coursework
1
To What Extent Does the Wave in which a Country Liberated Affect the Level of Party System Institutionalisation in Africa?
Introduction:
African party systems have been described as both ‘puzzling’ and ‘unique’
(Mozaffar and Scarritt 2005: 339). Political restrictions under authoritarian
regimes limited the freedom of political actors and parties to democratise in
many African Countries. As a result, ethno-political cleavages became an
alternative source of support which led to a unique party system structure in
Africa. However since the liberation of many of these countries, democracy has
been allowed to emerge which has again changed the political structure. The
way in which the countries adapted to these changes, in addition to other factors
can be used explain why there are differing levels of party system
institutionalisation (PSI) amongst African countries. Therefore, it is imperative to
study this unique area of the world that does not always conform to the Western
dominated literature and theory.
This paper will look at the extent to which the wave in which a country liberated
in affects its level of PSI. In addition to this, this essay will also consider a
number of other variables that could be used to explain the differing levels
including; the type of system a country has, the GDP per capita of each of the
countries and also the level of electoral disproportionality as a result of the type
of electoral system in place. This essay will compare two groups of countries
depending on which wave they liberated in. The first group will include Ghana
and Kenya which became liberated in the ‘first wave of democratisation’ when
they liberated from colonial rule and the second group will include South Africa
and Namibia which became liberated many years later than the previous group
when they moved from minority to majority rule. The essay will study these
2
groups by using a Most Similar Systems (MSS) design analysis. This paper will
then finally argue that it is not the wave in which the countries liberated from
that is the main cause of the differing levels of PSI but rather the type of system
that resulted from the type of liberation. However this paper will also highlight
that there are also other important factors that could be used to explain our
outcome variable.
In order to assess whether the time of a country’s liberation in Africa affects its
level of party system institutionalisation, the definition of both a party system
and of PSI must be established. Firstly Mainwaring and Scully define a party
system as 'the set of patterned interactions in the competition among parties'
(Mainwaring and Scully 1995: 4). Furthermore, PSI is a highly disputed and
contested term. Huntington defined PSI as the ‘way in which a party system
acquires value and stability’ (Huntington 1965), whilst Casal Bértoa and Mair
explain that it is the ‘process by which the patterns of inter-party competition
become stable, predictable and routinised over time’ (Casal Bértoa and Mair
2012: 3). However, despite disputes over what makes a party system become
institutionalised, in both definitions we can see that stability plays an integral
role in a countries PSI level. Thus, this essay will predominantly focus on the
stability definition of party system institutionalisation when analysing the
different variables that could be used to explain the differing levels of PSI.
In order to study the party systems in Africa, the context of Africa’s political
history must first be understood. From the early 1800s onwards, Africa was
under the control of European powers, most notably the French, Portuguese and
the British. However this began to change following the end of the Second World
War, which is a key milestone in African history when many European countries
realised they no longer had the economic power or determination to rule Africa
3
(Keller 1995: 157). Consequently, the 1950s marked a period in which many
African countries began to gain their independence. The liberation phase was
predominantly undertaken in two waves. The first wave took place between the
1940s and 1970s when after fighting in the War, many African Soldiers began to
demand more for themselves (Keller 1995: 159). This growing tension in
combination with the economic factors led to the end of colonial rule in many
African countries including; Ghana, Kenya, Malawi and Zambia. However,
Africa’s ‘first wave’ of democratisation quickly transformed into autocracies as
African militaries began to seize power in the mid-1960s. Thus, although most
African countries were at least nominally democratic during this independence
period, over 60 percent of Africa’s sovereign states had experienced military rule
by the 1990s (Keller 1995: 166).
However, the end of the Cold War marked another milestone in Africa’s political
history. During this time western powers began to realise the need to assist
Africa in its pursuit for stable democratisation. In the early 1990s, Africa entered
a ‘second wave’ of democratisation as autocratic regimes were replaced by more
democratic constitutions. Many countries moved from a position of minority rule
into one of majority rule. The countries involved in this so called ‘second wave’
of democratisation included; South Africa, Namibia, Rwanda and Uganda.
Literature Review:
The unique nature of Africa’s party system has led to a keen interest amongst
academics to study party system institutionalisation in the continent. In a study
of 30 African Countries Kuenzi and Lambright measured the differing levels of
PSI. Their findings of the study showed that although the level of PSI was
generally lower than in more western areas and places such as Latin America,
4
they did find that the length of democracy that each country had experienced
was an important factor in explaining the level of party system
institutionalisation (Kuenzi and Lambright 2001: 437). Although Kuenzi and
Lambright believe that the length of time during which a country has
experienced democracy is the key factor in explaining the differing levels of PSI,
many scholars dispute this and look to other variables in explaining differences
in the levels of institutionalisation.
Carbone agrees with the research question in this paper that the ‘wave’ in which
a country liberated is key to explaining its level of party system
institutionalisation (Carbone 2007: 4). The research shows that those countries
that liberated from colonial rule during the 1940s to 1970s on average had lower
levels of PSI usually characterised by fragmented multi-party systems compared
to those countries who liberated from minority rule in the period between 1980
and 2000 (Carbone 2007: 4-8). Furthermore, Carbone argues that this is due to
the reforms of the 1990s following the end of the Cold War when many countries
in the West realised the need to help those African countries that had just
become liberated and thus Carbone argues that the reforms led to the renewal
of many new political parties (Carbone 2007: 4). Those countries that became
liberated during this time are shown to be more institutionalised as they had the
guidance and assistance of the West in helping them to become more
democratised and stable. Therefore the time of liberation could be seen as a
reason in explaining the differing levels of PSI.
However, there are further explanations put forward by academics. Chandra
stresses the significance of ethno-political cleavages and the legacies of
colonialism when explaining PSI in Africa (Chandra 2004: 13). Support for this
argument comes from Ferree who states that sharing ethnicity with a candidate
5
might convince a voter that the candidate’s political position resembles their own
(Ferree 2006). Ethnicity thus may represent a cleavage structure similar to
relevant religious, class, and urban–rural divisions found throughout the rest of
the world (Cheeseman and Ford 2007: 22-26). In an environment where parties
fail to distinguish themselves according to policy, some countries such as Ghana
vote using a valence style of politics which means that ideological attachment is
very weak and can increase electoral volatility. The uncertainty in predicting the
electoral outcome weakens democratic regimes which in turn become more
unstable and thus decreases the level of PSI. Therefore, in countries that use
this valence style of voting, the level of PSI is often lower and thus highlights the
importance of ethno-political cleavages in explaining why some countries are
more institutionalised than others.
However in being critical of the selection of literature, it must be noted that
many of the theories in explaining the differing levels of party system
institutionalisation like Duvergers’s Law for example, is focused on well-
established and long standing democracies often in the West. Africa is unique in
that it does not conform to the usual ideological based politics and has a high
number of ethno-political cleavages. Therefore when applying such literature to
Africa we must be careful as not all theories will apply to these unique party
systems. This would suggest that further research into the uniqueness of Africa’s
party systems needs to be carried out.
Data and methodology:
This research paper will include the analysis of four sub Saharan African
countries which each fit into two groups; those that liberated from colonial rule
in the 1940s to 1970s and those that liberated from minority rule between the
6
1980s and 2000s. The data used in this research paper has been sourced from
the PARLINE database from the Inter Parliamentary Union. As only four countries
are going to be analysed in this research, a small-N analysis will be used in order
to closely examine the four countries in depth. Using a small-N analysis will
increase the internal validity and measurement validity of the findings and give a
more accurate assessment of the four countries rather than a brief
generalisation of all African countries. This is a more suitable research method to
use in this particular research a lot of African countries cannot be generalised
due to the unique nature of their party systems and thus a more focused
analysis of the countries researched in this paper needs to be used.
Dependent Variable:
In order to compare the levels of party system institutionalisation, two key
indicators will be used. This research paper has chosen to compare the electoral
volatility of the four countries and also electoral fragmentation by comparing the
effective number of legislative parties in each system. These dependent
variables have been chosen because they give a clear indication of how
institutionalised a particular country is when looking at the results of these
indicators. Pederson’s index of electoral volatility gives us an accurate
representation of how the political system has changed from one election to
another which is useful in analysing changes to a particular party system
(Pederson 1979). Furthermore Krupavicius argues that electoral volatility is ‘a
reliable indicator of relative strength and social rooting of political actors in
consecutive elections’ (Krupavicius 1999: 8). This is therefore an important
indicator to use to determine the level of institutionalisation in Africa as many of
the party systems are highly unpredictable and irregular. In addition to this, the
second variable chosen is electoral fragmentation using Laakso and Taagepera’s
7
Table 1: Average Electoral Volatility Score (2000-‐2016)
measure of effective number of legislative parties (Laakso and Taagepera 1979).
As many African countries have a large number of legislative parties, it is
important to not just count the amount of parties in a system but to weight the
count of them by their relative strength. Therefore this an appropriate measure
to use as this research study has chosen to analyse countries with often a high
number of political parties.
Independent Variables:
The main independent variable in this research paper is the wave in which a
country liberated whether that be from colonial rule or from minority rule.
However other variables will also be analysed in order to try and explain the
differing levels of PSI. These variables include the type of system present in
each country, the GDP per capita each country has and also the electoral
disproportionality level of the four countries. To test the main independent
variable, the countries will be split into two groups and then compared on their
levels of electoral volatility and electoral fragmentation. The main hypothesis
using the key outcome predictor is:
H1: Countries that liberated in the second wave of liberation will have higher
levels of PSI.
Analysis:
Table 1 shows the varying levels of
electoral volatility between the two
groups with those countries in red
becoming independent in the first
wave of liberation and the
Country
Electoral Volatility (EV)
Ghana 25.8
Kenya 32.6
South Africa 16.5
Namibia 17.9
8
Table 2: Average Effective Number of Legislative Parties Score (2000-‐2016)
Country Effective Number Legislative Parties (ENPP)
Ghana 3.06
Kenya 7.54
South Africa 1.98
Namibia 1.55
countries in blue becoming independent in the second wave of liberation. The
results clearly show that those that became liberated from colonial rule between
1940 and 1970 have a much higher electoral volatility score than those who
liberated from minority rule much later. Kenya which became liberated in 1963
is shown to have a volatility score of over 16 points higher than that of South
Africa which only became fully independent in 1994. The higher the electoral
volatility, the lower the level of PSI which would mean that Ghana and Kenya
have a lower level of party system institutionalisation.
Furthermore we again see
patterns of similarity when we
compare the effective number
of electoral parties between
the two groups shown in
Table 2. The fragmentation
score of the two countries
that became independent first are shown to be significantly higher than South
Africa and Namibia which became liberated many years later than Ghana and
Kenya. The lower the number of effective legislative parties, the more stable the
country should be and thus the higher the level of PSI. Therefore, South Africa
and Namibia would have a higher level of party system institutionalisation than
that of Ghana and Kenya who have a higher number of effective legislative
parties.
Country Effective Number Legislative Parties (ENPP)
Ghana 3.06
Kenya 7.54
South Africa 1.98
Namibia 1.55
9
In order to establish what explains the differing levels of party system
institutionalisation between the two groups of countries, a Most Similar Systems
Design (MSSD) of analysis will be used. All of the countries are similar in that
they are all located within the African Continent, were all former colonies of
either the British or the Germans and all now considered to be democracies. The
MSSD study will be used to explain why South Africa and Namibia are shown to
be more institutionalised (when using electoral volatility and electoral
fragmentation as indicators) than Ghana and Kenya.
In order to establish why South Africa and Namibia are considered to be more
institutionalised, we must first eliminate those factors that they have in common
with Ghana and Kenya and thus wouldn’t explain the difference in PSI levels.
Therefore using the information in Table 3 we can see that the level of political
culture, the time of democratic transition and the length of democracy are all
factors that either relatively similar or identical across all four countries. Thus,
we can rule out these factors in explaining the differing levels of
institutionalisation. In addition to this, there is a further factor that can be
eliminated. The type of regime present in South Africa and Namibia is not
0
5
10
15
20
25
30
35
Electoral VolaClity (EV) Electoral FramentaCon (ENPP)
Ghana Kenya South Africa Namibia
Figure 1: Bar Chart showing the differing levels of average electoral volatility and average electoral fragmentation in the four countries (2000-‐ 2016).
What are the Levels of Electoral Volatility and Electoral Fragmentation in Ghana, Kenya, South Africa and Namibia?
10
Table 3: Sources of PSI in Africa: MSSD (most recent data)
consistent as South Africa is shown to have a parliamentary regime whist
Namibia has a semi-presidential regime. Therefore regime type also cannot
explain why these two countries are more institutionalised compared to Ghana
and Kenya.
However, although the factors showing consistency can be ruled out, there are
many patterns that can be drawn from Table 3 which could help to explain the
differences in levels of PSI. The first factor that can be used to explain the
differing levels of party system institutionalisation is the variable addressed in
the research question- the time of liberation and most notably which ‘wave of
liberation’ it corresponds to. From Table 3, the information shows that South
Factors Ghana Kenya South Africa Namibia Liberation gained (Year)
From colonial rule in 1957 (1st wave)
From colonial rule in 1963 (1st wave)
From minority/majority rule in 1994 (2nd wave)
From minority/ majority rule in 1990 (2nd wave)
Type of regime presidential presidential parliamentary semi-‐ presidential
Political culture 6.25 5.63 5 5.63 GDP per capita ($)-‐ Taken in 2017
1697 1133 7593 6000
Length of colonial rule (years)
90 43 188 106
Time of transition 3rd wave 3rd wave 3rd wave 3rd wave Length of democracy (years)
24 26 23 27
Type of system multiparty but mainly 2 party system
multiparty dominant dominant
Electoral disproportionality (Lsq)
5.76 8.26 0.37 0.9
Type of Electoral System (in legislative elections)
First Past the Post First Past the Post Party List PR Party List PR-‐ closed list
11
Africa and Namibia, the two countries that have higher levels of PSI, both
liberated from minority rule in the second wave of liberation. This liberation took
place a lot later than Ghana and Kenya. The aforementioned argument put
forward by Carbone states that it was the influence of the west in helping
democracy to flourish that led those countries who liberated later to become
more stable (Carbone 2007). Therefore we can accept our H1 hypothesis.
However, perhaps it wasn’t the time of liberation that led to the varying levels in
PSI but the influence that the time in which the countries liberated that shaped
the type of parties that formed. There are two main types of parties in Africa;
Independence parties and Liberation Parties. Independence parties formed in the
first wave of liberation are usually characterised as being a lot weaker than the
Liberation parties and are usually fragmented which often results in multi-party
systems which can be seen in both Ghana and Kenya. However Liberation
parties have been shown to have a more lasting presence and are often more
structured. This has often led to one party becoming dominant in these countries
that liberated in the second wave. Therefore perhaps it is not the time of
liberation that is the key factor but the influence that the time of liberation had
on the type of party system which then followed.
The second factor that could also be used to explain the differing levels of PSI is
the type of system a country has. Our hypothesis when analysing this variable
is:
H2: Countries with a dominant party system are more likely to have higher
levels of PSI.
From Table 3, the information shows that both South Africa and Namibia have a
one party dominant system. This contrasts from the multiparty system existent
12
in Kenya and also that of Ghana which although has traditionally also had a
multiparty system, more recently has become dominated by two main parties,
the New Patriotic Party and the National Democratic Congress party. There is a
lot of conflicting arguments over the impact of having a one party dominant
system on a country’s level of party system institutionalisation. Giliomee and
Simkins argue that lack of opposition can be extremely detrimental to
democracy and believe that having a strong opposition is essential in preventing
a move towards authoritarian tendencies and abuse of power by the incumbency
(Giliomee & Simkins, 1999: 337).
However Arian and Barnes suggest an alternative argument and in fact hail a
one party dominant system as a “stabilising mechanism” (Arian and Barnes
1974: 600). Although this argument put forward by Arian and Barnes is based
upon their analysis of dominant parties in Israel and Italy, the findings of their
examination into the nature of these systems can be applied to other systems
especially the study of the dominant party system in Africa. Therefore, the
existence of a dominant one party system in South Africa and Namibia could
explain why the level of electoral volatility and fragmentation is lower in these
two countries and thus have a higher level of PSI. The lack of competition and
predictability of having a one party dominant system lowers the countries
volatility level and thus the prediction of what will happen from one election to
another is more stable and prominent. Thus, the type of party system could be
the key to explaining the differing levels of PSI between the two groups of
countries and we can therefore accept our H2 hypothesis.
However, this factor is not the only variable that can be used to explain why
South Africa and Namibia have different PSI scores to that of Ghana and Kenya.
13
A country’s GDP per capita is the next variable that will be analysed. Our
hypothesis for this variable is:
H3: Countries with a higher GDP per capita will have higher levels of PSI.
From the data presented in Table 3 we can also see that South Africa and
Namibia both have a much higher GDP per capita to that of Ghana and Kenya.
There have been a vast amount of research into the impact of a country’s GDP
per capita on its level of PSI. Much of this research however focuses on countries
in the west and fails to go into depth the impact of GDP in Africa. Despite this,
much of the research on the west can be applied to the countries in this study.
In one logistic regression analysis carried out by Casal Bértoa, the impact of a
countries GDP per capita was shown to be significant at .00, and the variable
was found to have a strong substantive impact on PSI (Casal Bértoa 2011).
Furthermore the research paper goes on to argue that the more advanced an
industrial democracy is, the more the stable party system will be when
compared to the less-developed democracies (Casal Bértoa 2011).
Beatty Riedel expands on this argument put forward by Casal Bértoa by
explaining the reason behind this contention. Beatty Riedel argues that when the
economy is good and the country is thriving, the voters are more likely to be
satisfied with the current political situation and thus more likely to vote a party
back into power. This in turn reduces the electoral volatility and makes a country
become more stable (Beatty Riedl 2014: 126). Therefore this could explain why
South Africa and Namibia have lower levels of electoral volatility compared to
the levels present in Ghana and Kenya. And thus these findings support our H3
hypothesis and a country’s GDP per capita could explain the differing levels of
PSI.
14
Finally, one other variable that can be used to explain the differing levels of PSI
is the disproportionality between the vote distribution and the seat distribution
which is a resultant from the electoral system existent in the country. Our final
hypothesis in this research paper is:
H4: Countries with lower electoral disproportionality as a result of a proportional
electoral system will have lower levels of PSI.
In order to measure vote-seat proportionality, the Gallagher Least-Squares
(LSq) Index is used. From the data in Table 3 we can see the huge variation in
the electoral disproportionality of Ghana and Kenya which is extremely high in
comparison to the electoral disproportionality of South Africa and Namibia.
However when we compare these findings with the type of electoral system
present in these countries we see a correlation between Ghana and Kenya which
both have high electoral disproportionality and also use a majoritarian First Past
the Post (FPTP) electoral system. Contrary to this, South Africa and Namibia
which both have low electoral disproportionality both use a proportional
representation electoral system when electing their legislature. We would expect
that those countries that possess a proportional electoral system to have a lower
electoral disproportionality score which is supported by our findings.
Although this trend fits with our countries, there is some divergence when
looking at the impact of disproportionality upon the effective number of electoral
parties and thus PSI. When applying Duverger’s thesis that a majoritarian
electoral system tends to favour a two party system to the African countries, we
would expect that the effective number of electoral parties would be lower which
would mean higher levels of PSI (Duverger 1954). However this is shown to not
be the case. The two countries that use a majoritarian electoral system are
15
shown to have a higher number of effective electoral parties than Namibia and
South Africa that use a proportional electoral system and have a lower ENEP.
Therefore we fail to accept the H4 hypothesis. However, in being critical of the
selection of literature, it must be noted that this thesis is based on established
democracies.
Therefore Mozaffar puts forward an alternative explanation by arguing that many
African countries are emerging democracies and are still unsure of the normal
voting behaviours. Voters in relatively new democracies may have difficulty in
understanding the workings of their countries’ electoral systems, and therefore
may struggle to understand how to behave strategically (Mozaffar and Scarritt
2005: 343). There is further support from Reynolds who argues that in highly
polarised ethno-regional areas with lots of cleavages like that existent in Africa,
a proportional electoral system tend to increase political stability and reduce
conflict and violence (Reynolds 2009: 78-79). Reynolds argues that by including
marginalised ethnic groups into legislative elections, the inclusion has a more
positive effect on social tolerance and increases the likelihood that voters will
support democracy. Thus, perhaps because Africa has relatively new
democracies and is highly polarised, a proportional electoral system like that
existent in Namibia and South Africa is more suitable for fostering a more stable
democracy which is shown in the lower electoral volatility and effective number
of legislative party levels for both of these countries.
Conclusion:
Finally to conclude, it is clear that the study of party system institutionalisation
in Africa is vital to understanding how new democracies are developing. From
the research presented in this essay, the main explanations for the varied levels
16
Word Count: 4,354
of PSI are shown to be the wave in which the countries liberated, the type of
party system, the level of GDP per capita and finally the level of electoral
disproportionality present in the four countries. Although all of these factors
could be argued to be a significant cause of the level of party system
institutionalisation a country has, it is clear that some factors are more
prominent than others.
It is clear that the wave of liberation did have an impact on these countries,
however the way in which countries liberated has been shown to be more
significant on the type of system and parties that formed from this event rather
than the event itself. It is also clear that the amount of GDP per capita a country
has assists in increasing confidence amongst voters of the political system in
place which in turn often leads to re-election and thus a more stable political
system. However although GDP does clearly play a big part, the political system
already needs to be in place and working in order for re-election, thus GDP can
be seen as increasing and prolonging PSI but not necessarily as the main
causation.
However by using a MSS design analysis we can only see what factors influence
our outcome and not the extent to which they influence. In order to analyse
which variable is most significant in explaining the differing levels of PSI, a
regression analysis. Thus, in order to study the party systems in Africa in more
depth further analysis needs to be undertaken. Furthermore, although this paper
has highlighted many of the possible explanations for the differing levels of PSI
amongst African countries, not all potential explanations are taken into
consideration and thus, again, further research into this continent does need to
be carried out.
17
Bibliography:
Arian, Alan, and Barnes, Samuel. H. (1974) ‘The Dominant Party System: A Neglected Model of Democratic Stability’. The Journal of Politics 36.3. [Online}. Available at: https://www.jstor.org/stable/421684?seq=1#page_scan_tab_contents (Accessed 1 May 2017).
Beatty Riedl, Rachel. (2016), Authoritarian Origins of Democratic Party Systems in Africa, Cambridge University Press.
Carbone, Giovanni. M. (2007) ‘Political Parties and Party Systems in Africa: Themes and Research Perspectives’. World Political Science 3.3. [Online]. Available at: http://adpm.pbworks.com/f/africa%20party%20systems%20carbone.pdf (Accessed 2 May 2017).
Casal Bértoa, Fernando. (2011). ‘It’s mostly about money! Party system institutionalization and its sources: reducing causal complexity in post- communist Europe’, Working Paper Series on the Legal Regulation of Political Parties. [Online]. Available at: http://www.partylaw.leidenuniv.nl/uploads/wp1511.pdf (Accessed 8 May 2017).
Casal Bértoa, Fernando and Mair, Peter. 2012. “Party System Institutionalization across Time in Post-Communist Europe.” In Party Government in the New Europe, ed. Ferdinand Müller-Rommel and Hans Keman. London: Routledge.
Chandra, Kanchan. (2004) ‘Ethnic Parties and Democratic Stability’. Perspectives on Politics 3.02. [Online]. Available at: https://www.cambridge.org/core/journals/perspectives-on- politics/article/ethnic-parties-and-democratic- stability/1D5DDC5687496C4157C3B2CDAEC88B2A (Accessed 29 April 2017).
Cheeseman, Nick, and Ford, Robert. (2007) ‘Ethnicity as a political cleavage’. The micro- foundations of mass politics in Africa 13.05. [Online]. Available at: https://www.researchgate.net/profile/Robert_Ford11/publication/241306316_Et hnicity_as_Political_Cleavage/links/0c96053bfcab8228cc000000.pdf (Accessed 1 May 2017).
Duverger, Maurice. (1954). Political Parties. New York: Wiley
Ferree, Karen. E. (2006) ‘Institutional Duration and Growth in Africa’. Studies in Comparative International Development 40.4. [Online]. Available at: https://link.springer.com/article/10.1007/BF02686302 (Accessed 6 May 2017).
Giliomee, Hermann and Simkins, Charles. (2005) The Awkward Embrace: One-party Domination and Democracy, London: Taylor & Francis e-Library.
"IPU PARLINE Database on National Parliaments". Ipu.org. N.p., 2017. [Online]. Available at: http://www.ipu.org/parline-e/parlinesearch.asp (Accessed 2 May 2017).
Keller, Edmond. J. (1995) ‘Decolonization, independence, and the failure of politics’. Africa, 3. [Online]. Available at: http://www.sscnet.ucla.edu/polisci/faculty/keller/papers/SelectedPub/decoloniza tion.PDF (Accessed 28 April 2017).
18
Kuenzi, Michelle, and Lambright, Gina. (2001) ‘Party System Institutionalization in 30 African Countries’. Party Politics 7.4. [Online]. Available at: http://journals.sagepub.com/doi/abs/10.1177/1354068801007004003 (Accessed 1 May 2017).
Krupavicius, Algis. (1999) "The 1997–1998 Presidential Elections in Lithuania". Electoral Studies 18.1. [Online]. Available at: http://onlinelibrary.wiley.com/doi/10.1111/1467-9248.00151/abstract (Accessed 9 May 2017).
Laakso, M. and Taagepera, R., (1979) “Effective” number of parties: a measure with application to West Europe. Comparative political studies, 12. [Online]. Available at: http://journals.sagepub.com/doi/abs/10.1177/001041407901200101?journalCo de=cpsa (Accessed 1 May 2017).
Mainwaring, Scott and Scully, Timothy. R. (1995) ‘Building Democratic Institutions: Party Systems in Latin America’. Foreign Affairs 74.4. [Online]. Available at: https://www.cambridge.org/core/journals/canadian-journal-of-political-science- revue-canadienne-de-science-politique/article/building-democratic-institutions- party-systems (Accessed 28 April 2017).
Mozaffar, Shaheen and Scarritt, James. R. (2005) ‘The Puzzle of African Party Systems’. Party Politics 11.4. [Online]. Available at: http://journals.sagepub.com/doi/abs/10.1177/1354068805053210 (Accessed 28 April 2017).
Pedersen, M.N., (1979) The dynamics of European party systems: Changing patterns of electoral volatility. European Journal of Political Research, 7. [Online]. Available at: http://onlinelibrary.wiley.com/doi/10.1111/j.1475-6765.1979.tb01267.x/full (Accessed 1 May 2017).
Reynolds, Andrew. (2009) ‘Elections, electoral systems, and conflict in Africa’. The Brown Journal of World Affairs, 16. [Online] Available at: http://www.jstor.org/stable/24590741?seq=1#page_scan_tab_contents (Accessed 8 May 2017).