Research Proposal
Questioning Ethnic Politics in Sierra Leone
Lydia L. Lundgren*
PhD Student, Political Science
University of California, San Diego
*I appreciate the guidance I’ve received from Clark Gibson and Karen Ferree, as well as helpful
suggestions from Danielle Jung, Sarah Knoesen, Jaime Settle, and Nick Warner. I am also
thankful to Ted Miguel, Liz Foster, Tristan Reed and the IRCBP for their generosity and help in
providing access to their datasets. All mistakes are my own.
This is a draft. Please do not circulate or cite.
Introduction
In 2007, some academics in Sierra Leone were quick to refer to election campaigning in
their country as exhibitions of ethnic politics. On August 26, 2007, Reuters published their
analyses on the Presidential and Parliamentary elections in an article entitled, “Tribal Voting.”1
Joe Ali, a history professor at Freetown’s Fourah Bay College was quoted in the article: “‘[the
political parties] have simply preached naked ethnic politics.’”2 Dr. Gbla concluded:
“‘Politicians say if you help your kinsmen you will survive; we will give you jobs, opportunities
and education. That’s why people vote along their tribe lines.’”3
Yet, Gbla also admits there are exceptions. He claims: “‘The majority of traditional
Mendes would rather decide to vote with their people [SLPP party], but if Margai [leader of the
PMDC] plays his cards well he can get 50 percent of them.’”4 Getting 50% of the Mende who
support the SLPP, of course, required splitting a political party that has claimed a support base
among the Mende people for four decades. And, in fact, Margai, was successful in this attempt
in certain locations. In addition to this surprise split within the SLPP, through interviews in
Freetown, Christensen and Utas (2007) heard a great deal about “Watermelon Politics.”
Watermelon politics refers mainly to youth in Freetown, who pretend to support the SLPP
(whose color is green) while also supporting the APC (whose color is red).5 Thus, although
ethnic politics is a prominent issue discussed in politics and utilized in political mobilization in
the country, it also seems to be a complicated issue involving two perspectives—that which is
advertised and that which is realized. This paper does not intend to challenge the statements of
the Gbla or Ali, above, regarding the methods by which politicians seeks to garner votes. Rather,
1
Manson (2007) 1.
Manson (2007) 1.
3
Manson (2007) 2.
4
Manson (2007) 1.
5
Christensen and Utas (2008) 530.
2
1
aside from what is spoken or written in propaganda in Sierra Leone regarding ethnicity and
politics, this paper seeks to explore whether ethnic politics takes the form of ethnic distribution
and/or voting.
The role of ethnicity6 in political development in Africa is not a recent subject of concern
in social science literature. (Bates, 1983; Horowitz, 1985) Recently, however, it has been
suggested that the role of ethnicity in politics is a complicated issue and varies between countries
as well as within them. (Cheeseman and Ford, 2008; Eifert, Miguel and Posner, 2007; Norris and
Mattes, 2003; Posner and Simon, 2002) Ethnicity, then, is part, but not necessarily the whole
story that explains political processes in Africa, and in Sierra Leone.
This paper asks two sets of questions. First: Does ruling party ethnicity affect provision
of goods and services, and perceptions of government among ethnic groups of same (or
different) ethnicity, and does it matter if it is a local or central government representative?
Further, are there alternative cleavages along which provision seems to be focused other than, or
in addition to, ethnicity? Second: Do perceptions, or changes in perceptions, of economy
(personal and national) or governance (local and central) among different ethnic groups predict
levels of vote share for an incumbent or opposition party? Or are ascriptive characteristics—such
as ethnicity—better predictors of political party support?
These questions will be developed and examined in this paper. First, I present the
theories in the literature on ethnic politics regarding both distribution of goods and services, and
voting. Second, I explain why Sierra Leone offers an excellent case study from which to derive
and test hypotheses on ethnic politics that may be applied outside the country. In the third
6
Although there is debate over what constitutes ethnicity, it is simply used in this paper as Horowitz (1985) defines
it as inherited. (83) and an attribute of common origin.” (41) In Sierra Leone, certain ethnic groups, especially the
Mende, Temne, Limba and Kono, hold significant meaning both in current politics as well as in the political and
social history of the country.
2
section, I present the data I plan to gather as well as the methods I plan to use in testing these
hypotheses.
Finally I offer three concluding remarks on possible findings. The first is whether Sierra
Leone is truly based on patronage networks, or whether the promises of patronage never realize
themselves. Perhaps in very poor countries, such as Sierra Leone, it is impossible to spread
benefits everywhere promised, and if so, there may be costs to pay on election day. Second, this
paper addresses theories on ethnic voting as well as whether the use of ethnicity by voters may
ever be evaluative. This paper supports the argument that ethnicity may be evaluative in certain
contexts, and seeks to test whether this may be the case. Finally, this paper questions whether
weakening the psychological benefit of voting for a co-ethnic, the ethnic voting theory most
often associated with Horowitz (1985), is possible, and if so, whether it produces unexpected
outcomes—such as voting for a non co-ethnic, or consideration other factors, such as the
economy.
I. Theories of Ethnic Politics
If voting is determined by demographic rather than democratic considerations, then the
quality of the democracy may be challenged. Horowitz (1985) acknowledges that in numerous
countries, even established democracies, party preferences are likely to be associated with ethnic
affiliations.7 However, it is not the support granted by an ethnic group that is necessarily
problematic, but rather the method by which that support is garnered and the expectations
associated with that support. As Sartori summarizes: “If a party is not a part capable of
governing for the sake of the whole...then it does not differ from a faction. Although a party only
7
Horowitz (1985) 295.
3
represents a part, this part must take a non-partial approach to the whole.”8 Focusing resources
and benefits on only one part, or ethnic group(s), rejects the democratic purpose of political
parties, and thwarts the development of national policy. Thus, the extent of expectation—the
relative service provided by a representative to co-ethnics—is in part what constitutes the
difference between ethnic voting that is acceptable and that which is objectionable. In addition,
if Horowitz (1985) is correct in stating that if politicians can count, they have something to count
on, then voting lacks evaluation of options and/or quality of representation. This, too, challenges
the purpose of checks and balances in democracy through elections. The following section will
address the role of evaluation and expectation in various theories on ethnic voting, as well as
assess whether the various theories are compatible with economic voting theories that may also
explain the evolving political process in Sierra Leone.
Part 1: Expressive and Rational Theories of Ethnic Politics
Two predominant theories regarding ethnic voting attempt to explain the reasons why
ethnicity is salient in politics. Horowitz (1985) understood voting for a co-ethnic to have
intrinsic value, or psychological value; thus voting is an expressive act of one’s identity. In this
case, then, the expectation is that voting for a co-ethnic will result in a psychological benefit. On
the other hand, others suggest ethnic voting may be a rational pursuit of one’s best interest.
Chandra (2004) presents such an instrumental (rational) ethnic voting theory where the voter
expects a material benefit from electing a co-ethnic.
Ferree (2004) offers a third variation of ethnic voting that suggests ethnicity provides
informational cues in low-information contexts. This is particularly relevant in many African
countries where technology is limited and information is often transmitted by word of mouth,
8
Sartori (2005) 23.
4
especially in rural areas. Establishing credible lines of communication may build on existing
mechanisms, such as ethnic groups. Chandra (2004) also addresses this problem of credible
information in new democracies, where ethnicity provides a means by which electoral decisions
may be made in situations of limited information. In these cases, then, the expectation is that
ethnicity provides a credible proxy for information either unavailable or not provided, in order to
evaluate choices and make better-informed decisions.
Lindberg and Morrison (2008), however, challenge the idea that ethnicity is evaluative.
They make an argument regarding evaluative and non-evaluative voting behaviors, asking “Are
[African voters] taking ‘primordial’ shortcuts, such as ethnic or family ties to candidates, using
patron-client relations to exchange their vote for individual favors, or are they using moreevaluative reasoning in respect to programs and performance?”9 This categorization of ethnicity
as non-evaluative, however, only addresses the expressive and patronage arguments. Perhaps
ethnic voting is not purely non-evaluative. This paper will return to this question in the
concluding remarks.
The expectations associated with each of these theories are also important, for they each
also invoke certain requirements. For the expressive and informational proxy theories, the ethnic
party in question must present certain characteristics that provide for positive psychological
benefits or evidence of credibility. For example, Horowitz (1985) claims that people do not
simply, blindly, follow their leaders. Rather, when a party begins to pursue a direction that is not
supported by its followers, thus threatening psychological benefits, intraparty alternatives may
arise in the form of alternative leadership.10 Ferree (2004), too, questions “under what
conditions” is it possible for the beliefs, or “ethnic images”—that make information proxies
9
Lindberg and Morrison (2008) 100.
Horowitz (1985) 343.
10
5
possible—to change.11 Ethnic voting, then, even when expressive or informational, is not
necessarily static. It depends upon certain actions on behalf of the political leaders and reactions
on behalf of the voters. Where these behaviors conflict or do not correspond, ethnic voting
patterns may weaken.
The rational theories of ethnic voting also rely on expectations. Chandra (2004)
develops the idea of “patronage democracies,” in which the state monopolizes access to jobs and
resources as well as discretion over the laws regarding those jobs and services, making proximity
to politicians a necessity.12 She contends that in a patronage democracy, “selective targeting”13
may occur, where a group or groups are selected to receive most or all of the benefits from state
resources. Cox and McCubbins (1986) also address a type of selective targeting in non-patronage
democracies. They find that political parties will reward their “reelection constituency,” which,
in ethnic politics, would be the ethnic group(s) forming a political party’s base. If the ethnic
party in power can maintain a secure majority by doing so, it is reasonable for a political party to
use “selective targeting” to reward only its ethnic group(s), or its “reelection constituency”, and
restrict services to other groups.14 Selective targeting of ethnic groups in the distribution of
public sector resources is not unique to developing countries. In describing the machine politics
of the 19th and 20th centuries in the United States, Erie (1988) suggests that secure machine
governments concentrated on rewarding their core supporters (often an ethnic group such as the
Irish) finding “little electoral incentive to mobilize newer ethnic arrivals [immigrants].”15 The
comments of Ali and Gbla above suggest that such a selective targeting is promised by Sierra
11
Ferree (2004)
Chandra (2004) 6, 12. Note: To qualify as a patronage democracy, there must exist a large public sector, not to be
threatened through competition from a private sector.
13
Chandra (2004) 61.
14
Chandra (2004) 62.
15
Erie (1988) 10.
12
6
Leonean politicians. Stokes (2005), however, offers an alternative to focusing benefits on core
supporters.
Stokes suggests that when “core constituents” are “ideologically committed” then they
have little bargaining power and little incentive to leave their political party, even when their
party is not responsive to their needs.16 Ideology could mean a political orientation or an ethnic
loyalty, inherent to expressive ethnic voting, for example. If such a condition is present, Stokes
contends leaders are likely to focus on “sway voters”, or those in the middle of the distribution.17
However, this depends on the closeness of the two parties, and the ease with which these voters
in the middle could choose between parties. These sway voters may be non-co-ethnics who do
not see a major difference between the two dominant parties, and thus can be free to choose
between them. On the other hand, sway voters may be more than simple constituents. It is
possible that sway voters are members of society who hold great sway over a certain voting
block. For example, van de Walle contends politicians will not always reward in-group
constituencies, but rather focus on “ethnic brokers and regional elites.”18 Thus, it is possible to
find that the elite of all or some ethnic groups may receive the benefits that otherwise would
have been provided to co-ethnics of the ethnic party. Selective targeting, however, can have
negative consequences. Stokes (2005) suggests certain constituents may be ignored by
patronage machines, in which “not all poor voters receive benefits.”19 This kind of selective or
elite targeting is especially critical in poor countries such as Sierra Leone, where the poverty
level is high and thus further depletion of resources may be dire.
16
Stokes (2005) 315.
Stokes (2005) 325.
18
Van de Walle (2007) 65.
19
Stokes (2005) 315.
17
7
In examining the success of patronage politics or selective targeting, it is important to
determine whether a certain ethnic group has or has not enjoyed benefits, and whether they
notice a decline or lack of increase in their economic status. Second, has another group enjoyed
selective targeting? Finally, considering the voter’s response to their own or the national
economic condition may help to establish whether they are in fact expressive voters—at times
acting against their own best interest—or if it is possible that their decisions are more complex.
In order to examine this question in detail, it will be necessary to examine economic voting
theories and establish whether and when economic voting is compatible or conflicts with ethnic
voting.
Part 2: Theories on Economic Voting
Perhaps a useful means by which to conceptualize the compatibility of ethnic and
economic voting is found in Fiorina’s (1976) work on consistent voters and cross-pressured
voters, which highlights the possibility of discord between political identification and political
interests.20 Here, the consistent voters are those whose party offers the more attractive platform,
and cross-pressured are those whose party does not do so.21 For Fiorina, party identification
served as an expressive factor for the voter and party platform or a party’s stated objectives acted
as a rational consideration.22 For Fiorina, then, expressive voting may, or may not, be
sympathetic to instrumental voting, just as ethnic voting may or may not be sympathetic to
economic voting. These dilemmas are not unique to developing countries. The question to ask,
then, is whether there are examples of cross-pressured voters in Africa—and if so, how do they
vote?
20
Fiorina (1976) 396.
Fiorina (1976) 396.
22
Fiorina (1976) 394.
21
8
There are several theories on economic voting. Harper (1999) contends economic voter
theories in general have assessed whether voters are pocketbook or sociotropic voters and/or
whether they vote prospectively or retrospectively. However, regarding pocketbook and
sociotropic voting, Kinder and Kiewiet (1981) argue that a voter’s personal financial status may
be linked, in their mind, to the national economy, and that their assessment of the national
economy from this perspective may not be entirely comprehensive. Thus, separating these two
types is perhaps more useful in theory than in reality, and does not clearly speak to the issue of
cross-pressured or consistent voting.
On the other hand, recent theories have argued for a more complex understanding of, or
variations on, economic voting, utilizing the ideas of retrospective and prospective voting. For
example, Stokes (2001) argues that retrospective and prospective perspectives will interact with
a voter’s support for or opposition to the incumbent government. In sum, she proposes
variations of economic voting, depending on whether the voter is optimistic or pessimistic about
the future of the economy and whether they support or oppose the government in response to
these views.23 She and Maravall and Pzerworski (2001) examine these four combinations:
normal retrospective (opposes and is pessimistic); intertemporal (supports and is optimistic);
exonerative (supports and is pessimistic); and oppositionist (opposes and is optimistic).24
These categories help to establish whether and when it is possible for ethnic voters to be
economic voters. One reason it may be so difficult to provide evidence that African voters are
also economic voters is that many of these theories are compatible with ethnic voting, and thus it
is difficult to tease the two apart for voters may simply be consistent. For example, oppositionist
voters, who according to Maravall and Przeworski (2001) “were overwhelmingly influenced by
23
24
Stokes (2001) 17.
Stokes (2001) 17 and Maravall and Przeworski (2001) 70.
9
noneconomic considerations,”25 may in fact be informed by ethnic voting. Further, exonerative
and intertemporal also seem compatible with ethnic voting. For exonerative, Maravall and
Przeworski suggest that voters either do not see the government as responsible for the state of the
economy or they simply find the incumbent government to be the better alternative.26 The
intertemporal voters weigh their future expectations more heavily than their present pains, and
thus believe the incumbent government will lead to a better future.27 Both allow for ethnic
considerations, ultimately preferring a co-ethnic, and economic considerations.
Only the normal retrospective voter, who would otherwise support the incumbent, seems
to present an example of cross-pressured voters, for whom economic voting conflicts with (and
possibly trumps) ethnic voting. Neither an ethnic expressive nor an ethnic patronage voter
would likely vote against their party if it is the incumbent. Thus, the question that is relevant to
any poor country that has either made no progress in development or has experienced economic
declines: do people who are assumed to be ethnic voters ever vote as normal retrospective
voters? Perhaps, as Magaloni (2006) describes an experience among PRI’s supporters in
Mexico, ethnic voters may “learn to distrust” 28 practices and promises of an incumbent
government, and thus vote for an opposition party. On the other hand, voters may simply stay
away from the polls. Posner and Simon (2002) find that abstention is the more likely choice for
voters who respond negatively to economic changes but are unwilling to support the opposition.
In addition, they also find that these negative responses are a result of a decline in economic
25
Maravall and Przeworski (2001) 70.
Maravall and Przeworski (2001) 66.
27
Maravall and Przeworski (2001) 66.
28
Magaloni (2006) 174.
26
10
conditions, not simply consistently poor conditions. 29 In these cases, voters would be conflicted
by ethnic and economic considerations, and would be cross-pressured.
These theories provide the motivations behind the hypotheses regarding Sierra Leone
described in the following section.
II. Ethnic Politics and Voting in Sierra Leone, Hypotheses
Sierra Leone provides an excellent case study for these questions regarding ethnic politics
for the following reasons. First, the country has been divided ethnically, and by region, since
independence in 1961. The Mende and Temne, two of the thirteen ethnic groups in Sierra
Leone, each comprise approximately 30% of the population.30 The equivalent group size of
these two ethnicities, as well as their regional separation lends to their mobilization as political
support bases. Posner (2006) and Ferree (2008) both suggest ways in which the sizes of ethnic
groups interact with the institutional contexts, specifically electoral rules, and result in varied
patterns of ethnic coordination or mobilization. In addition, Scarrit and Mazaffar (2005)
highlight the importance of geographic distribution of ethnic groups in determining the number
of democratic political parties in Africa. Sierra Leone requires 55% majority of total popular
vote for Presidential races, with a popular majority rule for any runoff.31 Although a coordination
between the Temne and the Mende would mean dominating the country politically, the animosity
between these groups stemming from independence32, as well as their regional separation
29
Posner and Simon (2002) 330.
Jackson (2005) 50.
31
Sierra Leone began its independence as a Parliamentary system. However, in 1971, Siaka Stevens, a member of
the APC, declared Sierra Leone a Republic and named himself its first President. The state has continued to operate
as a Republic since, though plagued with intermittent military regimes. The rule for 55% majority is recorded in the
1991 Constitution, Chapter 5:42.
32
Kandeh (1992) provides some context to the development of the animosity that began to grow between the
Mende and the Temne in the mid-twentieth century. Under colonial rule, Milton Margai had been named Prime
Minister in 1956. In 1962, he became the first Prime Minister of an independent Sierra Leone. However, the fact
that Milton Margai was Mende threatened the Temne and Kono leaders. In order to quell the opposition, Margai
30
11
(Mende in the South, Temne in the North), prevents such coordination. Thus, Mende and Temne
need to coordinate with other smaller ethnic groups—such as the Limba, who constitute about
8.4% of the northern population, and have traditionally coordinated with Temne .33
Secondly, the two political parties that have survived since independence and ruled the
country at different times34, the Sierra Leone People’s Party (SLPP) and the All People’s
Congress (APC), claim support bases among the Mende and Temne, respectively. In fact, both
originated in response to what they interpreted as bias toward another ethnic group. Regional
elites coordinated to confront the Creole hegemony during colonialism through the formation of
the SLPP, and northern politicians formed the APC in reaction to the dominance of Mende
within the SLPP.35
Finally, in the seven years since the civil war ended, both the SLPP and the APC have
ruled; the SLPP from 2002-2007, and the APC from 2007-present. This transfer of power—
which has been for the most part peaceful—provides both an interesting examination of
distribution, and whether it changed in any noticeable ways after 2007, as well as causal factors
underpinning the shift in power from the SLPP to the APC in 2007.
Provision—Patterns and Prospects
selected certain leaders from these groups to sit in his cabinet. However, the relative poverty in the Temnedominated regions in the north and the Kono district in the east did not ameliorate concerns of the residents in these
areas regarding their future under a Mende leader. The lack of resources to spread around became a major reason to
“capture the state” among the elite political leaders who were not part of the ruling party. Thus, although Margai had
made concessions to the other ethnic groups, they were insufficient in the eyes of the Temne.
33
Bangura (2000) 567.
SLPP: 1961-1967, 1996-2007; APC: 1967-1996 (Though with military regimes interspersed)
35
Kandeh (2003) 196-7.
34
12
To reiterate one of the two main questions examined in this paper: Does ruling party
ethnicity affect provision of goods and services, and perceptions of government among ethnic
groups of same (or different) ethnicity, and does it matter if it is a local or central government
representative? Up to this point, however, it is theories of distribution, not perceptions among
the constituents regarding distribution, that have been discussed. Both are important. This paper
examines what constituents have access to specific public goods or public sector employment, as
well as what their perceptions are regarding their access to public goods. In areas where
information regarding one’s relative economic status is limited, analyzing both actual and
perceived levels of provision is important to establish whether there is any difference between
relative national provision and personal perceived provision.
The benefit of there being a change in government in 2007 allows testing whether there is
any noticeable difference between the SLPP and the APC as a ruling party. Thus, for both
regimes, the hypothesis is:
H1: If ruling party (local, national) is representative of an ethnic group(s), then members of that
ethnic group(s) ought to have higher perceptions of government and economic factors.
It is also important to control for the source of provision of goods and services—whether
from the central government, local government, donors, or another provider.
A Local Government system was implemented in 2004 through the Local Government Act. In
addition, a local government initiative has been undertaken by the Sierra Leone government and
international actors to strengthen the capacity of these local government bodies.36 Thus, if certain
36
Organizations involved include: USAID (See www.usaid.gov/policy/budget/cbj2006/afr/pdf/sl636-002.pdf);
Institutional Reform and Capacity Building Project of Sierra Leone Government and World Bank (See
http://www.ircbp.sl/drwebsite/publish/index.shtml)
13
ethnic groups are privileged, then there may be an evening out of such inequity as local
government becomes more capable of providing for its constituents and distribution is more
equitable. The extent to which local government is seen to be central in the lives of citizens will
likely be revealed through the surveys conducted in 2005, 2007 and 2008, discussed in the final
section on the data used in this paper. In addition, the local government is part of the same party
system as the central government—thus it is important to test various combinations of ruling
party influence. It may be different to have a local government ruled by APC while the central
government is ruled by the SLPP, for example.
A possible threat to such success for the local government initiative stems from a
traditional system of local governance dating to colonialism: the chiefdom system. There are
153 chiefdoms in Sierra Leone and each of the thirteen districts has one Paramount Chief (PC)
who sits in the Parliament and maintains the same powers as other Ministers of Parliament (MP),
though he is not popularly elected. Rather, chiefs are elected from and by ruling families within
Sierra Leone. Traditionally, chiefs have been responsible for dispute resolution and tax
collection. 37 Each chief has a speaker who also serves as his deputy; and each chiefdom has a
committee. The committee must include a treasury clerk, bailiff, and court functionaries
(including police).38 In addition, although it was decreed by Siaka Stevens in 1967 that the
chiefs would have no power over granting mining licenses in their chiefdoms, in 2003, President
Kabbah announced he was reinstating their right.39 Thus, chiefdoms operate as a local governing
structure as well as arbiter of access to valuable resources in Sierra Leone. Thus, some conflict
has arisen in certain locations between official local government and chiefs, especially in regard
to the use of revenue from collected taxes. The chiefdom system, however situated vis-à-vis the
37
Hanlon (2005) 462.
Jackson (2006) 98-9.
39
Kabbah (2003)
38
14
local government, is not necessarily disconnected from politics at large within Sierra Leone.
Specifically, chiefs are not considered apolitical. 40 They have reportedly declared that no
independents would be allowed to stand for elections in their chiefdoms, rather political
nominations had to be obtained.41 Jackson (2006) suggests this is reflective of the chiefs’
relationships with party leaders, not their endorsement of the importance of political parties.
Toward that end, political parties may manipulate chiefs. For example, in 2004, during the local
elections, it was suggested that SLPP threatened depose chiefs who did not support the ruling
party in local elections.42
The chiefs are not only victim to hierarchical power, however, they are sometimes part of
perpetuating it. Over the years, the youth had become subject to what Jackson (2006) described,
the gerontocracy, and were forced to pay arbitrary fines to the chiefs.43 There persisted a general
feeling among Sierra Leoneans that the chiefs themselves were corrupt, and these feelings are
deeply entrenched in Sierra Leone’s history. 44 Nonetheless, in considering whether chiefs would
threaten the patterns of provision based on ethnic and party lines, it seems that the chiefs are well
integrated into the political system and would perpetuate inequitable distribution, rather than
rectify it. The chiefdom system does, however, provide insight into the ruling families in Sierra
Leone and their elite role in politics. Thus, it will be useful to examine differences in
distributional outcomes regarding ruling families as well as for youth—often noted to be a group
disenfranchised and excluded from politics in Sierra Leone. The following variations elaborate
on the original hypothesis, addressing both partisan inequities, dependent upon ruling party, and
national inequities, not dependent upon ruling party:
40
Jackson (2006) 101.
Jackson (2005) 53.
42
Fanthorpe (2005) 47.
43
Jackson (2006) 54, 98.
44
Fanthorpe (2005) 46.
41
15
H1.1: If ruling party (local, national) is of same ethnic group, then members of that ethnic group
who are of a ruling family will have higher perceptions of government and economic factors as
well as higher levels of economic income proxies than national non-elite.45
H1.2: If citizens are from a ruling family, then they will have higher perceptions of government
and economic factors as well as higher levels of economic income proxies.
H1.3: If ruling party is of same ethnic group, then members of that ethnic group who are youth
will have lower perceptions of government and economic factors as well as lower levels of
economic income proxies than national non-youth.
H1.4: If citizens are youth, then they will have lower perceptions of government and economic
factors as well as lower levels of economic income proxies.
It will also important to control for alternative theories of distribution. For example,
Alesina, Baquir and Easterly (1999) find that public goods are inversely related to ethnic
diversity. Thus, census information on ethnic diversity within various areas will be useful. In
addition, redistribution of public goods may be a product of democratization in Sierra Leone.
Although not all of Lake and Baum’s (2001) findings were hopeful, they do conclude that
democracy works to an extent, and “remains the most effective instrument available for
controlling the state and producing public services.”46 Foreign aid may contribute to the
democratization of a poor country, and in the provision of public goods.
A poor country, Sierra Leone receives 50% of its revenue from donors. The data
necessary for this paper will be discussed in the final section of this paper, but it is clear that this
hypothesis is threatened by alternative sources of benefits, such as donors. An argument against
such a need to control for aid donations does exist, however. Aid may require certain
representatives of the community to assist in distribution it. Those representatives are often
45
Since the APC won in 2007, this can be tested for both SLPP and APC, though for different periods due to the
data sets available—that is SLPP test will use the last three years in office and the APC test will use the first year in
office.
46
Lake and Baum (2001) 618.
16
either chiefs or respected elders in Sierra Leone. Hanlon (2005), however, argues that the
Paramount Chiefs may be corrupt, using aid money for their own personal gain and to reward
political allies.47 Thus, aid becomes part of the domestic inequitable circulation of goods and
services. In addition, Richards (2003) provides anecdotal evidence for the preferential treatment
of elders in receiving benefits provided by donors.48 Thus, if possible, examining the process by
which aid is distributed, not only the amount of aid provided by donors will be important to
understanding its role in the overall distribution patterns within the country.
In sum, addressing these patterns of provision provides insight into the extent to which
Sierra Leone is an example of a patronage democracy. It also attempts to distinguish between
patronage organized along ethnic lines, or along other cleavages such as class. Both of these
observations will be useful in establishing patterns of provision under the SLPP from 2002-2007
and at the outset of the APC regime in 2007. It will also seek to address when and how reforms,
such as the Local Government Act, or international assistance may help to ameliorate inequity in
distribution.
Vote Shares—A Puzzle
Reiterating the second question stated in the introduction: Do perceptions, or changes in
perceptions, of economy (personal and national) or governance (local and central) among
different ethnic groups predict levels of vote share for an incumbent or opposition party? On the
other hand, is Sierra Leone an example of ethnic voting? The recent loss of the SLPP to the APC
provides an interesting puzzle regarding the issue of ethnic politics in Sierra Leone and the ideas
in the literature regarding ethnic voting.
47
48
Hanlon (2005) 465.
Richards (2003) 35.
17
In 2002, the SLPP managed to win 70%49 of the vote, with 83% turnout.50 Yet in 2007,
with 75.8% turnout, the SLPP not only lost to the APC, each receiving 38 and 44% respectively,
but the SLPP also lost a major political leader, Charles Margai, who formed an opposition party,
the People’s Movement for Democratic Change (PMDC), which won 14% of the vote share in
the first round. No other party received more than 2% of the vote, so the PMDC’s 14% is
remarkable as it only formed a little over a year before the election in 2007. In the runoff, the
APC won with 55% of the popular vote share, and SLPP received 44%. Thus, the question
arises, what conditions led to the downfall of the SLPP? This paper examines two possible
causal factors: changing perceptions of government and the economy (testing economic voting
theories), and perceptions of the SLPP as a Mende ethnic group (testing ethnic expressive voting
theories).
Sierra Leone emerged from a civil war in 1999, though an official UN mandate for a
strengthened peacebuilding mission (UNAMSIL) was not implemented until 2001.51 Ahmad
Tejan Kabbah of the SLPP had been President since the 1996 elections,52 and his role as
President became associated with the establishment of more peaceful conditions within the
country during peacekeeping and peacebuilding.53 However, this success was not to last. After
re-election in 2002, the SLPP government had become known for its corruption and inefficiency,
leading to reported grave dissatisfaction of Sierra Leoneans, and lending a propitious platform
against corruption for the APC candidate, now current President, Ernest Bai Koroma.54
49
African Elections Database.
The Carter Center report suggests irregularities in the vote shares in 2002—including higher then expected turnout
in SLPP strongholds. But these implicit suggestions of fraud were not pursed or proven. See Carter Center
“Observing the 2002 Sierra Leone Elections” (2003) 28.
51
United Nations Background Report on Sierra Leone
52
Ibid.
53
See CNN.com/World (2002); NYTimes/World (2002); Sierra Leone.org (2002).
54
BBC News (2007)
50
18
However, the Afrol news reported the opposition party’s success was surprising: “Given the
fragile position Sierra Leone still experiences—regarding security and economy—voters were
not expected to turn against a government that has been able to guarantee increasing security
after decades of war and destruction.”55 Thus, testing both the perceived experience under the
SLPP as well as any effect that experience may have had on the 2007 elections is an important
step in understanding better how a strong ethnic incumbent party may lose to a challenger.
Further, this paper will test whether cross-pressured Mende voters acted as normal retrospective
voters, influenced by economics and trust in their government and party.
H2: Where there are lower perceptions, or a decline in perceptions about the government and
economic factors, there will be lower votes shares for SLPP and higher vote shares for PMDC or
APC in the first round, and lower proportional gain in votes for SLPP in the runoff.
H2.1: Where there are lower perceptions, or a decline in perceptions about the government and
economic factors, among Mende, there will be lower votes shares for SLPP and higher vote
shares for PMDC or APC in the first round, and lower proportional gain in votes for SLPP and/or
higher abstention rates in the runoff.
Sierra Leone presents an opportunity to test not only challenges to credibility regarding
governance and economic performance, but also to the ethnic psychological benefits and
credibility of ethnic informational cues. As stated above, the PMDC received 14% of the
Presidential vote, but perhaps more surprising is that it also won very large shares within four
previously SLPP-dominated regions. Table 1 shows the percent vote share for the SLPP and the
APC in these four districts.
Table 1: PMDC & SLPP % of total district vote share in former SLPP strongholds56
Bo District
49.7% SLPP
37.1% PMDC
55
56
Bonthe District
31.5% SLPP
61% PMDC
Moyamba District
40.4% SLPP
35.6% PMDC
Pujehon District
50% SLPP
43.5% PMDC
Afrol News (2007)
Data calculated from constituency-level vote shares available at Sierra LeoneDirect Special Report
19
In Bonthe, the PMDC also took all MP seats. Thus, although the PMDC only managed to win
14% of the national vote, it is clear they fared much better within certain previously SLPPdominated districts. In order to understand this shift, it is important to explain the role of
ethnicity in such a shift in support.
The PMDC represents a unique Mende split in recent history within the SLPP that
resulted in a loss for the SLPP.57 Although the formation of the PMDC may have resulted
initially from the SLPP choosing Vice President Solomon Berewa as their Presidential candidate
in 2007 rather than Charles Margai, Margai’s methods of mobilization stemmed from an issue
critical to the Mende people. In 2003, the Special Court of Sierra Leone (SCSL) indicted Sam
Hinga Norman,58 former deputy defense minister in 1996, who also supervised the Civil Defense
Force (CDF). The CDF first began as the Kamajor (or Kamajoi) movement, of Mende descent, in
the south and east in 1996, compensating for the lack of a national force to counter the rebel
Revolutionary United Front (RUF). The areas in which the CDF was entrenched were in the
south: Bo, Pujehon, Bonthe, and Moyamba.59
Although the SCSL is a national Court, the SLPP played a role in its creation in 2002,
as it was the party in rule at the time.60 Thus, the SLPP was directly linked to the indictment of
Norman, according to his Kamajor, Mende followers.61 A quote from the director of PRIDE, an
57
Due to disagreements regarding party leadership in the SLPP, parties, such as the National Unity Party in 1996,
have been formed. However, none has succeeded in reducing SLPP support to the extent that they lose to the
opposition. In addition, John Karefa-Smart formed the UNPP and competed somewhat successfully against the
SLPP in the 1996 elections, winning 40% of the Presidential vote and 17 MP seats. However, Karefa-Smart is not
Mende, but rather Sherbro, an ethnic group prominent in the Bonthe District and the Western Area of Sierra Leone.
Thus, although the UNPP represented a competitive group within the SLPP dominated South, it was not a member
of the Mende who led this dissent. The UNPP’s success was not repeated in 2002 or 2007. Also, the APC was not
allowed to run in the 1996 election.
58
Special Court of Sierra Leone (2003)
Bangura (2000) 558.
60
Special Court of Sierra Leone website
61
The Commonwealth Observer Group, (2007) 5.
59
20
NGO working with ex-combatants in Sierra Leone, reflects the politicization of the Special
Court: “The Special Court has become a big political issue…Many Kamajors thought the court
would only try rebels. They now feel betrayed.”62 When Norman died in 2007, the Kamajors
believed that their leader, Norman, “died an orchestrated death in the hands of the government
they reinstated.”63
The report continued to assert: “Hinga Norman’s death, the conviction of the
heroes of the Kamajors by the Special Court of Sierra Leone and the singular contributions made
by Charles Margai in most parts of the South and the Eastern Regions divided the previously
unshakable bond of the Mendes.”64 One author suggested: “In a sense, the Hinga Norman issue
helped in large measure in defeating the SLPP in the presidential election.”65 Although
previously an SLPP member himself, Margai was not only the founder of the PMDC, but also an
SCSL defense attorney and of Mende descent. 66 Thus, he was perfectly situated to create a
credible image as a true representative of the Kamajor concern.
It is important to note SLPP’s response to these developments. The SLPP Vice
President at the time, Solomon Berewa, responded to the association of the SLPP with the SCSL
in an interview with the Concord Times. It is reported that Berewa tried to remove the blame
from the SLPP for the indictment of Norman, and place it on Norman’s former followers within
the CDF, arguing it was they who pushed for his indictment.67 This attempt by Berewa to create
an alternative explanation was largely unsuccessful, and may have contributed to the loss of
PMDC supporters in the 2007 runoff. Through his association to the SCSL, Berewa may have
weakened the psychological benefit previously produced by voting for the SLPP, since the SCSL
62
IRIN UN Office for the Coordination of Humanitarian Affairs (2007)
Standard Times Press Sierra Leone (2007)
64
Ibid.
65
Dennis (2007)
66
See Notes from Special Court Trial Available (2004) as well as Sierra Leone Court Monitoring Program (2007)
67
Awareness Times Sierra Leone News and Information (2007)
63
21
threatened the SLPP’s claim to be representative of all Mende people. For the same reasons,
Berewa’s ability to be the arbiter of any informational proxy would have also been called into
question, for his credibility as a representative of all Mende people had been challenged.
Perhaps even more surprising than the success of the PMDC against the SLPP was
the eventual joining of former RUF rebels to the PMDC. Several leaders of the CDF and RUF—
former adversaries—signed a press release in January 2007, in which they denounced the use of
violence, gave thanks to the international community for bringing peace to Sierra Leone,
declared support for PMDC and also accused the SLPP government “of being the dividing force
in Sierra Leone.”68 The PMDC began to endorse non-ethnic voting. In a press release before his
death, Norman stated: “We refuse to believe that: Sierra Leoneans…can only associate
themselves with one or two political parties along ethnic and regional lines, [or that] there is no
other choice for our people especially of the South and the East with which to exercise their
political freedom of association.”69 In reference to the south and east, Norman was referring to
the SLPP strongholds.
Further political developments emerged over the course of the year in 2007. The
Republican United Front Party (RUFP) merged with the APC in July of 2007,70 and after the
announcement of a runoff between the SLPP and the APC, the PMDC moved to support the
APC.71 Thus, various ethnic groups, previously from different political parties and even from
68
Kabba, Karamoh, (2007 )
Ibid.
70
Ansumana, Juliet, (2007)
71
Kobah, Karamoh, (2007) and also see Standard Times Press Sierra Leone (2007)
69
22
different warring factions, joined under the same banner against the SLPP.72 Eventually, the
APC took the Presidency and 59 MP seats, the SLPP won 43 seats and the PMDC won 10 seats.
Supporters of expressive voting may argue that Mende voters merely replaced the SLPP
Mende leader for a PMDC Mende leader. Horowitz (1985) after all did predict intra-party
splits. However, the runoff provides a counter-argument. Charles Margai advocated supporting
the APC in the runoff, but Margai—due to the Constitution of the PMDC73—would not be part
of the APC government. Thus, any Mende who voted for the APC in the second round may have
been using informational cues from a Mende leader, Margai, but they would not be receiving an
expressive psychological benefit from voting for a co-ethnic. Perhaps it could be argued they
received an expressive benefit from voting with co-ethnics, but this seems to diverge too far from
the expressive voting thesis. Thus, the runoff in 2007 presents an interesting glimpse into the
question: provided with a second chance, does the PMDC voter continue to reject the SLPP, or
does he/she return to the SLPP party? Table 2 suggests that within the districts where the PMDC
had a significantly strong showing, the APC enjoyed a higher percentage increase in vote share
in the runoff than the SLPP.
Table 2: Percent increase in vote share in Second Round 2007 for Districts in which
PMDC had Strong Showing74
District
SLPP % Increase
APC % Increase
72
It should be noted, however, that not all PMDC members supported the decision to support the APC, see Standard
Times Press Sierra Leone (2007) This question regarding the runoff will be addressed in this paper as well.
73
Manson (2007)
74
Percentages calculated from NEC and Sierra LeoneDirect Data
23
Bo
Bonthe
Moyamba
Pujehon
18%
70%
55%
15%
100%
115%
83%
50%
Table 2 highlights the greater percent increases toward APC vote shares, although these regions
were former SLPP strongholds.75 Thus, it is possible that a significant number of Mende indeed
voted for the APC in the second round. The final hypotheses, then, are:
H3: Where there are higher proportions of Mende, there will be corresponding vote shares for the
SLPP in both rounds.
This hypothesis will determine whether the voters were ethnic expressive, but it would also be
interesting to establish whether perceptions of the government and/or economic factors also covaried with non-expressive voting:
H3.1: Where there are poor perceptions of both economic and/or government factors among
Mende, there will be a higher vote share for the PMDC, and a lower proportional gain in votes
for the SLPP in the second round.
III. Research Design and Data Collection
This paper attempts to tackle two aspects of ethnic politics—how it operates after an
election, through distribution, and how it operates during an election, though voting. Thus, there
may be a problem of endogeneity—election outcomes cause ruling parties, ruling parties cause
75
In order to determine whether these shares were in part due to Mende supporters, this paper will use survey
responses regarding vote choice in both rounds for this Presidential race. However, in addition, it will also be useful
to use demographic census data on the percent of Mende population in these districts, to support or challenge
findings in the surveys. This data is described in the next section.
24
distribution patterns, distribution patterns cause election outcomes and so forth. Nonetheless,
Sierra Leone presents an interesting case where this chain did not proceed in an expected
fashion. Thus, it is worth examining more closely in order to try and determine the reasons why.
The research design is drawn out below. For the first model, the treatment is the ruling
party and the unit of analysis is party support. The measures are derived from survey data on
perceptions of government, perceptions of the economy, and personal economic income proxies.
Thus, if the respondent is of the same ethnic party, or of a ruling family, it is likely the treatment
will produce a relatively increased level of perceptions and indicators. Since the cross-sectional
time series data spans two different ruling parties, these two tests are shown below.
For the second model, the treatment is either perceptions of the government and
economic factors or the weakened ethnic psychological benefit due to the SCSL trial. The unit
of analysis is vote choice or vote share, depending on the measure available. The measures are
vote shares and changes in vote shares in various regions, as well as a 2008 survey question on
the respondents’ vote choice in all elections (2007 Presidential, first and second rounds, 2007
MP, and 2008 local elections).76 These tests are drawn below.
Research Design
Model 1: Does Ruling party ethnicity affect provision and perceptions of government among
ethnic groups of same (or different) ethnicity, and does it matter if it is a local or central
government representative?
For SLPP, in power 2002-200777:
76
Note to WGAPE members: would it also be useful to look at competitive districts and vote shares to first and
second losers (Cox’s SF ratios?).
77
Although there is no pre-test for the SLPP test, it may be argued that due to the oversight of UNAMSIL, which
left Sierra Leone in 2004, replaced by UNIOSIL, the SLPP may not have immediately entered into patronage
25
X (Ruling Party-Central) O (2005 Survey/Service, Perceptions of Gov't by ethnic group) O
(2007 survey, same)
X (Ruling Party-Local) O (2005 Survey/Service, Perceptions of Gov't by ethnic group) O (2007
survey, same)
For APC, in power 2007-present:
O (2005 Survey/Service, Perceptions of Gov't by ethnic group) O (2007 survey, same) X (Ruling
Party-Central) O (2008 survey, same)
O (2005 Survey/Service, Perceptions of Gov't by ethnic group) O (2007 survey, same) X (Ruling
Party-Central) O (2008 survey, same)
Model 2: Do perceptions, or changes in perceptions, of economy (personal and national) or
governance (local and central) among different ethnic groups predict levels of vote share for a
ruling party? Are there other predictors, such as dissatisfaction with a certain policy among a
particular group, good predictors of where the ruling party may lose or gain support?
O (2002 vote share) X1 (Change in 2007-2005 survey: perceptions of economy and/or
government) X2 (Threat to Ethnic Pysch. Benefit from SCSL) O (2007 vote share, first round) O
(2007 vote share, second round) O (2008 survey: reported party support in 2007 elections)
Data Collection
This paper examines perceptions of governance and economic factors in years 2005,
2007 and 2008 as well as vote share patterns in 2002, 2007 and 2008. In order to measure
perceptions of government and economic factors, this paper uses the datasets from the
Institutional Reform and Capacity Building Project (IRCBP) household surveys conducted by
the Government of Sierra Leone and Statistics Sierra Leone. The survey questions address
perceptions of local government, central government, chiefdom systems, public goods provision,
and personal economic factors78—all of which are relevant to this study. These surveys also
provide information on ethnicity, age, ruling families and vote choices in 2007. Further, in 2007
relationships. In addition, leading up to the election in 2007, the SLPP may have increased benefits to their
supporters in order to garner votes.
78
The questions concerning economic factors may facilitate distinguishing between pocketbook and sociotropic
voters.
26
and 2008, additional community and opinion leader questionnaires offer additional data that may
enhance and be integrated with the household data. For example, in 2007, the opinion leader (a
member of a village identified as representative of the village) is asked why a challenger may not
have a chance to win in the upcoming 2008 elections (answers include the standing councilor is:
from a strong political party, has the support of chiefs, from a rich/influential family), and what
are the most important ways in which a community chooses its local councilor (answers include:
political party, promise of development, recommendation by chief and same ethnicity). Annex 1
provides further information on the specific household questions to be used from the surveys.
The household surveys include 6,341 respondents in 2005, 6,338 in 2007, and
approximately 6,340 in 2008.79 A reported 68% of respondents who were interviewed in 2005
were also respondents in 2007, and 20% of the 2007 were respondents of the same household in
2005. The remaining 12% were selected from new households at the same address, or a new
household in same area. The survey was designed to be representative at the local council level,
of which there are only nineteen80, but the large number of respondents allows disaggregation at
the chiefdom level (N=153), and the constituency levels (N=112), which are administrative
boundaries used to elect Parliamentary candidates.81
For election data, this paper refers to the African Elections Database, the LeoneDirect
Special Report, and Psephos: Adam Carr’s Election Archive. From this data, vote shares and
party strongholds may be identified. A detailed description of election data is provided in Annex
2. There is discrepancy in the disaggregation levels of the relevant years: 2002, 2007, and 2008,
79
The 2008 survey is being conducted as this paper is written, so the exact number of respondents is not yet known.
There is one local council per the 13 districts and five councils for major towns within those districts, as well as a
council for Freeport City.
81
The constituency to ward level data is currently being researched by the author, and is currently unavailable.
80
27
and thus further research must be conducted on gathering 2002 disaggregated vote share data at
the constituency or chiefdom level.
In addition, to this electoral data, it would also be useful to collect data on where and
what aid has been provided in Sierra Leone. There is a project, Who What Where, which is
reported to have such data. This paper is also in need of census data on ethnic distributions in
Sierra Leone. Currently, this information is derived from general consensus among academics
and organizations familiar with the country. A 2004 census was conducted with relevant ethnic
demographic information, though only population statistics by gender are available online.
Thus, further research on the census is necessary.
Concluding Remarks
There are several possible findings that may result from this research proposal. Three are
particularly interesting. First, numerous authors have suggested that Sierra Leonean politics is
ultimately a fight over state access and resources, resulting in a patronage democracy (Kandeh,
2003; Fanthorpe, 2005). The first of the original two questions attempts to test whether this is in
fact the case, and if it is, what groups are in fact benefitting. Advertisements of ethnic patronage
may never actualize themselves. In poor countries, such as Sierra Leone, it may also be the case
that there is simply not sufficient material benefit to be spread amongst necessary “re-election
constituencies”. These restrictions may have unexpected or unintended consequences, such as
losing previously strong support bases. The findings in Sierra Leone, then, may be applicable to
other poor countries with the same restrictions.
Second, hypotheses regarding the role of ethnic voting may challenge the differentiation
Lindberg and Morrison (2008) make regarding evaluative and non-evaluative voting behaviors.
28
This argument is weakened when “primordial” shortcuts require evaluation, thus calling into
question the decision to remove ethnic voting from the “programmatic voting” category, where a
vote is “based on either evaluation of past performance of incumbents in terms of public
decisions or on publicly declared intentions of policy if elected.”82 The SLPP’s support of the
SCSL’s indictment of Hinga Norman is such an example of a public decision reflective of past
performance (albeit not economic, but certainly policy-oriented), requiring evaluation of what
may have previously served as a ‘primordial’ shortcut. This example may also lend support for
the need for ethnic informational cues in low-information environments, since Margai may
represent an ethnic cue that does not support a co-ethnic.
Finally, and related to this observation, it will be important to determine—to the extent
we can—whether the weakening of the psychic benefit produces a change in a voter’s
evaluations of other factors, such as the economy. There may be an interactive effect between
economic perceptions and a break in the ethnic benefit that results in a change in expected vote
share. It will be important to assess, then, whether there were lower perceptions of government,
lower perceptions of the economy and/or lower levels of income proxies among Mende who
voted for the PMDC, whether there was a change in these levels between 2003, before the
indictment, and 2007, after the indictment, and whether there were regional differences,
especially in entrenched and non-entrenched CDF districts.
82
Lindberg and Morrison (2008) 101.
29
Annex 1:
Examples of Questions from IRCBP 2005, 2007, 2008 Surveys
2005, 2007, 2008
Position and Income Proxies
 Do you now, or have you held in the past, any of the following positions: paramount
chief, local councilor, traditional society head, etc.
 During the past 12 months, have you done any work for which you received a salary or
regularly paid wage?
 In school, what was the highest class or form you completed?
 Do you own: a radio, type of stove, television, bicycle, etc.
 Do you own or rent?
Public Goods Services and Access:
 Nearest source of drinking water?
 Time to market?
 Quality of roads?
 Location and attendance at school?
 Medical providers?
Interest in and knowledge of government
 Who should run schools, central or local government?
 If you would receive better services by local government, would you pay more taxes?
 If local government official was spending money on him/herself, what could you do
about it, report the problem, sue, not vote for him or her in next election, boycott
meetings, refuse to pay taxes?
 What are the most important sources of information regarding what the government
does?
 Did you register to vote, in what election(s)?
 To whom do you address problems, local councilor, chief, elder in community, NGO,
etc.?
 Are you participatory?
 How do decisions that affect you usually get made, chiefs decide by themselves, chiefs
talk to community, whole community meets, local councilors and chiefs meet, etc.
 Within community, how would you rate your own level of influence?
 Suppose local/central government in this area was given, say 500 million Leones. Do you
believe they would spend the money in an effective way that would benefit the needs of
your council area?
Social Cohesion Questions
 Are relationships in your community generally harmonious?
 Can you trust the people in this community?
 How much do you trust local government officials, central government officials, people
from your community, people from outside community, teachers, doctors?
Data Accuracy
 Were there other people present, including relative, local government official, village
elder or chief, etc.
30
2007, 2008:
Status
 Is your household from a ruling family?
 Are you, or has anyone in your household, been a paramount chief, local councilor, etc.?
Economic Activity
 Do you or anyone in your household do farming or agricultural work?
Chiefdom Governance
 Do you think the Paramount Chief listens to what people in this town say or what they
need?
 If the paramount chief was given 500 million leones to complete a project in this area, do
you believe he/she would spend all the money doing a good job or would he/she cut some
of the money?
Local Council
 Do you think the local council listens to what people in this town say or what they need?
 Do you know when local council elections will next be held?
 Do you think this community would like the same councilor to stand?
Central Government
 Do you personally know anybody who works in government in Freetown?
 Do you think the government in Freetown listens to what people in this town say or what
they need?
 Do you know about what the government is doing?
 Do you know when the next election will be held?
 Do you know who will stand for Presidential election?
 Are you planning on voting?
Opinions
 In our country it is normal to pay a bribe to a government official to encourage them/It’s
wrong to pay a bribe to any government official?
War Experience
 Did anyone in your household receive demobilization, DDR payments?
2008
Household Amenities
 Material of floors, roof, walls?
Central Government
 In the past year, have you talked with your MP?
 While in office, did your previous MP do anything to help your constituency?
 In your opinion, did the previous MP so a good or bad job while he/she was in office?
Voting
 Did you vote in the 2008 Local Elections/2007 National Election?
 Who did you vote for in the 2008 Local Elections?
 Which party did you vote for President (first round)?
 Which party did you vote for President (second round)?
 Who did you vote for in the MP elections?
31

Do you have your voter ID card? [check for punches for first and second round]
32
Annex 2:
Available Election Data
2002:
Presidential national data available from “Psephos, Adam Carr’s Election Archive,” available at:
http://psephos.adam-carr.net/countries/s/sierraleone/, as well as regional data available from
Kandeh (2003).
Parliamentary data by district available at “Psephos.”
2007:
Presidential data by constituency (N=112) available at “Psephos” and “Sierra Leone Elections
Results 2007: LeoneDirect Special Report”
http://www.leonedirect.com/marketplace/sierraleone_parliamentary_election_resultstable.jsp
Parliamentary data by constituency (N=112) available at “Psephos” and “Sierra Leone Elections
Results 2007: LeoneDirect Special Report”
http://www.leonedirect.com/marketplace/sierraleone_parliamentary_election_resultstable.jsp
Presidential Runoff results also from: the National Electoral Commission: “NEC Statement to
the Media,” Thursday 17 September 2007. Available at:
www.dacosl.org/encyclopedia/7_elect/7_1/nec_statement_runoff.pdf
2008:
Local Council Data at local council ward levels (N=394) available from National election
Services, available at: http://www.nec-sierraleone.org/Results.htm
33
Bibliography
African Elections Database, available at http://africanelections.tripod.com/about.html
Afrol News (2007) “Historic Opposition Takeover in Sierra Leone,” September 17, 2007,
Available at: http://www.afrol.com/articles/26736
Alesina, Alberto, Rexa Baquir and William Easterly (1999) “Public Goods and Ethnic
Divisions,” Quarterly Journal of Economics 114:4, 1243-1284.
Ansumana, Juliet, (2007) “As RUFP Merges with APC, Youth Groups Say ‘Ernest Koroma is
Salone’s Saviour,” Awareness Times July 9, 2007, available at
http://news.sl/drwebsite/publish/article_20055947.shtml
Awako Correspondent (2008) “Over APC symbol Brookfields community asks if their vote will
be counted,” Awako Newspaper, available at www.awako.org
Awareness Times Sierra Leone News and Information (2007) “Public Release from Chief Hinga
Norman” February 20, 2007, Available at
http://news.sl/drwebsite/publish/article_20054788.shtml
Bangura, Yusuf (2000) “Policy Failure and Governance in Sierra Leone,” Journal of Modern
African Studies 38:4, p 551-577.
BBC News (2007) “S. Leone Riddled with Corruption,” November 14, 2007, Available at:
http://news.bbc.co.uk/2/hi/africa/7092861.stm
Carter Center (2003) “Observing the 2002 Sierra Leone Elections” available at:
www.cartercenter.org/documents/1349.pdf
Chandra, Kanchan (2004) Why Ethnic Parties Succeed Cambridge: Cambridge University Press.
Christensen, Maya M. and Mats Utas (2008) “Mercenaries of Democracy: The ‘Politricks’ of
remobilized combatants in the 2007 general elections, Sierra Leone,” African Affairs 107:
429 p. 515-539.
CNN.com/World (2002) “Kabbah Declared Winner in S. Leone,” May 19, 2002, Available at:
http://archives.cnn.com/2002/WORLD/africa/05/19/sierra.leone.result/index.html
The Commonwealth Observer Group, (2007) “Sierra Leone Presidential and Parliamentary
Elections 11 August 2007 and Run-off Presidential Election 08 September 2007,”
available at
http://www.thecommonwealth.org/document/39079/152078/170436/sierra_leone_report.
htm
Cox, Gary (1997) Making Votes Count: Strategic Coordination in the World’s Electoral Systems
Cambridge: Cambridge University Press.
Cox, Gary (1999) “Electoral Rules and Electoral Coordination,” Annual Review of Political
Science 2 p. 145-61.
Dennis, Arthur B. (2007) “Sierra Leone 2007 Presidential Election: A Lesson for Liberia,” The
Analyst October 3, 2007, available at:
http://www.analystliberia.com/featurearticle_oct03_07_arthur_dennis.html
Eifert, Benn, Ted Miguel, and Daniel Posner (2007) “Political Sources of Ethnic Identification in
Africa,” Afrobarameter Working Paper No. 89.
Erie, Steven P. (1988) Rainbow’s End: Irish-Americans and the Dilemmas of Urban Machine
Politics, 1840-1985 Los Angeles: University of California Press.
Fanthorpe, Richard (2005) “On the Limits of Liberal Peace: Chiefs and Democratic
Decentralization in Post-War Sierra Leone,” African Affairs 105: 418 p 27-49.
34
Ferree, Karen, (2004) “The Micro-Foundations of Ethnic Voting: Evidence from South Africa,”
Afrobarometer Working Paper No. 40.
Fiorina, Morris (1976) “The Voting Decision: Instrumental and Expressive Aspects,” The
Journal of Politics 38: 2 pp. 390-413.
Hanlon, Joseph (2005) “Is the International Community Helping to Recreate the Preconditions
for War in Sierra Leone,” The Round Table 94:381, 459-472.
Harper, Marcus (1999) “Economic Voting in Post-Communist Eastern Europe,” Center for the
Study of Democracy Paper 99’06.
Horowitz, Donald (1985) Ethnic Groups in Conflict Berkeley: University of California Press.
IRCBP Public Service Reports, 2005, 2007, 2008.
IRIN UN Office for the Coordination of Humanitarian Affairs (2007) “Sierra Leone: Election
Could Turn on Kamajor War Heroes/Criminals” September 7, 2007, available at:
http://www.irinnews.org/Report.aspx?ReportId=74173
Jackson, Paul (2006) “Reshuffling an Old Deck of Cards? The Politics of Local Government
Reform in Sierra Leone,” African Affairs 106:422 p. 95-111.
Kabbah, President Ahmad Tejan (2003) “Speech Delivered by President Kabbah of Sierra Leone
in Kono on Wednesday of August, 2003,” available at www.statehouse-sl.org
Kabba, Karamoh, (2007 ) “Voices from Confinement: Former Warriors Declare Their Support
for the P.M.D.C. in Sierra Leone,” WorldPress.org March 5, 2007, available at:
http://www.worldpress.org/Africa/2702.cfm
Kandeh, Jimmy D. (1992) “Politicization of Ethnic Identities in Sierra Leone,” African Studies
Association 35:1 p. 81-99.
Kandeh, Jimmy D. (2003) “Sierra Leone’s Post-Conflict Elections of 2002,” Journal o Modern
African Studies 41:2 p. 189-216.
Keefer, Phillip and Razvan Vlaicu (2008) “Democracy, Credibility, and Clientelism,” World
Bank Policy Research Working Paper No. 3472.
Kinder, Donald R. and Roderick Kiewiet, (1981) “Sociotropic Politics: The American Case,”
British Journal of Political Science 11, p. 129-141.
Kitschelt, Herbert and Steven I. Wilkinson (2007) Patrons, Clients and Policies: Patterns of
Democratic Accountability and Political Competition Cambridge: Cambridge University
Press.
Kobah, Karamoh, (2007) “Why the PMDC Went APC” PMDC website, undated, available at:
http://www.pmdcsl.org/WhyAPC.html
Lake, David A. and Matthew A. Baum (2001) “The Invisible Hand of Democracy: Political
Control and the Provision of Public Services,” Comparative Political Studies 34:6, 587621.
Lindberg, Staffan I. and Minion K.C. Morrison (2008) “Are African Voters Really Ethnic or
Clientelistic? Survey Evidence from Ghana,” Political Science Quarterly 123:1 p. 95122.
Local Government Act (2004) available at: www.sierra-leone.org/laws.html
Manson, Katrina (2007) “Freetown Opposition Party Wins Majority; Sierra Leone Opposition
Unites for Run-off,” Reuters AlertNet 24 August 2007, Available at: www.alertnet.org
Manson, Katrina (2007) “Analysis-Tribal Voting key to Sierra Leone run-off,” Rueters AlertNet,
26 August, Available at www.alertnet.org
Maravall, Jose Maria and Adam Przeworski (2001)Political Reactions to the Economy: The
Spanish Experience,” in Susan C. Stokes (2001).
35
Notes from Special Court Trial Available (2004) “Special Court for Sierra Leone Monday 29,
November 2004, Trial” at: www.sc-sl.org/Transcripts/CDF-111704.pdf
National Electoral Commission (2007) “NEC Statement to the Media,” Thursday 17 September
2007. Available at: www.dacosl.org/encyclopedia/7_elect/7_1/nec_statement_runoff.pdf
NYTimes/World (2002) “President of Sierra Leone Wins Re-election by a Wide Margin,” May
20, 2002, Available at:
http://query.nytimes.com/gst/fullpage.html?res=9E01EFDF1538F933A15756C0A9649C
8B63&n=Top/Reference/Times%20Topics/People/K/Kabbah,%20Ahmad%20Tejan
Ohman Magnus (2008) “The 2007 Presidential and Parliamentary Elections in Sierra Leone,”
Notes on Recent Elections/Electoral Studies 27 p. 740-773.
Posner, Daniel N. and David J. Simon (2002) “Economic Conditions and Incumbent Support in
Africa’s New Democracies: Evidence from Zambia,” Comparative Political Studies 35:
3, p. 313-336.
Richards, Paul (2003) “Working Paper 21: The Political Economy of Internal Conflict in Sierra
Leone,” Netherlands Institute of Sierra Leone pp. 1-38.
Sartori, Giovanni (1976) Parties and Party Systems: A Framework for Analysis Cambridge:
Cambridge University Press.
Sierra Leone Court Monitoring Program (2007) “Special Court for Sierra Leone Holds
Sentencing hearing on the CDF Trial,”, September 19, 2007, Available at:
http://www.slcmp.org/drwebsite/reports/Special_Court_for_Sierra_Leone_Holds_Senten
cing_hearing_in_the_CDF_Trial.shtml
Sierra Leone Elections 2007: LeoneDirect Special Report, available at
http://www.leonedirect.com/marketplace/sierraleoneelections.jsp
Sierra Leone.org (2002) “Reaction to IGC Africa Report No. 49 of July 2002 on Sierra Leone,”
Available at: http://www.sierra-leone.org/gosl-icgresponse.htm
Special Court of Sierra Leone (2003) “The Special Court for Sierra Leone: Case No. SCSL-03-I,
The Prosecutor Against Sam Hinga Norman, Indictment,” March 3, 2003, Special Court
of Sierra Leone, available at http://www.sc-sl.org/normanindictment.html
Special Court of Sierra Leone website “About the Special Court of Sierra Leone,” at:
http://www.sc-sl.org/about.html
Speech Delivered by President Kabbah of Sierra Leone in Kono on Wednesday of August, 2003,
accessed at www.statehouse-sl.org on 20 September 2008.
Standard Times Press Sierra Leone (2007) “APC, PMDC Marriage: Will the In-Laws Bless this
Courtship,” (August 21, 2007), available at http://standardtimespress.net/cgibin/artman/publish/article_1904.shtml
Stokes, Susan C. (2001) Public Support for Market Reforms in New Democracies Cambridge:
Cambridge University Press.
Stokes, Susan C. (2005) “Perverse Accountability: A Formal Model of Machine Politics with
Evidence from Argentina,” American Political Science Review 99:3, p. 315-325.
United Nations Background Report on Sierra Leone at
http://www.un.org/Depts/dpko/missions/unamsil/background.html
van de Walle (2007) “Meet the new Boss, Same as the Old Boss? The Evolution of Political
Clientelism in Africa,” in Kitschelt, Herbert and Steven I. Wilkinson (2007).
36