The Determinants of Party System Change in Sub-Saharan Africa

This article examines what socio-economic factors are conducive to changes in the patterns of interparty competition in Sub-Saharan Africa (SSA). The literature has in recent years paid considerable attention to measuring party system change, to identifying the consequences of party system instability for the proper functioning of democratic regimes, and to understanding what factors are responsible for the instability of party systems. In contrast to previous studies that view political change in general and party system change more specifically as the result of social transformation, development, modernization and change in the cleavage structure, this paper shows instead that poverty is the primary driver of party system change in the SSA region. In countries with high levels of poverty, political elites do enjoy little to no performance-based legitimacy. The lack of performance-based legitimacy is the reason why voters in such countries are willing to alter their voting habits and parties are unable to preserve their electoral fortunes over time—which is precisely why party systems do change. The literature showed that stable party systems are good for democracy. This paper shows that to enhance the stability of party system in SSA, poverty has to be reduced and possibly eradicated.


INTRODUCTION
In 1976 Giovanni Sartori published a book entitled Parties and Party Systems.The book presented a new framework for the analysis of party systems.In this book the Italian political scientist advanced four main claims, namely that party systems are structured or stable patterns of inter-party competition, they could be categorized on the basis of their stability, fragmentation and ideological polarization, the fluidity (instability) of party systems was in many ways the result of the fact that political parties were not properly institutionalized, and that African party systems were a puzzle for party system scholars.The puzzling nature of African party systems was primarily due to the fact that, lacking properly institutionalized parties, they displayed high levels of fluidity.
In recent years party system scholars have considerably advanced the understanding of African party systems.Some party system scholars attempted to assess the level of institutionalization of African parties (Basedau & Stroh, 2008), others (Mozaffar & Scarrit, 2005) showed that African party systems are characterized by low levels of fragmentation coupled with high levels of volatility (electoral change), while Bogaards (2008) noted that the level of volatility varies across levels of democracy and party system types.A fourth line of research (Mozaffar, Scarritt, & Galaich, 2003;Ferree, 2010) reported that in Africa ethnopolitical divisions or cleavages (Lipset & Rokkan, 1967) are significant determinants of volatility.
While volatility, as measured by the Index of Volatility devised by Pedersen (1979), provides a useful indication of whether and how much the electoral fortunes of political parties change in consecutive elections, it does not provide an indication of whether (and how) the party system has changed (Bartolini & Mair, 2007).It fails to so because it does not provide any indication as to whether the cleavage structure or the pattern of interparty competition have changed.
Nwokora and Pelizzo devised a new measure, the index of fluidity (Nwokora & Pelizzo, 2015;Nwokora & Pelizzo, 2018), to capture the frequency and the magnitude of the changes in the functioning of party systems.This index has been computed for all the countries in Sub-Saharan Africa (Pelizzo & Nwokora, 2016) from the year in which they achieved independence to 2013.
In contrast to other studies (Nwokora & Pelizzo, 2016;Pelizzo & Nwokora, 2018), that investigated the impact of fluidity on the qualities of democracy, the purpose of the present study is to examine some of the possible causes of party system instability.In doing so, we will explore the relationship between the instability of a party system -as measured by the index of fluidity-and the ethnic fragmentation of the population because the literature has consistently identified a strong relationship between volatility and ethnic fragmentation.But we will also investigate whether variation in the level of party system instability is affected more significantly by rapid economic growth and development, as Olson (1963) and Huntington (1968) once hypothesized, or whether it is affected more significantly by poverty.

RESEARCH METHODS
Party systems, as Sartori (1976) noted, are systems that result from the pattern of inter-party competition.Depending on how parties relate to one another, depending on how they affect the direction of competition, depending on whether and to what extent such effect is stable over time, the patterns of inter-party competition can properly be understood as party systems.
What makes a party system a system is not simply the existence of a specific pattern of party competition but also the stability of such pattern.
The absence of this kind of stability, the changeability, the propensity that party systems may have to change is what Sartori (1976) defined as fluidity.
Knowing that the fluidity of a party system is its changeability does not provide any indication as to what actually amounts to proper party system change.In this regard, Pelizzo and Nwokora (2016), summarizing the party system change literature, noted that party system change has been associated with party change, electoral change, change in the cleavage structure and functionalist change, which, for Pelizzo and Nwokora, represents the most appropriate way of conceptualizing party system change.
Having identified party system change with change in the functioning or mechanics of a party system, Nwokora andPelizzo (2015, 2018), proposed a new methodology for measuring the fluidity of party systems.
The fluidity of a party system they argued, can be measured by the Index of Fluidity, which is computed by estimating the frequency, the scope and the variety of change.In addition to proposing such methodology, Pelizzo & Nwokora (2016) and Nwokora & Pelizzo (2018) estimated the index of fluidity for all the countries from Sub-Saharan Africa.In the empirical section of this paper we will rely on the data made available in these two studies to perform our statistical analyses and develop a better understanding of what makes party systems change in the African context.
Party system scholars have proposed several ways to categorize party systems (Pelizzo & Nwokora, 2016).The best known framework for the analysis and categorization of party systems was proposd by Sartori, who noted that there are seven types of structured party systems or party systems proper: one party systems, hegemonic party systems, pre-dominant party systems (Nwokora & Pelizzo, 2014), two-party party systems, moderate pluralist party systems, polarized pluralist party systems, and atomized party systems.One party, hegemonic and predominant party systems have only one relevant party; two party systems have two relevant parties, moderate pluralist party systems have more than two relevant parties but the pattern of competition and the pattern of alternation in government is identical to what we detect in two party systems, polarized pluralist have at least between five and six relevant parties and considerable amount of ideological polarization, while some party systems are considered atomized because they have so many parties than any additional increase in the number of parties won't affect their functioning.
A change between any of these types (from one to atomized, from predominant to polarized pluralist, and so on) is a party system change.The frequency with which it occurs is what is used to assess the 'frequency' dimension of the index of fluidity.Some changes are bigger than others, which is why the scope of change that a party system experiences evolving from a one-party to a hegemonic party system is less than it would be by evolving from a one-party to a polarized pluralist party system.And since, some party systems, in their historical development, adopt a larger or smaller number of party system types, the variety of party system change is the third dimension of party system change that is used by Nwokora & Pelizzo (2015, 2018) to assess the fluidity of party systems.
Knowing what is a party system change, how we can detect it and how we can measure the fluidity of party system does not provide much information about why party system change occurs-which is precisely what this paper aims to uncover.

RESULTS AND DISCUSSION
In this section we want to present the results of some preliminary analyses on the relationship between the fluidity of party systems on the one hand and poverty, development, the urban rural cleavage and ethnic fragmentation on the other hand.Before we do so we need to say a few words about our data and datasources.
The fluidity of party systems is measured by the index of fluidity created by Nwokora and Pelizzo (2015).
Poverty is measured on the basis of the percentage of the population living in slums in 1990 and 2014.The data are taken from the World Bank's Development Indicators Database.
Development is measured on the basis of the change in the percentage of the population living in slums from 1990 to 2014.The idea is that as people become richer, move out of the slums, society becomes more pluralistic, the demand for political pluralism grows and the number of parties may grow accordingly.
The transformation of the percentage of rural population provides an indication not only of the socio-economic transformation of the country -if the rural populations moves to the cities, industrialization may be occurring -but also of the fact that as the urban population grows, a rural-urban cleavage may pit urban and rural interest against each other, creating the condition for greater political pluralism and party system change.
Finally, since in the African context, tribalism is often blamed as the root cause of many of the problems that the continent is experiencing, it is interesting to see whether and to what extent the instability of party systems is influenced by the ethnic fragmentation.
By correlating the data on party system fluidity with the data on poverty, development, rural population and ethnic fragmentation, we find that higher levels of poverty are correlated with, and possibly are conducive to, higher party system instability.In fact the correlation between fluidity of party systems and the percentage of people living in slums in 1990 and 2014 yields strong, positive and statistically significant coefficients.This means that regardless of when we measure poverty, higher rates of poverty are associated and are possibly responsible for higher levels of party system fluidity.
Contrary to what Olson (1963) and Huntington (1968) had hypothesized, development is not a destabilizing force.Changes in the percentage of people living in urban slums has no impact whatsoever on the stability/instability of African party systems.The correlation yields a negative but statistically insignificant coefficient ( r = -.049,sig.817).
Ethnic fractionalization, which may influence Africa in other ways, also has no impact on the stability/instability of party systems.The correlation yields a weak, positive, but statistically insignificant coefficient (r = .083,sig.581).
The percentage of people living in urban areas is unrelated to the instability of the party systems.The correlation between fluidity and the percentage of rural population in 1990 and 2014 yields very weak and very insignificant coefficients.The analysis also reveals that the fluidity of party system is also not affected by changes in the percentage of people living in rural areas and that therefore urbanization and the associated emergence of a possible urban-rural cleavage have not had any meaningful impact on party systems and on the changes thereof.See table 1.

CONCLUSION
Political scientists had long been aware of the political consequences of party system attributes.Government instability, political instability, democratic breakdowns, quality of legislation, size of the deficits, had all been seen at one point or another as the consequence of the fragmentation and ideological polarization of party systems.This has been the main claim to fame of the system attributes approach from Lowell (1896) to Tsebelis (2002).Nwokora & Pelizzo (2015) have shown, however, that the stability/ instability of party system is as important as the system attributes and possibly even more important.The quality of South East Asian democracies was more deeply shaken by the instability of party systems than by their fragmentation.A similar conclusion was reached by Pelizzo & Nwokora (2018) in their analysis of the relationship between fluidity and quality of democracy in East Africa.
This line of research was able to show that the fluidity of party system matters as there are clear and clearly visible consequences of party system change and instability.For instance, Nwokora &Pelizzo (2015) and Pelizzo & Nwokora (2018) documented that there is an inverse relationship between the instability of party systems and the qualities of de-mocracy.In other words, as the instability of party systems increases, the quality of democracy declines.
Previous studies showed how to measure party system change and why party system change matters.The present study shows what factors are responsible for party system change in Sub-Saharan Africa.Specifically, we were able to show that while ethnic fragmentation, urbanization, and development have no detectable impact on the stability of party systems, poverty can greatly compromise and undermine the stability of party systems.
It is easy to understand why.By failing to promote development and to take people out of poverty, political elites lose the performance-based legitimacy (Huntington, 1993) that they would be otherwise able to enjoy, voters lose faith in the political system and, as a result, instead of simply desiring a government change, they desire a system change, which includes a change in the format and in the functioning of the party system.
These findings have a simple, but in our opinion rather important, implication: to enjoy the dividends of party system stability, party systems need to be stabilized, and the reduction of poverty is the most important step to reach this outcome.