banda’s rise to power within
the MMD shows the extent to which intra-party machinations can leave a
party saddled with an unpopular leader, demonstrating the need to devote
far greater attention to internal party dynamics. Second, the continual re-
positioning of party platforms on major issues of economic policy and
campaign style reveals that not all African elections take place in an ideo-
logical vacuum.
5
As LeBas has recently argued, the policy positions and
strategies adopted by parties matter, but they can only be understood fully
within the context of the wider party system and the way in which parties
interact over time.
6
Finally, the ability of controversial opposition leader Michael Sata to
mobilize a cross-ethnic support base of the ‘dispossessed’in urban areas
supports Larmer and Fraser’s claim that his rise to prominence derives in
part from his ‘populist’stance, and lays bare the limits of the ‘ethnic cen-
sus’model of party support.
7
However, the fact that Sata also received a
third of his support from his rural ethno-regional ‘heartlands’suggests
that his success derives from his ability to knit together a diverse support
base, and makes clear the importance of avoiding simplistic characteriza-
tions of the complex strategies adopted by leaders and parties. Drawing
together these three ‘lessons’, we conclude that parties matter, and that
the tendency to divorce the study of African elections from the study of
how parties function and interact impoverishes our understanding of Af-
rican politics.
Parties in historical perspective
Africanist perspectives on the function and significance of political parties
have fluctuated wildly over the past seventy years. In order to appreciate
the determination of a recent ‘third wave’of literature to reinstate parties
at the centre stage of the study of African politics it is necessary to reflect
briefly on the core assumptions of previous work on parties and party sys-
tems. The emergence of nationalist movements in the late colonial period
inspired a first wave of ‘party literature’which focused on the social con-
nections and political aspirations of African political parties, reflecting the
5. Nicolas van de Walle, ‘Meet the new boss, same as the old boss? The evolution of political
clientelism in Africa’in Herbert Kitschelt and Steven I. Wilkinson (eds), Patrons, Clients and
Policies: Patterns of democratic accountability and political competition (Cambridge University
Press, Cambridge, 2007), pp. 62–3.
6. Adrienne LeBas, ‘Polarization as craft: explaining party formation and state violence in
Zimbabwe’,Comparative Politics 38, 4 (2006), pp. 419–38.
7. Miles Larmer and Alastair Fraser, ‘Of cabbages and King Cobra: populist politics and
Zambia's 2006 election’,African Affairs 106, 425 (2007), pp. 611–37.
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 53
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

optimism of the time. Although critical of the leadership role of the African
middle class, Thomas Hodgkin’s seminal work described parties that were
successful because they tapped into widely felt frustrations with the injus-
tice of colonial rule, and as a result shaped party systems that reflected a
range of social and communal cleavages.
8
Significantly, the first wave of
literature both took African parties seriously and attempted to employ con-
ventional political science methods and frameworks in their analysis. Most
notably, Ruth Schachter-Morgenthau suggested that, like those in Europe,
West African parties could be divided into elite parties, which relied on the
charisma or popularity of a small number of leaders to attract support, and
mass parties, which depended on a more developed organization based on
the continual and active participation of party members. Indeed, writing in
1965, Schachter-Morgenthau suggested that the mass party was on the rise
in Africa, predicting that the greater inclusivity of the mass-party model
would give it a clear advantage in terms of mobilizing electoral support.
9
Although Bienen and Zolberg found that on closer inspection many sup-
posedly mass parties relied on patron–client structures,
10
calling the
distinction between mass and elite parties into question, the early literature
on nationalism in Zambia implicitly recognized that parties with a broad
urban base did exist in Africa and that they could be highly effective. Most
obviously, Mulford, and later Scott, documented the rapid development of
the United Independence Party (UNIP) in urban areas along the line of rail
through a mass-party organization which divided households into party
cells and demanded far-reaching sacrifices of both time and money from
party members.
11
In line with these findings, the influential work of Robert
Bates traced UNIP’s social roots, arguing that much of the party’s early
success depended on the support and organizational capacity of the ever-
growing ranks of Zambian mineworkers.
12
A second wave of literature arose from the 1970s onwards, largely in re-
sponse to the perceived failure of African political parties to make good the
dreams of the nationalist movement. As noted by Allen, this literature docu-
ments the decline of competitive multi-partyism into either ‘centralized-
8. Thomas Hodgkin, African Political Parties (Peter Smith, Gloucester, 1971).
9. Ruth-Schachter Morgenthau, Political Parties in French-speaking West Africa (Clarendon
Press, Oxford, 1967).
10. Henry Bienen, ‘One-party systems in Africa’in Samuel Huntington and Clement Moore
(eds), Authoritarian Politics in Modern Society (Basic Books, New York, NY, 1970); Aristide
Zolberg, One-Party Government in the Ivory Coast (Princeton University Press, Princeton,
NJ, 1969).
11. D. Mulford, The Politics of Independence (Oxford University Press, Oxford, 1976); Ian
Scott, ‘Party functions and capabilities: the local-level UNIP organization during the first
Zambian republic’,African Social Research 22, 45 (1976), pp. 107–28.
12. Robert Bates, Unions, Parties, and Political Development: A study of mineworkers in Zambia
(Yale University Press, New Haven, CT, 1971).
AFRICAN AFFAIRS54
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

bureaucratic’one-party states or ‘spoils politics’.
13
Parties in single-party
systems tended to lose their sense of purpose and identity, stagnating in
the absence of formal political competition and rendered irrelevant by the
existence of a more effective state apparatus. Gertzel, Baylies, and Szeftel
provided one of the most important descriptions of this process of party at-
rophy in Zambia following the creation of a relatively open one-party state in
1972.
14
As Bratton’s work later recorded, economic decline eroded party
funds and undermined the party’s slogan that ‘it pays to belong to UNIP’,
leading to a period of irreversible decay in formal political structures that
was particularly acute at the local level.
15
The rapid decline of a diverse
range of parties in the 1970s led to a negative re-evaluation of the signifi-
cance of nationalist-era parties more generally, in which critics such as
Lemarchand detailed the limited reach of formal party institutions.
16
As van de Walle has noted recently, the second wave of party literature
ultimately concluded that ‘as the independence era parties ventured out into
the rural hinterland, they typically had little choice but to rely on traditional
patrons as “intermediaries”or “brokers”between the political centre and
the countryside, given the initial weakness of the party, and the often greater
legitimacy enjoyed by these local actors’.
17
The implications of this critique
for the study of parties was significant, because it suggested that political
mobilization in Africa depended on informal structures that could not sim-
ply be analysed using tools borrowed from the study of parties in Western
Europe. This reappraisal reached its zenith with Hyden’s suggestion that as
African nationalist parties often lacked formal party structures, cohesion,
and longevity, they should not be classified as ‘parties’at all, and would
be better conceptualized as ‘movements’.
18
Hyden’s intellectual reclassifi-
cation of African parties epitomized a wider trend in which academic
energies were redirected towards the study of the politics of the belly, leaving
the analysis of party structures behind.
Following the re-introduction of multi-party politics in the early 1990s,
the rapid turnover of parties and the tendency to mobilize support via
appeals to community identity reinforced the view that parties were not
worth studying in their own right. Both scholars and journalists spoke of
13. Chris Allen, ‘Understanding African politics’,Review of African Political Economy 22,65
(1995), pp. 301–20.
14. Cherry Gertzel, Carolyn Baylies, and Morris Szeftel, The Dynamics of the One-Party State
in Zambia (Manchester University Press, Manchester, 1974).
15. Michael Bratton, Peasant and Party-State in Zambia (University Press of New England,
London, 1980).
16. Rene Lemarchand, ‘Political clientelism and ethnicity in tropical Africa: competing so-
lidarities in nation building’,American Political Science Review 64, 1 (1972), pp. 68–90.
17. van de Walle, ‘Meet the new boss’, p. 52.
18. For the latest statement of this point see Goran Hyden, African Politics in Comparative
Perspective (Cambridge University Press, Cambridge, 2006), Chapter 2.
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 55
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

‘paper parties’that were willing and able to change their names and logos
at will, operating without any real sense of internal democracy and depen-
dent on a small number of wealthy leaders for both policies and resources.
19
Rakner’s work on the period 1991–2001 shows that many of these stereo-
types held true for Zambia: the victory of the MMD in 1991 was followed by
political fragmentation and the increasing dependence of opposition parties
on their leaders for resources and direction.
20
The tendency to discount parties as important factors in their own right
was underpinned by an influential approach to the study of African elections
which suggested that they could be modelled as an ‘ethnic census’.The
most famous proponent of this approach was Donald Horowitz, who argued
that as a result of the psychological and historical associations between cer-
tain ethnic groups and political parties in ethnically segmented societies,
ethnicity enjoys an overwhelming influence over political behaviour.
21
The assumption that formal party organizations were all but non-existent,
and that political mobilization was best explained via a focus on the personal
ties of individual leaders to their ethno-regional communities, ensured that
the study of parties represented little more than a footnote in the early anal-
ysis of electoral competition in Africa.
Over the past decade a third-wave of party literature has emerged which
has employed in-depth empirical research and survey data to challenge
many of the received wisdoms of previous scholarship. Led by Daniel
Posner’s work on ethnicity, Carrie Manning’s analysis of party systems,
Nicolas van de Walle’s discussion of dominant-party regimes, and Salih’s
collection of case-study material, this new literature has demonstrated
that the strategies adopted by parties, and the responses of African voters,
are complex and require substantial further study.
22
In line with the main
thrust of this new body of literature, this article argues that divorcing the
study of elections from the study of parties leaves us with a jaundiced under-
standing of the political process; only by taking parties seriously can we
reach an adequate understanding of the Zambian election of 2008. In seek-
ing to contribute to the broader reappraisal of party politics in Africa, we
draw out the lessons from the Zambian by-election of 2008 for three of
19. For example, Ngugi wa Thiong’o, ‘The choices before us: reflections on Mwai Kibaki and
the 2007 Kenya general elections’,Pambazuka and Daily Nation, <http://www.pambazuka.
org> (24 June 2009).
20. Lise Rakner, Political and Economic Liberalisation in Zam bia 1991–2001 (Nordiska
Afrikainstitutet, Uppsala, 2003).
21. Donald Horowitz, Ethnic Groups in Conflict (University of California Press, Berkeley,
CA, 1985).
22. Daniel N. Posner, Institutions and Ethnic Politics in Africa (Cambridge University Press,
Cambridge, 2005); Carrie Manning, ‘Assessing African party systems’,Party Politics 11,6
(2005), pp. 715–6; van de Walle, ‘Meet the new boss’; M. A. Mohamed Salih (ed.), African
Political Parties (Pluto Press, London, 2003).
AFRICAN AFFAIRS56
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

the key issues that have been taken up by the third-wave literature: the sig-
nificance of intra-party dynamics in dominant-party systems, the need to
understand a party’s election campaign strategy in the context of the wider
party system, and the limits of the ‘census politics’approach in explaining
party support in Africa.
Lesson one: dominant-party government and the importance of succession politics
Despite the reintroduction of multi-partyism, political competition has re-
mained limited in countries as diverse as South Africa, Cameroon,
Namibia, Zimbabwe, and Zambia. Consequently, a major theme within re-
cent writing on parties has been the causes and consequences of dominant-
partyism. Writing after most countries had only experienced one or two
multi-party elections, van de Walle and Butler are not alone in finding that
most African party systems feature dominant parties, meaning that the larg-
est party holds more than 60 percent of the seats.
23
Zambia offers an
instructive illustration of this phenomenon. Following the MMD’s landslide
election victory in 1991, commentators were quick to identify Zambia as a
classic dominant-party state. As Baylies and Szeftel noted, since the MMD
had effectively been able to take over the state wholesale post-1991, and con-
stitutional reform was effectively delayed indefinitely, the party retained the
advantages of incumbency previously abused by UNIP during the one-party
era.
24
Many of these abuses, including the refusal to allow former President
Kenneth Kaunda to contest the 1996 elections, the use of state resources to
support the MMD presidential candidate, and the ability of the executive to
influence the police, electoral commission, and media, have been documen-
ted by Rakner and Svasand.
25
As a result, the MMD has won all five multi-party elections since 1991,
illustrating Bogaards’claim that ‘dominant parties dominate’in sub-Sahar-
an Africa.
26
The party’s electoral strength reflects a wider trend identified
by Posner and Young, in which 68 percent of electoral contests in Africa
since 1990 have been won by the incumbent, with the consequence that
rotations of the presidency have occurred not as a result of transfers of pow-
er between parties, but through internal succession battles within parties.
27
23. Nicolas van de Walle and K. S. Butler, ‘Political parties and party systems in Africa’s
illiberal democracies’,Cambridge Review of International Studies 13, 1 (1999), pp. 14–28.
24. Carloyn Baylies and Morris Szeftel, ‘The 1996 Zambian elections: still awaiting demo-
cratic consolidation’,Review of African Political Economy 24, 71 (1997), pp. 113–28.
25. Lise Rakner and Lars Svasand, ‘Stuck in transition: electoral processes in Zambia 1991–
2001’,Democratization 12, 1 (2005), pp. 85–105.
26. Matthijs Bogaards, ‘Counting parties and identifying dominant party systems in Africa’,
European Journal of Political Research 43, 2 (2004), pp. 173–97.
27. Daniel N. Posner and Daniel J. Young,‘The institutionalization of political power in
Africa’,Journal of Democracy 18, 3 (2007), pp. 126–40.
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 57
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

As with other dominant-party states on the continent, in Zambia the iden-
tity of the President has been determined by the succession process and
intra-party struggles within the MMD. Because of the centrality of the ex-
ecutive to African political life, itself rooted in the failure to conduct
genuine constitutional reform in most new democracies, these succession
battles have typically been high-stakes games with the potential to be both
controversial and destabilizing. Govea and Holm estimate that 61 percent
of the successions they identify in Africa were ‘unregulated’, with many de-
teriorating into political crisis and violence.
28
The potential implications of a succession battle are well illustrated by the
Kenyan case, where Daniel arap Moi’s mishandling of the appointment of
his successor within the ruling KANU party in the run-up to the 2002 elec-
tion led first to a major split within the party, and later to its electoral
defeat.
29
Managing succession politics has proved equally challenging for
the MMD. In 2001 the party successfully negotiated the transfer from
Chiluba (constitutionally barred from a third-term) to Levy Mwanawasa,
although not without cost; soon afterwards, Michael Sata quit the party –
having been overlooked by Chiluba –and established the Patriotic Front
(PF), which quickly became the MMD’s main electoral rival.
30
In 2008
the party faced a potentially even greater challenge, as it struggled to main-
tain party unity amidst an unregulated succession struggle following the
unexpected death of Mwanawasa. The way in which internal party dynamics
and the need to minimize defections in the run-up to the presidential by-
election led to the selection of Rupiah Banda as presidential candidate, in
the process saddling the MMD with an uninspiring and largely unpopular
leader, demonstrates the importance of studying succession processes in
dominant-party states for an understanding of wider political realities.
The Mwanawasa succession
Following Mwanawasa’s stroke on 29 June 2008, a behind-the-scenes suc-
cession battle raged within the MMD, as the party appeared reluctant to
disclose the status of the President’s health. As Kabanda has argued, the
failure of the constitution to specify the process for replacing an incapaci-
tated president left the country in a state of limbo until the announcement
of Mwanawasa’s death on 19 August 2008 and the timetabling of a presi-
dential by-election for 30 October.
31
The proximity of the polls thrust the
28. Rodger M. Govea and John D. Holm, ‘Crisis, violence and political succession in Africa’,
Third World Quarterly 18, 1 (1998), pp. 129–48.
29. D. M. Anderson, ‘Briefing: Kenya’s elections 2002 –the dawning of a new era?’,African
Affairs 102, 407 (2003), pp. 331–42.
30. See Jeremy Gould, ‘Zambia’s 2006 elections: the ethnicization of politics?’(article in Nor-
dic Africa Institute newsletter, No. 1, January 2007), <http://www.nai.uu.se/> (24 June 2009).
31. S. Kabanda, ‘A crisis when the President falls sick’,The Challenge 10, 3 (2008), p. 13.
AFRICAN AFFAIRS58
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

MMD into an urgent search for a presidential candidate, which many com-
mentators and party leaders believed the party would not survive intact.
32
Internal wrangling in part derived from Mwanawasa’s failure to specify a
chosen successor, despite previously being seriously injured in a road acci-
dent, suffering a minor stroke in 2006, and having endured prolonged bouts
of ill-health.
The lack of internal control over the succession process reduced the risk
that a favoured candidate might be foisted on the party, but also increased the
likelihood that internal manoeuvring would descend into an unsightly free-
for-all. As Habasonda rightly observed, ‘if [MMD] don’t handle this succes-
sion issue properly . . . they risk becoming the opposition themselves’.
33
In
contrast to usual protocol, the party’s National Executive Committee (NEC)
asserted its right to choose Mwanawasa’s successor, sparking speculation
that the party leaders feared that a convention would deny them the oppor-
tunity to install their favoured candidate. However, even the NEC had not
arrived at a consensus and a long list of pretenders to Mwanawasa’s crown
emerged, including Acting President Rupiah Banda, Finance Minister
Ngandu Magande, Home Affairs Minister Ronnie Shikapwasha, former
Works and Supply Minister Ludwig Sondashi, and Former Vice-Presidents
Enock Kavindele and Nevers Mumba.
34
Early on it became clear that the
main battle was between Banda and Magande, and, against the early pre-
dictions of many commentators, Banda was ultimately chosen to be the
party’s presidential candidate by 47 votes to Magande’s11.
35
Critics of
the appointment pointed to Banda’s undistinguished record in public ser-
vice, his alleged loyalty to the rival United National Independence Party
(UNIP), and his lack of a following within the ruling party. Despite this,
three main factors, all rooted in internal MMD politics, contributed to Ban-
da’s victory.
First, Banda was able to use his position as Acting President during
Mwanawasa’s lengthy illness, along with the support of his wealthy backers,
to construct a patronage network in preparation for the inevitable succes-
sion battle. At the same time, he had a clear advantage over his competitors
because, as acting leader, he enjoyed a monopoly over the main media
sources. Traditional chiefs, provincial MMD committees and leaders of
opposition parties were paraded on government-owned television and
32. ‘Zambia: Mwanawasa illness reveals reform fragility’,Oxford Analytica, 7 July 2008.
33. Lee Habasonda is a member of the Southern African Centre for Constructive Resolu-
tion of Disputes (SACCORD); see also ‘Dangerous divisions in the ruling party’(IRIN news
report, 3 September 2008) <http://news.irinnews.org/PrintReport.aspx?ReportId=80156>
(23 January 2009).
34. It was feared that, within the context of a national convention, Katele Kalumba would
have an opportunity to manipulate the elections and win. Interview with Dr C. Beyani, UPND
Director of Research, 7 January 2009.
35. ‘Zambia: Mwanawasa illness reveals reform fragility’.
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 59
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

newspapers, talking up Banda’s virtues and suggesting that, as he was the
natural successor to Mwanawasa, the presidential by-election was superflu-
ous. It could therefore be argued that whoever had been Vice-President
would have won the candidacy; that Banda simply found himself ‘in the
right place at the right time’.
36
However, the choice of Banda was not an
innocent accident. The second key factor underpinning his victory was the
recognition of senior MMD ‘king makers’, many of whom had had worked
together in the United National Independent Party (UNIP) during the one-
party state era, that Banda was a figure from a similar political generation
with whom they could do business. As a result, leaders such as the wealthy
and influential former Foreign Minister Vernon Mwaanga threw their
weight behind his campaign. Mwaanga’s support for Banda, and the inter-
nal divisions within MMD between pro- and anti- Banda camps, is best
understood in the context of an intra-party tussle between competing fac-
tions that represent ‘two distinct political ideologies that have characterized
the struggle for control during the MMD’s rule’.
37
The Mwaanga faction is mainly comprised of figures who gained prom-
inence under Chiluba’s reign, some of whom, like Mwaanga himself, were
survivors from the UNIP era. This group viewed the MMD ‘as an instru-
ment for regaining their lost political and economic power’but saw their
political clout decline during Mwanawasa’s presidency and supported
Banda because his candidacy promised them a return to the political spot-
light.
38
A second faction of reform-minded ‘young Turks’, some of whom
are relative newcomers to the party, has resisted the resurgence of the
Chiluba faction within the MMD and the recycling of the old UNIP-
era elite. This group has called for reforms to breathe fresh life into the
party, and supported Magande largely because he was the candidate most
likely to defeat Banda. Although competition between these factions is
fierce, the boundaries between them are not always easy to draw. While
Mwaanga considers himself to be a ‘true-blue’party loyalist, and the young
Turks to be alendo ‘visitors’, Banda’s late entry into the MMD, and contin-
ued close ties to UNIP, left him vulnerable to the accusation that he too was
alendo, and hence lacked the necessary qualifications to exert authority over
‘true blue’party activists. Despite these contradictions, the greater wealth
and influence of the Mwaanga faction ultimately underpinned Banda’s vic-
tory, and guaranteed the old guard a return to political influence. However,
it is worth noting that Banda’s reliance on Mwaanga lost him significant
support in the country at large. Most obviously, the private media in the
36. N. Simuntanyi, ‘MMD’s choice of successor to the party’,The Post, 8 September 2008.
37. K. Kalumba, ‘Political concessions in a fragile democracy’,The Times of Zambia, <http://
www.times.co.zm/news/viewnews.cgi?category=8&id=1004734345c> (24 February 2009).
38. Ibid.
AFRICAN AFFAIRS60
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

shape of The Post newspaper openly opposed Banda, citing its concern at
the rise of ‘corrupt’elements within MMD, referring explicitly to Katele
Kalumba and Vernon Mwaanga, who is widely believed to have been in-
volved in drug smuggling.
39
The final factor that contributed to Banda’s victory was his ability to
present himself as a candidate who would not use the presidency to build
his own dominant exclusionary clique within MMD, and hence as a less
threatening candidate to those with presidential aspirations in the near fu-
ture. Although his reliance on Mwaanga called this pledge into question,
Banda’s relatively weak standing in the party and his advanced age (he will
soon be 72) conferred credibility on his promise that he would not stand in
the 2011 elections, offering his rivals an opportunity to re-launch their own
presidential campaigns. Of course, Banda may well decide that the taste of
power is too sweet to be given up lightly, and since the election he has
moved to further consolidate his position by taking on the mantle of party
leader at the February 2009 NEC meeting. This concentration of power in
Banda’s hands appears to be a breach of his previous statements, and may
exacerbate tensions within the party and undermine future party unity.
However, in the run-up to the 2008 polls it was his ability to present himself
as a ‘one-term’leader that enabled the MMD to hold together. Despite
deep divisions at the heart of the MMD, all of the losing candidates de-
clared their commitment to the party, which did not suffer any damaging
defections in the run-up to the 2008 polls.
Banda’s victory highlights the complex process of succession politics as
highlighted by Govea and Holm, and has much to tell us about internal
party dynamics within the types of dominant-party systems described by
Bogaards. A wide range of factors played into the selection of the MMD
presidential candidate, including the need to keep the party together, the
ability of Banda to deliver patronage, the national popularity of different
candidates, and the age and profile of the leading contenders. Ultimately,
the MMD selected the candidate who made the most sense in terms of the
internal party debate, rather than the looming national contest. The conse-
quence of the internal party dynamics described here was therefore that the
MMD entered the 2008 elections with a candidate who lacked significant
popular support at the national level, creating fresh opportunities for the PF
opposition. Banda’s victory is thus further evidence of the significance of
internal party dynamics for national-level political outcomes, and of the val-
ue of studying internal party dynamics in dominant-party systems.
39. Donatella Lorch, ‘In Zambia, a legacy of graft and a drug scandal taint democratic re-
forms’,New York Times, <www.nytimes.com> (24 February 2009).
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 61
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

Lesson two: party platforms and the party system
A second key theme within the third wave of party literature concerns the
way in which parties seek to articulate the demands of their supporters.
Party platforms have long been an under-studied aspect of African poli-
tics, in part because of the assumption that African elections are
ideology-free zones. As Manning has noted, ‘The role of ideology in party
formation and competition tends to be weak in African third-wave democ-
racies, not least because, with the advent of structural adjustment and
high aid dependency, all political contenders are constrained by the same
economic model and policy parameters.’
40
Zambia offers an interesting
opportunity to reassess the role of ideology in political competition, be-
cause the salience of the trade union movement and the relatively open
nature of the one-party state resulted in the emergence of a culture of ne-
gotiation between what are now called civil society groups and the
government. As Lungu has argued, in the late 1970s the churches were
able to force the first President, Kenneth Kaunda, to rethink the introduc-
tion of ‘scientific socialism’, while trade unions proved adept at using their
‘uneasy alliance’with the ruling party to push for better living condi-
tions.
41
Similarly, in the late 1980s pressure from business interests and
urban workers forced the government to turn its back on the International
Monetary Fund and to attempt a more popular ‘home-grown’economic
strategy.
42
Significantly, the relative import of urban interests, and most
obviously of unionized mineworkers whose activity in the copper industry
has accounted for most of Zambia’s post-colonial foreign earnings and
economic growth, has been a continual theme of political tension and de-
bate. For Bratton, it was UNIP’s inability to maintain relations with the
union movement that undermined the party’s hold on power in the late
1980s.
43
In recent years, the ability of Sata and the PF to capture urban
areas has raised the question of whether a similar fate awaits the MMD.
The centrality of these debates to political competition over the last six-
ty years reveals the Zambian experience to be an important corrective to
the assumption that African politics is ‘devoid of ideology’. This debate,
which hinges on the question of whether or not African politics ignores
policy issues because it is dominated by communal identities, is of the
utmost significance. If parties simply seek to capture the support of dis-
tinct ethnic communities based on Horowitzean ties of psychological
40. Manning, ‘Assessing African party systems’, pp. 715–6.
41. Lungu, ‘The Church, labour and the press in Zambia’.
42. D. M. C. Bartlett, ‘Civil society and democracy: a Zambian case study’,Journal of
Southern African Studies 26, 3 (2000), pp. 429–46.
43. Michael Bratton, ‘Civil society and political transition in Africa’in J. W. Harbeson, D. S.
Rothchild, and N. Chazan (eds), Civil Society and the State in Africa (Lynne Rienner, London,
1994), pp. 33–51.
AFRICAN AFFAIRS62
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

association, then they do not compete to provide a more attractive bundle
of policies to floating voters. Consequently, the connection between party
platforms and trends in national opinion is broken, and the classical
Downsian argument that party competition in a two-party system will
lead to convergence on the set of policies favoured by the median voter
does not hold.
44
As Claude Ake forcefully argued, where voters cease to
punish and reward leaders for good performance, and rather vote on the
basis of community identity, the accountability of the government to the
wider public is undermined, whatever the political system.
45
However,
while such a characterization may hold for certain African countries,
the evidence from Zambia suggests something very different, namely
the presence of a relatively responsive party politics, in line with recent
developments in Ghana, Cape Verde, and South Africa.
In tracing the evolution of the Zambian party system since the re-
emergence of multi-party politics, we focus on the role of the PF and
the MMD, leaving aside for lack of space the important question of
the UPND and the distinctive patterns of regional voting that play out
in its Southern Province support base.
46
To demonstrate the evolution of
the PF/MMD relationship over time, we borrow from the innovative
work of Adrienne LeBas, who has demonstrated that in Zimbabwe the
ruling ZANU-PF and the opposition Movement for Democratic Change
(MDC) do not form policies and election strategies in isolation but in
anticipation of the strategy most likely to be adopted by the rival party.
47
In other words, the stances selected by parties reflect the continual in-
teraction of parties over time and must be understood as a product of
this dynamic, ever-changing, relationship –the party system understood
in the broadest sense. In the Zimbabwean case, LeBas argues that both
parties adopted a strategy of ‘polarizing’the electorate by adopting more
extreme messages and strategies in order to increase the distance be-
tween the parties in the minds of voters, thus consolidating their hold
over their core supporters. Although the Zambian party system has been
divided by very different factors to those operating in Zimbabwe, LeBas’s
framework for understanding the evolution of the party system is none-
theless illuminating and highlights the extent to which the two major
parties have repositioned their electoral platforms in an attempt to cap-
ture the public mood.
44. Anthony Downs, An Economic Theory of Democracy (Harper and Row, New York, NY,
1957).
45. Claude Ake, The Feasibility of Democracy in Africa (Codesria, Dakar, 2000).
46. For more on the historical roots of Tonga-based politics see Giacomo Macola, Liberal
Nationalism in Central Africa (Palgrave Macmillan, New York, forthcoming).
47. LeBas, ‘Polarization as craft’.
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 63
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

The Zambian party system, 2004–8
Under Mwanawasa, the MMD was continually repositioned in an attempt
to head offthe threat resulting first from the United Party of National
Development (UPND) and later from the rise of the PF. Following his
appointment as Chiluba’s successor in 2001, Mwanawasa led the
MMD to a narrow victory in a hugely controversial election that many
believe was really won by Anderson Mazoka of the UPND.
48
Recogniz-
ing that the popularity of his leadership and of his party could be
damaged by his association with the MMD old guard, who were becom-
ing increasingly tainted by allegations of democratic backsliding,
Mwanawasa turned against his original patron, Chiluba, seeking his pros-
ecution on multiple counts of corruption. At the same time, Mwanawasa
recognized that unlike Chiluba he could not rely on support from Bemba-
speaking areas of the Copperbelt, Luapula, and Northern Province. Con-
sequently, he set about creating a new alliance which strengthened the
MMD’s ties throughout rural Zambia. Over the next four years, urban
Bemba-speaking communities, which had played such a vital role in the
MMD’s rise to power in 1991, came to feel increasingly marginalized by
the government.
Following the formation of the Patriotic Front in 2001, Sata first looked
to build support among his own Bemba-speaking community in those areas
neglected by the MMD. However, having polled just 3 percent of the vote
in 2001, Sata refocused his message in 2006 to reach beyond a Bemba sup-
port base. To do this, he set about articulating the frustrations of the
working class and urban poor more generally, a strategy which propelled
him from seventh place in 2001 to runner-up in 2006.
49
In their analysis
of the 2006 election Fraser and Larmer wrote that Sata’s‘skill is consistent-
ly to identify the popular mood (or prejudice) of the day, to ally himself
with it to his advantage, and to associate himself with mobilizations that
demonstrate popular engagement with the issue’.
50
They concluded that
while Sata was no Chavez, his ability to mobilize support in non-Bemba-
speaking urban areas such as Lusaka derived from a genuinely ‘populist’
appeal.
51
Between 2004 and 2008, Sata’s rise to prominence, backed by a coali-
tion of the dispossessed, transformed the Zambian party system, putting
the living conditions of the urban poor at the heart of political debates.
By running a confrontational campaign which played on the differences be-
tween the PF and the MMD, Sata effectively pursued a polarizing strategy
48. Gould, ‘Zambia’s 2006 elections’.
49. Ibid.
50. Ibid., p. 624.
51. Larmer and Fraser, ‘Of cabbages and King Cobra’.
AFRICAN AFFAIRS64
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

designed to heighten the tension surrounding the election, in a manner
similar to –if somewhat more benign than –that described by LeBas. Sata’s
main aim was to create a party system in which the primary issue cleavage,
leaving aside ethno-regional divisions, was a party’s attitude towards for-
eign investors and the urban poor. In doing so, he was deliberately
playing on Zambia’s long history of urban political mobilization and suspi-
cion of international financial institutions.
52
However, it was not simply
Sata’s platform that polarized the party system; in 2006 the manner of
the PF campaign was as notable as its content. As Larmer and Fraser have
argued, ‘Sata’s sheer rudeness about those in “power”allows supporters to
imagine in him potential for subversion and radicalism’.
53
Most infamous-
ly, Sata destroyed a cabbage at a campaign rally, a direct reference to
Mwanawasa, who had been given the nickname ‘the cabbage’following a
debilitating car accident.
54
Although the MMD won the 2006 presidential election by a comfortable
margin, Sata alleged electoral manipulation based on opinion polls which
had predicted a PF victory. As relations between Mwanawasa and Sata
reached an all-time low, the PF leader threatened to use his party’s new-
found urban dominance and strong performance in local elections to estab-
lish a ‘parallel government’that would render the country ungovernable.
Although Mwanawasa publicly dismissed Sata’s claims, he recognized the
strength of the PF threat and set about repositioning the MMD once again
by moving the party closer to the more populist policy space occupied by
the opposition.
55
Most notably, the MMD introduced a series of tax cuts
which Mwanawasa accepted were a direct response to ‘criticism over high
taxes during the election campaign’.
56
As the extent of urban discontent
with foreign investors became clear after a series of protests over the work-
ing conditions in Chinese-run mines, Mwanawasa was also forced to
respond to criticism that foreign investors were not doing enough to con-
tribute to Zambia’s long-term development. Most obviously, the MMD
finally adopted the PF’s hugely popular proposal of introducing windfall tax-
es on copper and increasing mineral royalty tax in order to increase
investment in the social service sector.
57
Consequently, the fierce competi-
52. Larmer, Mineworkers in Zambia; LeBas, ‘Polarization as craft’;Rakner,Political and
Economic Liberalisation in Zambia.
53. Larmer and Fraser, ‘Of cabbages and King Cobra’, p. 613.
54. ‘Zambia's mourning turns political’, BBC New Online, <http://news.bbc.co.uk/1/hi/
world/africa/7582398.stm> (20 January 2009).
55. ‘I will not take a populist approach: Mwanawasa’Lusaka Times, <http://www.lusakatimes.
com/?p=656> (26 May 2009).
56. ‘Mwanawasa to cut taxes after election’,Times of Zambia, <http://www.sadocc.at/news/
2006/2006-142.shtml> (7 March 2009).
57. Interview, Given Lubinda, PF leader, Lusaka, 16 March 2008.
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 65
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

tion between the MMD and the PF in the period 2004–6 had a significant
impact on the MMD party platform, and ultimately on government policy.
The sensitivity of national political leaders to the public mood was sub-
sequently illustrated by the 2008 election, when Mwanawasa’s death led to
competition over the right to appropriate his legacy and so encouraged the
main political parties to converge around the centre ground of Zambian
politics. Despite the many criticisms levelled at Mwanawasa during his
tenure, the period of Mwanawasa’s ill-health saw a positive re-evaluation
of his performance, as newspapers carried numerous features assessing
the positive legacies of his rule.
58
Significantly, the orderly nature of the
Mwanawasa administration stood in stark contrast to the rampant corrup-
tion and disarray of the Chiluba era. At the same time, the combined
impact of significant debt relief and a boom in copper prices looked even
more impressive when compared to the ongoing political and economic
crisis in neighbouring Zimbabwe. Even critical opponents were forced
to admit that ‘[Mwanawasa] was an outspoken person on corruption is-
sues. He did win back some international confidence of investment,
security and positive relations.’
59
The favourable re-evaluation of Mwanawasa’s leadership encouraged
Sata and Rupiah Banda to compete over his legacy. Although their relation-
ship had frequently been hostile, Sata and Mwanawasa had enjoyed an
unlikely rapprochement in May 2008. Following a heart attack on 25 April,
Sata was evacuated to Johannesburg at the expense of the Zambian govern-
ment and on his return the two leaders held a series of closed-door meeting,
resulting in an unexpected reconciliation. Recognizing the widespread sym-
pathy for Mwanawasa, and the prevailing public desire that his economic
achievements not be undermined, Sata sought to manipulate the rapproche-
ment to change his own public image, stressing his respect for the dead
President and toning down his combative manner. The new ‘statesmanlike’
Sata even took to bookending his speeches with appeals to his ‘brother’s’
memory:
‘Greetings people of Zambia. I, Michael Chilufya Sata, was privileged to reconcile with the
late President, Levy Mwanawasa who revealed his vision for the people of Zambia to me. . . .
We have just put to rest our beloved President, his Excellency, Dr Levy Patrick Mwanawasa,
SC. The most fitting and lasting tribute we can render to him is acceleration of the devel-
opment of our great country.’
60
58. Newspapers such as The Times of Zambia were saturated with advertisements wishing the
President a speedy recovery, which in some cases took up the entire page.
59. Fr Miha Drevensek and Fr Umberto Davoli, ‘Zambia –presidential legacies?’(Africa
Files article), <http://www.africafiles.org/article.asp?ID=19083> (1 February 2009).
60. Michael Sata, ‘The way forward 2008’(statement published on Michael Sata’s personal
website), <http://www.michaelsata.co.zm/> (2 May 2009).
AFRICAN AFFAIRS66
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

As part of this remarkable process of re-invention, the PF leader sold him-
self as a member of the political establishment well-placed to offer
continuity and stability, claiming that Mwanawasa had told him, and only
him, of his vision for the country. To make this new approach credible to
Zambian voters, Sata had to move the PF towards the centre ground of the
Zambian political spectrum, watering down the party’s position on a num-
ber of key issues. Although Sata continued to speak out against the presence
of Chinese workers in Zambia, he promised that foreign investors would be
protected. In an apparent U-turn on the party’s policy on Zambia's Asian
population, the PF took out a full-page advertisement entitled ‘The Asian
contribution to Zambia since 1905’which celebrated the role of Asians in
Zambia’s economic and political progress.
61
This approach was far removed
from the PF’s rhetoric in 2006, when Asian, Chinese, Lebanese, and foreign
'infestors' in general bore Sata’s wrath. Although Sata increasingly fell back
on old slogans towards the end of the campaign, the softer PF message in
2008 appears to have inspired a positive re-evaluation of Sata’s presidential
credentials. Prominent commentators suggested that ‘his near-death situa-
tion has refined Sata into a more sober and reliable political leader’.
62
Others suggested that ‘Sata's transformation can be likened to that of Saul
the prosecutor of saints who became Paul, a powerful saint, after a life-
changing encounter on the road to Damascus.’
63
The combination of the
MMD’s adoption of PF policies, and Sata’s strategy of moderation, signif-
icantly reduced the distance between the two parties in 2008.
Sata’s attempt to occupy the centre ground of the Zambian political
system by appropriating Mwanawasa’s legacy did not go uncontested.
His desire to accompany Mwanawasa’s body on a tour of Zambia’s pro-
vinces was rejected and he was banned from attending the funeral by First
Lady Maureen Mwanawasa, prompting a public spat. Political opponents
pointed out that the reconciliation between the two men had occurred
only months before Mwanawasa’s death, while many in the MMD had
worked closely with the President for over a generation. At the same time,
Acting President Rupiah Banda adopted the slogan ‘For Continuity’, ar-
guing that his appointment as Vice-President by Mwanawasa in 2006 was
evidence enough that he was the true ‘chosen successor’.
Of course, local campaigns often deviated from the national narrative and
the significance of Mwanawasa’s legacy varied according to location. How-
ever, the complexity of presidential election campaigns notwithstanding,
Sata’s decision to move the PF’s closer to the centre ground of Zambian
politics in response to the changing public mood following Mwanawasa’s
61. ‘The Asian contribution to Zambia since 1905’,The Post, 30 September 2008.
62. Rev. M. C. Mutambala, ‘Give Sata a chance’,The Post, 22 October 2008.
63. Mahundo’ona Chowa, ‘Uncovering misdeeds’,The Post, 22 October 2008.
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 67
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

death, along with the continual repositioning of the MMD in response to
the PF challenge, demonstrates the way in which the continual interaction
of parties influences the dynamics of election campaigns and ultimately
shapes government policy. The Zambian experience thus provides a neces-
sary corrective to the oft-stated assumptions that debates over policy issues
are wholly absent from African elections and that party competition rarely
results in more responsive government. As Mulafulafu has put it, Zambian
leaders have come to respect the public’s‘preference for an issue-based
campaign'.
64
Lesson three: electioneering and political mobilization
Perhaps the most salient debate within the third wave of party literature
concerns how parties attempt to mobilize support. Are African parties all
‘ethnic’, or do other cleavages structure the political system? Following
Horowitz and Scarritt, an ‘ethnic party’is usually defined as one which ‘de-
rives its support overwhelmingly from an identifiable ethnic group (or
cluster of groups) and serves the interests of that group’.
65
Looking at sup-
port for the ruling party cross-nationally, Norris and Mattes argue that
ethnicitydoesplayakeyroleindeterminingsupportforrulingparties,
but that it is not always the primary cleavage.
66
Expanding this research
to incorporate opposition parties, Cheeseman and Ford suggest that most
ruling parties are actually multi-ethnic in composition, and that ethnicity is
far from a dominant factor in determining political behaviour in countries
such as Mali, Senegal, Zimbabwe, and Botswana.
67
Partly because Posner completed much of his path-breaking research on
ethnic politics in Zambia, the support base of parties such as UNIP, MMD,
and PF has frequently been used as a testing ground for competing hypoth-
eses regarding the politicization of ethnicity in Africa. Posner’sinfluential
early assessment was that ‘class differences may be objectively identifiable
in Zambia, but, except for perhaps on the Copperbelt, they are peripheral
to politics’.
68
Instead, Posner argued that the move to multi-party politics
gave Zambians incentives to invest in larger communal identities that could
feasibly be used as vehicles to compete for state power. Consequently, his
64. S. Mulafulafu is Caritas Zambia executive director and former President of the Founda-
tion for Democratic Progress (FODEP). See The Challenge 10, 4 (2008), p. 12.
65. Horowitz, ‘Ethnic politics in conflict’, p. 237; James Scarritt, ‘The strategic choice of
multiethnic parties in Zambia's dominant and personalist party system’,Commonwealth and
Comparative Politics 44, 2 (2006), pp. 234–56.
66. Pippa Norris and Robert Mattes, ‘Does ethnicity determine support for the ruling party?’
(Afrobarometer Working Paper No. 26, 2003), pp. 1–28.
67. Nic Cheeseman and Robert Ford, ‘Ethnicity as a political cleavage’(Afrobarometer
Working Paper No. 83, 2007), pp. 1–41.
68. Posner, Institutions and Ethnic Politics in Africa, p. 87.
AFRICAN AFFAIRS68
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

analysis concluded that post-1990 political mobilization quickly came to be
restructured in terms of the country’s four main ethno-linguistic blocs:
Bemba, Nyanja, Lozi, and Tonga.
69
Early on, the PF’s heavy reliance on support from Bemba-speakers
appeared to conform to Posner’s model of ethnic politics, leading Scarritt
to classify the party as ‘potentially ethnic’, while Erdmann concluded that
‘PF must also be seen as a Bemba party’.
70
However, Sata’s change of tack
in the run-up to the 2006 election, and his ability to mobilize a broad urban
support base, encouraged qualitative studies of the PF’s approach which
have challenged this characterization. Most notably, Larmer and Fraser
have attributed Sata’s popularity to his ability to give voice to commonly
held frustrations and grievances in Zambia’s historically politicized urban
areas, which cut across ethnic lines.
71
On this interpretation the PF’s appeal
is best understood as a brand of populism founded on the ‘identification of
particular unmet demands of distinct social groups, and their re-presenta-
tion to these groups not only as legitimate but also as aspects of a wider set of
linked and unmet demands, sharing few characteristics beyond their frustra-
tion’.
72
The 2008 election suggests that Posner and Larmer and Fraser both
have part of the answer: the rise and rise of Sata has been underpinned by
the ability to simultaneously recruit ‘ethnic’and ‘populist’support.
Interpreting the Zambian election results of 2008
Assessing the nature of Michael Sata’s support base in 2008 is complicated
by questions regarding the reliability of the official results. Zambia is far
from a consolidated democracy, as the stalled constitutional review process,
state-dominated media, and the use of government resources to support the
MMD’s election campaign ably demonstrate. While the polls were no doubt
far cleaner than the controversial election of 2001, representatives of foreign
governments have admitted in private that they saw evidence of vote buying
on both sides, and have grave doubts about the reliability of the results.
73
As
in previous elections, credible opinion polls suggested that Sata was set for a
substantial victory, and the announcement of his defeat was met with incre-
dulity in urban areas. The refusal of Sata supporters to accept Banda’s
victory was underscored by their partisan preferences, but also reflected
widespread doubt regarding the government’sdemocraticcredentials.
74
69. Posner excluded the 1991 elections on the basis that it was mainly an anti-UNIP vote.
70. Jim Scarritt, ‘The strategic choices of multiethnic parties’; Gero Erdmann, ‘Ethnicity,
voter alignment, and political party affiliation –an African case: Zambia’(Giga Research
Programme, Working Paper No. 45, 2007), p. 24.
71. Larmer and Fraser, ‘Of cabbages and King Cobra’.
72. Ibid., p. 621.
73. Interview, donor representative, Lusaka, 12 November 2008; Interview, Dr Beyani.
74. Drevensek and Davoli, ‘Zambia –presidential legacies?’.
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 69
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

Of particular concern was the absence of a fresh voter registration process
prior to the 2008 election; given that Sata was well supported among urban
youth, this decision is likely to have worked to the advantage of the MMD.
In the absence of survey data at the time of writing, the analysis presented is
based on the official election results and so may underestimate the breadth
and depth of Sata’s real popularity.
The official results suggest that the 2008 election followed a similar path
to the 2006 polls. The MMD retained by far the broadest geographical
spread of support, winning just over 40 percent of the vote and securing a
majority of the presidential ballot in 73 of 150 constituencies (the geograph-
ical unit in which presidential results are declared). In addition to winning a
clear majority of the rural constituencies in Central, Eastern, Western, and
North-Western Province, Banda was the only candidate to win a majority in
constituencies in all eight provinces. The spread of support reflects the
MMD’s ability to utilize the advantages of incumbency to pursue an effec-
tive mixture of electioneering strategies. In 2008 the party continued to rely
on the greater organizational capacity conferred by its monopoly over state
resources and traditional leaders,
75
the co-option of a select number of local
leaders who could be relied upon to deliver problematic constituencies, and
heavy emphasis on the positive re-evaluation of Mwanawasa’s tenure as de-
scribed above.
76
In contrast, the United Party for National Development
(UPND) candidate Hakainde Hichilema remained reliant on the party’s tra-
ditional regional base of support among the Tonga community in Southern
Province, winning 19.7 percent of the vote. Hichilema won only 21 consti-
tuencies, just three of which were located outside of Southern Province
(Table 1).
At first glance, the most striking aspect of the 2008 results was how close
Michael Sata came to winning the presidency despite the absence of a level
playing field. However, Sata’s increased share of the vote in 2008 should
not be interpreted as a sign that he had significantly expanded his support
75. Author observation; see also ‘Rupiah has gone beyond limit’(commentary, Muravi
Blogspot), <http://maravi.blogspot.com/2008/10/rupiah-has-gone-beyond-limit-in-abusing.
html> (24 February 2009).
76. Interview, Given Lubinda, PF leader, Lusaka, 16 March 2008.
Table 1. Zambian presidential elections results, 2008
Party Candidate Constituencies Votes Percentage
MMD Rupiah Banda 73 718,359 40.1%
PF Michael Sata 56 683,150 38.1%
UPND Hakainde Hichilema 21 353,018 19.7%
Heritage Party Godfrey Miyanda 0 13,683 0.76%
Source: Electoral Commission of Zambia.
AFRICAN AFFAIRS70
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

base: voter turn-out was significantly down in 2008 at just 45 percent, and
all parties received fewer votes than in 2006. Sata’s strong performance de-
rived from the ability of the PF to mobilize a similar number of votes in
2008 as compared to 2006, while support for the MMD fell by over
450,000, reflecting the lack of popular enthusiasm for Banda’s candidacy.
Despite this, the durability of the PF’s support base in an election where all
of the party’s rivals struggled to get the vote out remains impressive, and
the gap between Sata and Hichilema demonstrates how much progress
the PF has made towards transforming Zambia into a two-party system.
A populist in Zambia
To analyse the foundations of Sata’s support, and contribute to the debate
regarding whether Zambian politics is purely ‘ethnic’, we break down the
election results according to province and population density. Using data
on constituency size and population provided by the Zambian Electoral
Commission, we classify all Zambian constituencies as either ‘high-density’,
‘medium-density’,or‘low-density’.
77
High-density constituencies are urban
centres, medium-density constituencies tend to be administrative centres
and peri-urban locales, while low-density constituencies are predominantly
rural. However, as our classification does not map directly onto existing clas-
sifications of ‘urban’or ‘rural’Zambia, which usually focus simply on
whether or not a constituency is located along the line of rail, we avoid using
the labels ‘urban’,‘peri-urban’and ‘rural’. It is important to note that while
the number of constituencies a candidate wins is a good indication of the
geographical breadth of their support, this is not the same thing as their over-
all popularity; because constituencies vary greatly in size, a candidate may
win the election without winning a majority of constituencies.
If the PF’s support were predominantly ‘ethnic’, we would expect the party
to win votes in Bemba-speaking provinces (Copperbelt, Northern, Luapula)
and to receive little support in constituencies in other parts of the country.
Conversely, if the PF’s support were predominantly ‘populist’and based on
mobilizing disenfranchised urban militants, we would expect to see most of
Sata’s votes coming from a broad range of high-density constituencies in
which trade unions and urban political movements operate, with little sup-
port deriving from low-density constituencies. Of course, analysing election
results based on constituency type runs the risk of the ecological fallacy: we
cannot be sure exactly who is voting for a candidate within a given constitu-
ency, and beyond that we cannot be sure of exactly why individuals are
voting. However, in the absence of survey data that would provide a more
77. Low-density < 30 people per square kilometer; medium-density < 250; high-density >
250. This classification is somewhat arbitrary but was chosen as it was found to broadly reflect
the understanding of party activists, survey workers, and demographers.
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 71
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

precise insight into voting patterns, constituency-level data represent a useful
resource when used sensitively and in conjunction with feedback from re-
search assistants and interviewees.
Broken down by constituency, the election results suggest that Sata’s
support base is both ‘ethnic’and ‘populist’. On the one hand, as Table 2
shows, Sata won all 13 constituencies in high-density areas, sweeping up
in both Bemba-speaking and non-Bemba-speaking areas. In 2008, Sata al-
so performed strongly in medium-density areas, winning a majority of these
constituencies (18 to Banda’s 17) and gaining a number of constituencies
he failed to win in 2006, including Chilanga, Chipata Central, Mpulungu,
Kapiri Mposhi, Solwezi, and Kafue (see Table 3). It is significant that Sata
also won victories in the MMD’s heartlands of Central and Eastern Prov-
ince, winning the Bwacha and Kabwe Central constituencies in Central
Province, both of which had returned MMD MPs in 2006. The geograph-
ical expansion of Sata’s support suggests that the PF’s strategy of holding
mass rallies in all provinces, and visiting all major provincial and district
centres for the first time in order to spread the PF message to new urban
and peri-urban settings, met with some success.
Significantly, in these urban areas the PF eschewed an ethnic rhetoric
and focused instead on popular concerns and local needs. Interviews and
Table 2. Presidential election results 2008, by type of constituency
Type of constituency Banda Hichilema Sata Miyanda Totals
Low-density 56 16 26 0 98
Medium-density 17 4 18 0 39
High-density 0 0 13 0 13
Totals 73 20 57 0 150
Table 3. Constituencies won by Michael Sata in 2008, by constituency type
and province
Province Low-density Medium-density High-density Total
Central 0 2 0 14
Copperbelt 0 11 6 21
Eastern 0 1 0 20
Luapula 10 4 0 16
Lusaka 0 0 7 10
Northern 16 0 0 21
North-Western 0 0 0 12
Southern 0 0 0 19
Western 0 0 0 17
Total 26 18 13 150
AFRICAN AFFAIRS72
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

feedback from research assistants suggest that the economic downturn ex-
perienced by the country in 2008, and in particular rapid inflation which
increased the price of food, fuel, and fertilizer, enhanced the effectiveness
of Sata’s message. Based on the slogans ‘For Lower Taxes, More Jobs’and
‘More Money in Your Pocket’, the PF sought to focus attention on the eco-
nomic difficulties facing the poorly paid and the unemployed. Significantly,
while all parties spoke out on this issue it was only the PF which had cam-
paigned on these same issues in 2006 and hence could credibly present
itself as the defender of the poor and disadvantaged. Perhaps the best evi-
dence that Sata’s message hit home is that many affluent voters in formal
employment were put offby the radical potential of Sata’s rhetoric. Indeed,
in a number of different urban constituencies research assistants found
that, with the exception of the mineworkers, most Zambians lucky enough
to have jobs, especially white-collar workers, rejected the PF because of a
fear of economic instability. As one interviewee reported, ‘The talk about
the Chinese and anti-foreign investors actually alienated people . . . they felt
insecure about it.’
78
It is therefore important to note that the core of Sata’s support base in
urban areas appears to lie not so much among those in full-time paid wage
employment, but among more radical mineworkers and the unemployed in
Zambia’s poorest urban compounds. The unquestioned support of these
groups, who have little to lose from the alienation of foreign investors
and everything to gain from the promise of ‘jobs for Zambians’, suggests
that in high- and medium-density constituencies Sata’s message resonates
less with the mood on the street than with the mood in the slums.
However, in addition to dominating high-density constituencies, Sata re-
ceived a significant proportion of his support from low-density Bemba-
speaking constituencies in rural areas, suggesting the continued existence
of an ‘ethnic’component to his support base. As shown in Table 3, Sata
won 26 of the 98 low-density constituencies, and these predominantly rural
constituencies contributed 38 percent of his votes, compared to the 30 per-
cent from medium-density constituencies, and the 32 percent from high-
density constituencies. Significantly, Sata’s ability to win in these rural con-
stituencies has little to do with his ‘populist’message or urban credentials.
Indeed, the policies advocated by the PF in 2008 would have been detri-
mental to the predominantly agricultural interests in these areas. Most
obviously, the PF’s promise to lower food prices in 2008, while a great
boon to urban consumers, would have lowered takings for rural producers.
Sata even made a number of faux-pas during the campaign which might
have been expected to undermine any rural support base, ridiculing the
78. Interview, V. Katanekwa, Director of the Livingstone Museum, Livingstone, 8 November
2008.
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 73
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

backwardness of Zambian 'stone-age' farming practices. As Gould noted in
his analysis of the 2006 election, the fact that the ‘Bembaphone’North of
the country is willing to support Sata despite his seeming lack of concern for
rural issues and the predominantly urban focus of his campaign is an im-
portant reminder of the continued power of ethnicity in Zambian politics.
79
That Sata only won low-density constituencies in predominantly Bemba-
speaking provinces (Luapula, Northern) indicates that his ability to gener-
ate support in rural areas is derived from his status as a Bemba-speaking
presidential candidate able to exploit the type of group ties described by
Horowitz and Posner. It is no coincidence that in both 2006 and 2008,
while consciously pushing a ‘populist’message, Sata also delivered the vast
majority of his speeches in the Bemba language.
80
Significantly, Erdmann
estimates that, even in 2006, Bemba voters made up some 60 percent of the
PF’s support base, and Bemba-dominated constituencies made up the ma-
jority of the PF’s parliamentary haul.
81
In short, the ‘ethnic’element of the PF support base initially identified by
Scarritt remains, and now sits happily side by side with Sata’s cross-ethnic
support base in urban areas. It is worth noting that the proportion of Sata’s
support derived from each constituency ‘type’has varied little over the
three elections the PF has fought, as shown in Table 4. The emergence
of the PF as a political force in the period 2006–8 cannot therefore be at-
tributed solely to the rapid expansion of its urban support base; Sata’s
support among low-density constituencies in his ethnic heartlands has ex-
panded to keep pace with his growing popularity elsewhere.
The 2008 election suggests that Posner and Larmer and Fraser are both
half-right. Sata does appear to have created a durable cross-ethnic support
base which has expanded beyond the mining areas of the Copperbelt into
medium-density constituencies in a number of provinces. Whether or not
Sata believes in his populist message, it resonates in the mines and in the
slums, and as the economy hits hard times its popularity is likely to grow. In
the light of this evidence, there is a pressing need to revise Posner’s conten-
Table 4. Location of support for Sata in presidential polls 2001–8
Constituency type 2001 (%) 2006 (%) 2008 (%)
Low-density 39.09 34.35 37.98
Medium-density 31.07 30.56 30.42
High-density 29.84 35.09 31.60
79. Gould, ‘Zambia’s 2006 elections’.
80. The authors are grateful to Rachel Beatty Reidl for this point.
81. Erdmann, ‘Ethnicity, voter alignment’, p. 23.
AFRICAN AFFAIRS74
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

tion that class only plays a significant role in Zambian politics on the Cop-
perbelt. However, it is also clear that not all of Sata’s support derives from
his populist positioning. Larmer and Fraser’s interpretation must make
space for the fact that over a third of Sata’s votes in 2008 most likely came
from ethnic partisans in Bemba-speaking low-density constituencies. The
great success of Sata’s campaign in 2008 was that he managed to mobilize
these two bases simultaneously, despite the potential conflict of interests
lying at the heart of such an alliance. Zambian politics is not an ethnic cen-
sus, but neither is it accurately portrayed as class struggle. The literature on
elections would do well to learn the lesson that parties need not be simply
‘ethnic’or ‘populist’, and are often most successful when they utilize di-
verse strategies of political mobilization.
Party politics in Zambia: towards institutionalization?
The Zambian election of 2008 reveals that an understanding of party pol-
itics lays the foundation for an understanding of electoral politics more
generally. Rupiah Banda’s rise to power had little to do with his personal
appeal, wealth, or ethnicity, and can only be explained with reference to
internal MMD politics and the party’s ability to use the advantages of in-
cumbency to sell a largely uninspiring candidate. Furthermore, the party
platforms adopted by the PF and the MMD in the run-up to the 2008 polls
were not ideologically vacuous, nor did they simply reflect the interests of a
distinct ethnic support base. Instead, the PF and MMD debated the appro-
priate role of foreign investors in economic policy and the standard of living
owed to urban workers, while carefully assimilating their rivals’most pop-
ular policy proposals. Finally, the rise of Michael Sata shows that election
campaigns and mobilization strategies adopted by Zambian parties cannot
be simply generalized into one model of ‘ethnic’or ‘populist’politics; in
2008, the PF campaign effectively combined both.
However, the ability of Sata to win power in the future, and the likeli-
hood that the PF will survive its leader as a political force in Zambian
politics, depends on whether he can embed his appeal in a more effective
party organization. Sata is a classic populist in the structure of his political
organization: the PF is heavily dependent on the force of Sata’s personality
and there is no clear indication as to who will lead the party after him, or
what would keep the party united. The PF’s predicament reflects the low
level of party-system institutionalization in Zambia more generally. While
the MMD has won every election since the return to multi-party politics, it
is only in the last four years that the top two opposition parties have re-
mained the same between elections. Although the MMD is able to carry
Banda because of its access to state resources, the electoral chances of op-
position parties rarely survive a change of leadership. On Kuenzi and
THE ZAMBIAN PRESIDENTIAL ELECTIONS OF 2008 75
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from

Lambright’s0–10 scale of party-system institutionalization (with 10 repre-
senting the highest level of institutionalization), Zambia scores just 5,
making it the 23rd least institutionalized country out of 30 African cases.
82
The weak institutionalization of the party system is significant, as it is
widely believed that ‘a high level of institutionalization fosters democratic
consolidation’.
83
Afluid party system with rapid changes in composition
may undermine the ability of the political system to integrate diverse com-
munities and to reach stable and mutually acceptable bargains over the
priorities for government policy. In the Zambian context, the limited for-
malization of party structures has two main roots. First, opposition parties
struggle to raise the funds necessary to contest elections from their suppor-
ters, many of whom are desperately poor. They are therefore dependent on
their leaders for election resources, a dependence which undermines the
evolution of more democratic governance structures.
84
Second, the lack
of ideological cohesion within most parties, combined with intense compe-
tition over leadership positions, makes it extremely difficult to hold
elections for party posts without triggering splits within the party. From
UNIP onwards, parties have tended to fragment along factional and ethnic
lines whenever the question of the internal distribution of power is raised.
In the case of the PF, Sata has attracted a broad range of cadres including
former MMD colleagues, representatives of urban communities, and indi-
viduals who left the UPND following the death of former leader Anderson
Mazoka. Although Sata has recognized the value of populist political rhet-
oric, and his PF colleagues are more than happy to ride on his bandwagon,
many joined the party because they were in need of a political home and
appear to lack a deep commitment to the PF’s stated ideological position.
Sata’s reluctance to hold elections for party posts or to announce a chosen
successor can thus be interpreted as a tacit recognition of the fact that with-
in the PF internal disunity is the norm, rather than the exception. Most
recently, the party had to dismiss sixteen ‘rebel’MPs who, against Sata’s in-
structions, agreed to take part in the National Constitutional Council which
had been established by the MMD to investigate constitutional reform, but
was dismissed as a sham by senior PF leaders.
85
In the absence of radical
changes to political party funding and the emergence of a new generation
of genuinely committed populist politicians, the continued stability of the
PF, and of the Zambian party system more generally, remains uncertain.
82. Michelle Kuenzi and Gina Lambright, ‘Party system institutionalization in 30 African
countries’,Party Politics 7, 4 (2001), pp. 437–68.
83. This opinion predominates within the literature on party system institutionalization in
Latin America –see ibid., p. 438.
84. Rakner, ‘Political liberalization in Zambia’.
85. ‘16 expelled MPs no longer ours –Lubinda’,Lusaka Times, <www.lusakatimes.com/?
p=6861> (27 May 2009).
AFRICAN AFFAIRS76
at Bodleian Library on December 29, 2014http://afraf.oxfordjournals.org/Downloaded from