This week’s reading discussed the emergence of political party systems in new democracies. The reading focused on three areas where new democratic nations were developing and trying to establish a system of political parties; the Former Soviet States, Eastern Europe and Africa. All of these areas have had challenging histories which lead from them leaving an authoritarian system to a new democratic emergence, but how will these nations develop lasting and strong sytems of political parties? Will they follow the traditional view that strong party systems will develop over time or will these nations be able develop a meaningful party system quickly in their evolution as a democratic nation? Then will these party systems fall along the same social cleavage lines that we see in most other western democratic nations, and if so what other factors including institutional factors will affect the outcomes of the party systems and the positions the parties in the system take.
The Miller piece contradicts the traditional view of party system development which states that emerging democracies will take a long time to fully develop strong party systems. He shows that in three former Soviet Union States, (Russia, Ukraine, and Lithuania) political party systems are developing and doing so in a manner far quicker than expected by many academics. Miller examined all three of these nations for six years, from 1992 to 1998. He found, in all three countries, both the populace and the elites party identification has increased significantly over the period studied with the percentages of elites and populace being relatively equal, thus showing that the average person feels an equal amount of representation between the populace and the elites. This fact is a good indicator of a strong functioning political party system. This study further discredits the socialization theory to political party development showing that the highest level of growth in partisanship ‘fell among the young and middle class who should have been the least affected by the socialization of the previous communist regime.’ This finding helps to show the strength of the party system by showing party identification and loyalty is centered around real and salient issues as opposed to just socialized ideals.
The work of Miller sets up the developments in the understanding of how party systems develop furthered by Zielinski. Miller showed that party systems can develop quite quickly and it is possible for these systems to develop around salient issues and for the parties in the system to develop in a manner where it’s partisans have a strong identification to the party. Zielinski furthers this argument by showing that parties can choose which cleavages to cement around and make the entrenched divisions between the parties. The Zielinski piece helps to merge the competing ideas of the social cleave and institutional theories that have dominated our study to date. He shows how the parties utilize the institutional factors to consolidate the political parties and they also selectively choose which social cleavages to center around in the consolidation of political parties. This discovery helps to merge the two competing theories by stating that every social cleavage does not have to be politicized. Therefore the parties choose which cleavages to focus on. This finding is seconded by the work of Mozaffar et al who shows that party systems are affected by both institutional factors and social cleavages simultaneously, parties do this strategically. Mozaffar seconds the work of Zielinski, finding it in Africa as opposed to Eastern Europe.
The Zeilinski work also noted a fascinating point in that the Eastern European Nations may not develop along the same social cleavage lines that we see through out the rest of Europe. He shows that it can not be assumed that one of the if not the most important social cleavage is social class. It is theorized that politicized social cleavages develop out of the current situation as a party system develops. Since these emerging democracies developed after the class struggles that affected Western Europe, it seems that they may be immune to the class struggle cleavage. In these Eastern European countries we see that the membership of the political parties is centered around different lines other than social class.
Angrist examines the case of Turkey and the formation of its party system in a region that contains authoritarian regimes. She includes 10 cases in the study: Turkey, South Yemen, Algeria, Iraq, Iran, Jordan, Egypt, and Syria. Two theories an external one and an internal one about Turkish pluralism state that externally, international Cold War influences affected Turkey or internally the “middle class is politically pluralistic,” Authoritarianism is explained by class structure, socioeconomic development and political culture (229). A third explanation is that “the nature of nascent part systems significantly affected the type of political regime” that emerged (229).
“Single preponderant parties,” parties that “exerted control in all (or nearly all) rural and urban areas,” along elite preferences and cultural norms, led to authoritarian regimes after either Britain or France left the region. This was the case of Tunisia, South Yemen and Algeria.
In the other cases, multiple parties were formed prior to independence by indigenous elites and survived independence. All the countries, except Turkey, became authoritarian eventually. The cases of Iraq, Iran, Jordan, Egypt and Syria all exhibited polarization of the parties. One segment of society, the Conservative forces, rallied for continuation of the political economic status quo, while the challengers wanted to alter it. This led to suppression of the challengers and eventual dissolving of democratic rule.
Turkey exhibited the same patterns as the previous mentioned multiple parties in which authoritarian rule was the result. Turkey differed, however, because unlike the previous cases, in which the challenger parties also defected from democratic rule to pursue other means of power, Turkey’s challenger party (Progressive Republican Party) did not defect. Other challenger parties (Free Republican Party and Democrat Party) arose to contest the Conservative status quo party (Republican People’s Party). In 1950, after the Democrat Party(DP) successfully convinced the Republican People’s Party(RPP) that DP’s followers were opposed to secularism and republicanism, RPP allowed free and fair elections. The challengers succeeded in gaining the support of other elites and elite networks that “opposed unchecked central government power” (242). Angrist concludes that using a class approach to the study of party systems because it uses agents to specify the party system, thereby bringing the actors into the study of the system and allowing for decisions to be made by those actors as to the course of the party system.
Bielasiak, Jack. 2002. “The Institutionalization of Electoral and Party Systems in Postcommunist States.” Comparative Politics 34: 189-210.
Bielasiak evaluates the institutionalization of party systems in post communist countries by assessing the electoral volatility and fractionalization of parties in these countries. He uses two perspectives to view these countries, “tabula rosa” and the structure perspective. The tabula rosa perspective sees weak party systems as a condition of the rapid breakdown of communism. The structure perspective sees these countries as defined party systems, some weak and some strong. To address this debate Bielasaik assess the institutionalization of part systems. Institutionalization is defined as “the extent to which the process of party formation and electoral competition is ‘well established and widely known, if not universally accepted’” (191). He does this by looking at the electoral system (the “rules of the game”) and political actors and their supporters.
Bielasiak finds that the electoral system is being altered. He finds a trend toward proportional representation from single member districts. He also find that election rules have been altered with most adjustments taking place “ in the initial period of the transition for communist to democratized politics, between the breakaway and founding elections. He interprets this as support for the structure argument. He also finds that electoral volatility is higher in post communist nations averaging 28 percent, “twice that of earlier democratizers (28 versus 14 percent)” (198) with little change over time. These countries also exhibit higher number of parties than other party systems. Bielasiak concludes that the party systems are still being structured in many post communist states.
Mair, Peter. 1979. “The Autonomy of the Political: The Development of the Irish Party System.” Comparative Politics 11: 445-465.
Mair analysis the development of the Irish party system through the use of framework developed by Sartori. The Irish party system, Mair explains, has been difficult for researchers to crossnationally because to the development of the Irish party system. The system developed out of civil war and the two largest parties, Finna Fail and Fine Gael, developed because of this war and have unique social bases. Another majority party, the Labour party, is weak and has “not developed a level of voting support equivalent to most of its European counterparts” (445).
Mair divides Ireland’s party system history into three distinct periods. The first is what he calls a modified form of Sartori’s “polarized pluralism” which exhibited a relevant antisystem party (Finna Fail), center placement of a second party, extreme polarization, and “highly centrifugal competition between the two major protagonist, Finna Fail and Cumman na nGaedhel [Fine Gael]” (447). The second is a “moderate pluralism” system in which there was small ideological space and coalitional tendencies. The center was very small but “the party system stopped short of being reduced to simply a two-party system” (450). The final period was a predominate party system in which the Finna Fail was the majority party.
The assent of Finna Fail to the majority party required it to become essentially a “catch-all” party. It began as a party “primarily based among small farmers and petit bourgeoisie,” and extended is support base to encompass “all sectors of the population, from all regions of the country” (452). It did this by removing the cleavage structure from its foundations, and emphasizing its “ability to govern” over all other parties. When other parties gained enough votes to form a coalitional government, as was the case in 1950’s, divisions in the coalition strengthened Finna Fail claim to government, thus solidifying support for the party and solidifying the party system.
Rose, Richard, Neil Munro, and Stephen White. 2001. “Voting in a Floating Party System: the 1999 Duma Election.” Europe-Asia Studies 53: 419-443.
“The electorate can only hold representatives accountable if parties persist from on election to the next” (419). Rose, Munro and White examine the emerging party system in Russia and its prospects for stabilizing; specifically they look at the elections in 1999 and compare them to previous elections. They establish four criteria for a stable and accountable party system in which (1) parties compete nation wide, (2) these parties “persist from one election to the next,” (3) these parties win a significant portion of the vote, (4) these parties’ candidates affiliate with the parties they ran under (420-21). Under these criteria, the authors speak of Russia as a “floating party system” in which “parties competing for popular support change from one election to the next” (420).
In Russia the authors find that parties are unable to survive from one election to the next and most do not run nationwide, and are therefore unaccountable. They form instead to win single districts rallying around a single candidate. Candidates also change their affiliation after the election, and form new parties in the Duma to gain party privileges. This is done because the rules of the assembly define a party as “group to which 35 members affiliate” (425).
They also examine the electorate in Russia using the New Russia Barometer for the 1999 election. They find that because to the “floating system” respondents were unable to identify with a party. Using discriminate function analysis, the authors are able to discern 55% of the voters who voted for the six successful list parties in the 1999 election. The results highlight three areas that divide voters: evaluation of Russia’s political and economic transformation in the past decade, political values ideology and ideological antipathy and divisions on social structure and political personality.
Leff, Carol Skalnik and Susan B. Mikula. 2002. “Institutionalizing Party Systems in Multiethnic States: Integration and Ethnic Segmentation in Czchoslovakia, 1918 – 1992.” Slavic Review 61: 292-314.
Leff and Mikula examine the development party systems in Czechoslovakia during the first Czechoslovakia Republic (1918 – 1938) and postcommunist Czechoslovakia (1989 – 1992) emphasizing the role of ethnonational cleavages and the role of institutional factors that shaped competition in each system. Ethnic cleavages form voting cue for the electorate, especially in postcommunist electorates where “transitional class structures and anomalies of class under socialism are slippery electoral reference points for voter” (295). They also “focus voters attention on potent evocations of basic identity” (295). Institutional factors in newly formed systems can help and hinder the formation of ethnic parties.
The cleavages present in Czechoslovakia at the time of party formation under the first republic “could arise from either multinational diversity or socioeconomic asymmetry” (299). In the case of Czechoslovakia, two distinct regions taken from Austria and Hungry formed the new republic: Czech lands, which were industrialized and populated by German and Czech ethnicities, from the Austrian half; and Slovakia, which less populated and less economically developed, from the Hungarian half. Prior to formation, the Czech lands under Austrian authority had experienced party competition. Slovakia on the other hand had no such experience. When a provisional government was established, Slovakia was also excluded from the negotiations. As such, one faction in the newly formed state, the Slovaks, was heavily disabled in the party system. This led to unequal representation in government and fragmentation in the system, as well as heavier fragmentation among Slovaks.
Miller et al. do find that information has a higher impact in supporting specific parties than socialization. The reasoning behind this is that the older members of the population in these areas have backgrounds in regards to the government than the current situation. Therefore, information on the new style of governance is likely to have a greater impact than socialization in developing partisanship ties. The article also shows evidence of parties representing varying issues that go hand in hand with social cleavages. These cleavages are typically centered on different issues in each country, for example education and income in Lithuania or gender and residence in Russia.
Sitter (2002) not only utilizes the idea of cleavages in illustrating party system emergence, but also the influence of party leader strategy on party system stability and change. This article maintains that along with cleavages, voting patterns, the organization of political parties, and institutions also affect the party strategy of mobilizing voters and building alliances. Once these alliances are made and voters are mobilized, stability or change can happen within a party system.
Sitter’s theory is tested by using party systems of Eastern Central Europe. After incorporating opposition and competition among parties to his theory, Sitter is able to show that competition predicated on the left-right dimension creates volatility for parties because personality of candidates is significant for voters. If parties engage in cross-cutting competition, voters are confused and are not sure on how they should align themselves. And the parties at the extreme left or right are shown to have a declining membership base. All of these different outcomes stimulated from the competitive strategies might usher in a change in the current party system.
Like Sartori, Zielinski (2002) notes that the theory of social cleavages is pretty unspecific about the conditions under which a cleavage is activated and political conflict grows or diminishes. Zielinski also believes that cleavages are important to the party system but he has a different point of view on how these cleavages and the number of parties affect party systems. Zielinski’s believes that the number of political parties in a political systems gives way for certain cleavages to be taken up by these particular parties. This in turn, causes political competition among parties and then a strong and stable party system arises.
And finally, Mozaffar et al. (2003) are interested in the uniqueness of African political party systems with high volatility and low fragmentation. Their article gives evidence proposing that African party systems are somehow different from party systems in the rest of the world. To me, this advances the idea of African exceptionalism; akin to how the South is sometimes treated when analyzing various aspects of American politics. Mozaffar et al.’s information is taken from 62 legislative elections in thirty-four different countries in Africa that held multiparty elections from 1980 to 2000. They sought to analyze how ethno-political cleavages and electoral institutions interact to determine party system size. It seems as if much of the newer literature is trying to mesh both the social and institutional framework when explaining a specific independent variable. In this case, it just happens to be party system emergence.
Mozaffar et al. show that district magnitude greatly decreases the number of electoral and legislative parties and that high ethno-political fragmentation is likely to decrease the number of parties in Africa. These findings are a bit surprising since much research has shown that in other regions of the world district magnitude and ethnic fragmentation actually increase party system size (Duverger 1954, Neto and Cox 1997). Such a reverse of several previous findings by other articles begs one to ask the question of whether Africa really should be treated differently from the rest of the world, or if the authors were wrong in their specifications. Is it okay to assume that all regimes in the region operate in the same way regardless of whether they are truly democratic or not? I have a problem believing that all of the elections they analyzed in 34 countries were conducted in the same way and in a democratic manner. This could have very well contributed to the reason why their findings run counter to past research. Perhaps future research would want to look at truly democratic elections (not just “emerging democracies”) and compare those results to the research that has been done in other parts of the world.
March 31, 2008
POL 628- Comparative Parties and Elections
Party System Emergence
The debate between the institutional approach and the social cleavage approach to the study of party systems is one that is ongoing in comparative politics. The purpose of this paper is to provide a brief overview of each approach, discuss, relative to this week’s readings, what each approach can tell us about the development of party systems in new democracies, and offer some comments as to how or whether reconciliation between the two approaches may occur.
The social cleavage approach argues that party systems result from social conflicts (or cleavages) that are played out over centuries (Ware 1996,186). These cleavages come to define the party system when political parties form around them. How the conflicts are resolved impacts subsequent conflicts, how those conflicts will be resolved, and the types of alliances that will form. For example, party systems in
The social cleavage approach has its theoretical origins in Lipset and Rokkan’s Party Systems and Voter Alignments (1967). In particular, this work introduces the idea that party systems become “frozen”. More specifically, Lipset and Rokkan find that the party systems of Western European nations in the 1960s reflected the cleavages of the 1920s, not the 1950s or 1960s. The social cleavage approach views institutions as endogenous to the party system, noting “they result from the interplay of social forces” (Ware 1996, 189). Adherents to social cleavage theory reject the idea that changing an institution would change political outcomes. That is, institutions do not cause a particular party system to develop.
By contrast, the institutional approach contends that party systems are shaped not by social cleavages but by institutions, like electoral rules or the structure of the state. For example, Duverger argued that “the simple-majority single-ballot system favors the two-party system,” whereas proportional representation prevents two-partism and encourages an increase in the number of minor parties (Ware 1996, 191).
Neither the social cleavage approach nor the institutional approach can fully account for party system development. Each approach tends to ignore what the other has to offer. For example, Ware (1996) notes that early institutionalists tended to focus on electoral systems at the expense of other institutions. As a consequence, the independent effect of electoral systems may have been exaggerated (often). Similarly, institutionalists note that the social cleavage approach cannot explain the development of radically different party systems in countries that have similar sociological characteristics (e.g. religious homogeneity, common language, very small ethnic minority, industrialization).
This week’s readings apply the social cleavage approach and the institutional approach to emerging party systems in
Of particular interest to this paper are three axioms of the social cleavage theory as developed by Lipset and Rokkan (1967). First, party systems are defined by the social cleavages present in a society. Secondly, party system development is a long process that takes many decades. Finally, at some point, party systems will “freeze”; after freezing, a party system will exhibit political divisions that correspond to the cleavages that defined a society decades earlier. This week’s readings note that these principles do not apply to emerging party systems.
With regard to the first axiom, that party systems reflect social cleavages, Sitter (2002) finds that party systems in
Turning to the second axiom, the assertion that party system development is a long process, Miller et al. (2000) find that citizens in former Soviet states are beginning to develop partisan attachments. This suggests that party identification need not be a byproduct of childhood socialization (487). With the increasing availability of information, individuals are able to educate themselves about the platforms of new political parties and choose the party that best represents their interests. This suggests that a party system can develop in only a few years as opposed to a few decades.
Finally, Lipset and Rokkan (1967) also posit that, after a time, party systems will “freeze”. Zielinski (2002) notes that Eastern European party systems froze before the emergence of class conflict. This begs the question of whether class can be translated into a cleavage, if class conflict develops at some later point. Using formal logic, Zielinski argues that new cleavages can be politicized but only under a very specific and limited set of conditions (203). Thus, the class cleavage may never be politicized in Eastern European states, making their party systems different from their Western European counterparts. If Zielinski’s argument holds when tested against data, it would refute the notion that once a party system freezes, no new cleavages can be politicized.
As noted earlier, neither the institutional approach nor the social cleavage approach can fully explain the emergence of new party systems. Mozaffar, Scaritt, and Galaich (2003) illustrate this point. They incorporate both electoral institutions and ethnopolitical cleavages into their account of the development of party systems in
Because earlier explanations of party system development in
 For example, with regard to Duverger’s Law, the causal link between electoral rules and two-partism is not necessarily clear. The suggestion that social cleavages drive the development of a two-party system is equally likely.
The volatility of a party system is the significant shifts in the support of parties in a party system. But volatility does not necessarily preclude a political realignment, where there is a long-term shift of a subset of the electorate to support of a party. Both Scotto et al. (2004) and Roberts and Wibbels (1999) show this is the case. Volatility also does not necessarily show that a change in a party system is evident. Ware (1996) points out the Danish case in which there was a “dramatic decline in support for the Radical Liberal Party and the emergence of the Progress Party and the Centre Democrats,” which had little impact on the party system in the long-term effect. This case suggests the durability of party systems. The question explored then, is what attributes to the stability and instability of party systems, which brings us back to institutional and sociological explanations mentioned above.
Institutional explanations of party system stability are found in Mair (1990) in an essay by Wolinetz and in Roberts and Wibbels (1999). Wolinetz argues, “Party alignments change when there are either significant changes in the number of parties competing or where there are substantial changes in the proportion of the electorate won by different parties” (Mair, 1990, 221). In examining Kirchheimer’s argument for the emergence of “Catch-all” parties, Wolinetz examines the same countries as Kirchheimer. He finds that three of the countries share “common legacies of disrupted political development;” and that France and West Germany share “legacies of crisis” and “changes in the rules of electoral competition,” all of which are institutional factors. Roberts and Wibbels (1999) show that volatility in Latin America is due to institutional factors, in turn causing instability in party systems. “Electoral volatility reflects broader patterns of political turbulence and institutional change or uncertainty” (585). Realignment of cleavages in Latin America is not evident in Roberts and Wibbels (1999), “class cleavages in Latin American clearly have not reached the level of organizational closure or structured partisan competition as they did historically in Western Europe” (585).
Evidence for the Sociological argument again deals with the cleavage structures in party systems. Where Roberts and Wibbels found that cleavages had not developed sufficiently for voter alignments in Latin America, Scotto et al. find this to be the case in Canada. Scotto et al. find a cleavage between Quebec voters and the rest of Canada along ideological lines. The voters in Quebec continue to support Bloc Quebecois (BQ), which is a minority party, and the Liberal Party continues to dominate. The voters for BQ have to found to vote their preferences for Quebec issues in elections and garner a sizable bloc of seats. This along with the emergence of distinct ideologies has led to a change in the party system of Canada from a “two plus” party system to a genuine multiparty system with a dominate party. The evidence of a unique “Quebec cleavage” attests to the sociological argument. Flanagan and Dalton (Mair 1990) argue that the “rise and ellipse of social cleavages” help to explain party system instability and in turn change trough models of realignment and dealignment. The major assumption of this argument, which is questionable (Scotto et al.), is that party systems reflect the social cleavages in a nation (232). The argument follows that resolution or emergence of cleavages may alter party alignment and in turn party systems. Flanagan and Dalton argue that over time the issues dividing parties diminish and parties positions begin to blur, which causes a “weakening of party attachments and increased volatility. This volatility can cause realignment as other issues become more salient. Likewise dealignment can lead to similar changes. Party system are dependent on the established parties in the system as to weather party systems change because of these cleavages, “weather a new-issue cleavage is represented by the existing parties, or new parties will depend, in part, on the established parties’ response to the new issue” (245). Ware offers some insight into the previous mentioned dependency of the party system on established parties and on the social cleavages. Ware sees parties as “office seekers” looking “to attract voters by what ever means they could and from whatever social groups they could,” a catch-all party. These parties would steal be vestiges of the old “frozen” class cleavages though (227). The composition of the electorate can also cause change the party system of a nation and I include it under Sociological explanations because they deal with new members of the electorate that share a common ground and cleavages can be potentially formed from these groups. The in-migration of new citizens has a affect on the party system by altering the existing cleavage lines and be incorporated into the existing parties or forming new parties (Ware 221). The same can be said of territorial change in a nation. The coming of age and dying of voter also alters the party system and can change the balance of support for the parties or in the long term the number of parties in a system can be affected in rare cases. The generational change though is very slow, as most voters entering the electorate are socialized by the previous generation, never the less party system change occurs (223).
March 24, 2008
POL 628- Comparative Parties and Elections
Party System (In)stability
With little exception, the party systems of Western Europe have been relatively stable over the past several decades. There are several reasons that this should be the case. Persistent attachments to existing political parties should help ensure that a party system remains stable. Wolinetz (1990) finds that “strong and persistent attachments to political parties put a brake on the emergence of catch-all parties” (221). Similarly, the continued electoral success of older parties imparts continuity and stability to party systems (Ware 1996). Institutions serve as an important source of party system stability. As Ware (1996) argues, institutions change slowly; any dramatic changes tend to come in response to democratic failure. “[P]arty systems tend to be preserved by the institutional context in which they were formed” (220). The transmission of partisan loyalties from one generation to the next is also an important source of stability. Children tend to inherit the partisan predispositions of their parents. However, as several authors note, generational replacement is also an important source of change, insofar as children orient themselves to the political system using different cues from those used by their parents.
That there are several factors which serve to maintain stability does not suggest that party systems are immune to change. As Flanagan and Dalton (1990) note, party system change can be explained either in terms of realignment or dealignment. There are several possible theories regarding the forces driving these two phenomena. For example, realignment may be attributed to shifts in voter loyalties, changes in the ideological leanings of the masses, or a rise in postmaterialist issues. The purpose of this paper is to examine these competing explanations for party system change.
Realignment is one possible explanation for party system change. The social cleavage model suggests that “change in party systems is attributable to the rise and eclipse of social cleavages;” realignment occurs “as parties and their electorates adjust their positions along a new cleavage dimension” (232). Party systems change as the cleavages that defined the original system decline in salience and are replaced by new cleavages. Unlike the frozen cleavages posited by Lipset and Rokkan (1967), Flanagan and Dalton argue that increasing affluence has caused class cleavages to become less central to defining the party system. Generational replacement explains the weakening of the class cleavage—younger voters are less likely to use class to orient themselves to politics. In place of class, party systems are becoming structured around non-economic issues, like environmental protection. Flanagan and Dalton’s argument is not that old cleavages are disappearing, only that they are declining in relevance and are being overlaid by new, postmaterialist cleavages. While class-based voting has declined across advanced industrialized nations, there has also been resurgence in some traditional cleavages (e.g. ethnic and religious) that were not fully resolved before they were eclipsed by economic cleavages. Currently, postmaterialist issues are embraced most fully by younger generations; however, as the cleavages become institutionalized, they are likely to take on an identity with particular social groups.
While Dalton and Flanagan maintain that postmaterialist issues are driving realignment, Wolinetz (1990) contends that party system change is attributable to shifts in voters’ loyalties, but is limited to countries that experienced a disruption in democratic development or where partisan loyalties were weak. Beginning the late 19th and early 20th centuries, party alignments reflected a combination of organizational pressures (like those from churches or trade unions), class or group loyalties, and ideological predispositions (223). As long as these forces pointed clearly to the choice of one party, voters were likely to develop a durable attachment to a particular party that persisted over time and was passed from one generation to the next. Changes in party alignments should occur only if voters with substantially different loyalties enter the electorate or if voters shift their preferences.
Similarly, Scotto, Stephenson, and Kornberg (2004) find that shifts in ideology can produce realignment. Examining the situation in Canada following the 1997 and 2000 elections, the authors note that “the voter’s ideological positions on issues had a significant impact on the probability of their casting a vote for a particular party” (474). These shifts led the Canadian party system to transition from a two-party-plus system to multiparty system following these elections.
There are at least two possible criticisms of the argument advanced by Scotto, Stephenson, and Kornberg, however. First, the authors fail to provide their argument with a solid theoretical foundation. Indeed, they begin testing their hypotheses with an exploratory factor analysis. Rather than having some theoretical reasons for believing that particular ideological dimensions will be associated with each other, they simply perform a factor analysis using all the available data and then explain ad hoc why various ideological dimensions are related. This is not good social science. Good social science is driven by theory, not data. Secondly, the authors conclude after only two elections that Canada has developed a multiparty system. An equally plausible explanation is that these two elections were anomalies; it may be that 2003 witnessed election results that resemble the two-party-plus system. Perhaps their conclusions should be accepted with the proverbial grain of salt.
As an alternative to realignment, Flanagan and Dalton (1990) contend that dealignment may also explain party system change. Realignment views instability as a temporary state; eventually, new cleavages will be institutionalized as parties form around them. However, with regard to dealignment, instability may be permanent. Citing evidence for the decline of parties, Flanagan and Dalton note that parties eventually may be replaced by other institutions, like interest groups, that will link voters to government. However, interest groups are not parties. Interest groups represent a single issue, while parties represent a wide range of issues. Unlike interest groups, parties put forth candidates for election. Additionally, as quasi-public institutions, parties are subject to government regulations, whereas interest groups are purely private organizations. Given these important differences, it seems unlikely that interest groups could truly serve as replacements for parties. Flanagan and Dalton offer no evidence to support their claims.
Accepting realignment as the general explanation for party system change, one can conclude that party systems transform because they no longer meet the needs of advanced industrial societies. New political cleavages emerge and are reflected in the party system. Rather than the historical forces cited by Lipset and Rokkan (1967) that shaped the original party systems of Western Europe, new party systems are shaped by the demands and pressures of and advances made by modern industrial societies.
The author’s purpose is not so much to test her own theory of electoral volatility but to set three approaches against one another. She looks at the institutional approach which states that institutions make it easy for new parties to enter the system, older parties usually have less volatility, increased ideological distance between the parties means less volatility, and the more effective number of political parties equates more volatility. The next approach, economic voting, concerns the fact that if voters make decisions based on recent economic performance, volatility is highest when inflation is high. And finally, the cleavage approach states that “electoral volatility will be lower in societies with well-structured and salient social cleavages.”
She finds no evidence confirming that presence of cleavages had a stabilizing result on the party system in Central and Eastern Europe. She gives extra attention to the ethnic and urban-rural cleavages, and concludes that these cleavages only influenced electoral stability during declines in the economy. The nature of the party system therefore became fluid since the cleavages did not have the ability to construct party competition. The arrangement of parties and voters through cleavages had to be administered in a political environment without any previous sources of institutional and social identities.
One issue with this article could be that Russia is a major outlier when it comes to ethnic heterogeneity (and could be with other variables as well), given how much larger it is that the other countries in her study. A more in-depth analysis might have also shed some light on how her regression analysis showed insignificant cleavage variables. She could have considered how many potential winning coalitions were in each country, if there was a rural coalition or an anti-Russian majority. Then she could investigate whether or not each country’s cleavages overlap.
Heath, Oliver. 2005. "Party Systems, Political Cleavages, and Electoral Volatility in India: A State-wide Analysis, 1998-1999," Electoral Studies, 24, 177-199.
Heath presents persuasive survey evidence that electoral volatility in India can be illustrated by the extent to which social cleavages are politicized and polarized by the party system. His cleavage polarization index seeks to gauge the extent to which a distinctive political party depicts each social cleavage. States in which parties can produce cross-cleavage support are accordingly less polarized. To assemble his index, he analyzes the relationship between caste-community and the cluster voted for, and he utilizes an index of dissimilarity to calculate the extent to which political competition is polarized along these caste-community lines.
Mainwaring, Scott and Edurne Zoco. 2007. "Political Sequences and the Stabilization of Interparty Competition: Electoral Volatility in Old and New Democracies," Party Politics, 3(2), 155-178.
This article questions whether parties in new democracies are being institutionalized. After centering their analysis on electoral volatility, Mainwaring and Zoco discover that interparty competition in new democracies is not being stabilized over time. The authors maintain that parties in established democracies helped construct citizens, developing supportive networks and acquiring voter loyalty, therefore extending electoral stability. Parties in new democracies, however, have not had the same mobilizing and empowering effect and therefore have not bought about the loyalty essential for stable party systems. Mainwaring and Zoco are concerned that high levels of volatility signify a level of political uncertainty that endangers democratic consolidation. They demonstrate that there is little evidence to back the idea that parties in new democracies are being institutionalized. The evidence that time (the age of a democracy) does not induce less volatile party systems counters received wisdom. Path dependency and historical sequencing are shown to be the main explanations.
Bogaards, Matthijs. 2008. "Dominant Party Systems and Electoral Volatility in Africa: A Comment on Mozaffar and Scarrit," Party Politics, 14, 113-130.
This article examines the work by Mozaffar and Scarritt in 2005 who identify different puzzling features of African party systems and tries to explain both the low fragmentation and high volatility aspects. The main puzzle is the unique combination of the two. But Bogaards has doubts about these particular findings for several reasons. First, he indicates that the analysis is founded on a database of electoral system features and election results that is only briefly explained and characterized, but seems to incorporate almost all sub-Saharan African countries with multiparty elections, regardless of the nature of the regime, the quality of the elections and ,the number of consecutive elections. But other studies of electoral volatility (including Tavits 2005) have either selected on regime type or controlled for regime type (Remmer 1991). Unfortunately, he points out that Mozaffar and Scarritt do neither of these and this may have skewed their results.
Ferree, Karen. 2005. “The Social Origins of Electoral Volatility in Africa,” Harvard University, Unpublished manuscript.
Ferree asks the question of why volatility varies, especially across young democracies. The two obvious answers to this question concerns institutions and social cleavages. While quite a bit has been written about electoral rules, federalism, presidentialism, etc., not as much has been focused on social cleavages after Lipset and Rokkan or other European analyses. Her paper is primarily concerned with social cleavages, and ethnic cleavages in particular, as an explanation in Africa.
Her model indicates that when no ethnic group (or more than one ethnic group) can form a wining coalition on its own, there will be high volatility and vice versa for when there is a single ethnic group that can form a winning coalition. If there are however multiple possible winning coalitions, volatility will be high. Her empirical test measures average volatility for each country and counts the number of possible winning coalitions in each country along. Interestingly enough, she does not account for other cleavages that might exist in the societies, only ethnic cleavages.
The Sociological Approach emphasizes cleavages, both class cleavages and non-class cleavages. These different cleavages originated during different time periods and still have an effect on party systems today. Ware (1996) defines four lines of cleavages identified by Lipset and Rokken: centre-periphery, state-church, land- industry, and owner-worker. The first cleavage, the centre-periphery cleavage, originated over disputes as to weather a societies religion should be national or international (Roman Catholic) and conflicts over language in nation-states (using Latin or national languages). The second cleavage, the state-church cleavage, arose around disputes over issues, such as education of citizens, between the state and the church. The third cleavage, the land-industry cleavage, centered on conflicts “between agricultural and industrial interests” (Ware, 187). The final cleavage, the owner-worker cleavage, arose around “a conflict of loyalty… those of a commitment to an international revolutionary movement and those of the national polity” (Ware, 188). The sociological approach’s emphasis on cleavages “links the pattern of parties evident in particular countries to patterns on social cleavages” (Ware, 188). The lecture by Kriesi (1998) falls in line with the sociological approach outlined by Ware (1996). Kriesi argues that since World War II a new middle class has emerged which he divides into managers and sociocultural professionals at the extremes and technocrats in the middle. The managers, who are “employees in administrative hierarchies” (168) identify on the right while the professionals, who possess “specialized knowledge and expertise,”(168) identify on the left. This distinction between the professional and managers is similar to the owner-worker cleavage outlined above. The manager has a “high level of loyalty to the organization” while the professional is less likely to identify with the organization. Kriesi advances this cleavage into an argument for a “value cleavage,” where the sociocultural professionals developed ‘left-libertarian” values through their daily experiences at work, and the managers develop an adversarial role to the sociocultural professionals in a similar fashion. Kriesi’s analysis also identifies cleavages other than this class cleavage such as urban rural cleavages, religious cleavages and linguistic cleavages; all of which relate back to the previously discussed cleavages in Ware (1996). Using these lines of cleavage Kriesi shows that electoral systems within Switzerland differ greatly from canton to canton with cleavages overlapping. These cleavages in turn, following Ware (1996), allow for parties to draw upon these cleavages and manipulate them to their benefit, linking the party to a social group on one side of the cleavage.
The institutional approach, outlined by Ware (1996), is concerned with the role that institutions, rules, procedures and structures play in the differences between party systems. The most famous argument for the institutional approach was made by Duverger and is known as Duverger’s Law: “the simple-majority single-ballot system favours the two-party system” (cited by Ware, 1996, 191). This is a plurality electoral system with a single member district. Duverger also argues, “where there is proportional representation… and where there are second ballot systems…there is a tendency for multipartism to persist” (Ware 1996 191). Threshold rules and state institutions also affect party systems. Countries that are presidential or parliamentary or semi-presidential affect party systems, where presidential systems tend toward cooperation among parties. Coppedge’s article falls in line with the institutional approach by examining elections in Latin American party systems. Coppedge (1998) focuses on party competition, fragmentation and volatility as well as ideology in party systems across Latin America in the 20th century. His research is clearly more attune to institutional approach by looking at factors such as the parties as institutions, the rules of the system and state institutions that make up a party system. Among the ideology of parties he finds parties that are extremely ideological and those that are not. He finds highly fragmented and institutionalized and those that are not. What he finds is that Latin America is very heterogeneous with regard to differences in party systems. Party systems vary greatly among countries. Based on the effective number of parties (based on vote share)[ENPV] for instance, and the effective number of ideological blocs (ENB), Coppedge calculates the fragmentation and ideological diversity of different party systems and compares them. He finds that in terms of ENPV that Chile, Ecuador and democratic Brazil are more fragmented then the other countries of study while Chile and Ecuador are more ideologically diverse then other countries. This particular example shows the difference between the institutional approach and the sociological approach. Both approaches have their own merits. Institutions do matter but so do sociological factors. Ware (1996) argues that institutions become more important after the first few years of a newly formed democracy (200). Ware also argues that the sociological approach is better suited to explaining “variations between party systems at periods, and in regimes where, most voters have become tied to a party through identifying with a particular social group to which the party is linked” (201).
The Differences in Party Systems
Ware shows three distinct approaches to the study of the differences between political party systems. The three models Ware uses to describe the differences found between various political party systems are the institutional, sociological and competitive approach.
The institutional approach to understanding the differences between political party systems relies upon a nation’s institutional factors. It uses these institutional factors to explain the differences and allows the scholar to make predictions concerning the type of party system that will develop in a nation knowing its institutions. One of the clearest examples of the institutional approach is from Duverger who states that a system that has single member districts and utilizes a simple majority system will naturally develop a two party system. This is due to the fact that in a system where a single member district and a simple majority naturally sparks a competitive environment where the parties must go head to head and only one may win in any given race. Therefore, two strong parties will emerge in order to try and compete with greater parity.
The sociological approach relies heavily on aspects of the society in which the society resides. It uses the history of the nation to explain the differences in the way political party systems develop. The different social cleavages which develop in a nation will effect the way in which the political party system take form. These historic cleavages will have a lingering effect on the political environment of a nation, and can potentially last well after the cleavage has dissolved. Kriesi discusses this cleavage in his piece where his examines the reduction of traditional cleavages in Western European nations. Some of these traditional, historic cleavages are based on different economic classes, the role of the church, differing religions and different ethnic groups. It is hypothesized that these traditional cleavages helped to set up and create the party systems we see in Western Europe and have held to the modern day.
While Kriesi uses the socialistic approach to explain the party systems of Western Europe and the finds that social cleavages are the most efficient tool for understanding these party systems, it can be argued that this theory does not travel. This is exactly what Coppedge finds in his piece which discovers that the socialistic model does not fit well with the study of party systems in Latin America. Coppedge finds through a long historic examination of a nation and party system is not effective in study of Latin America, this is due to the changing environments and regimes in Latin America. Instead Coppedge opts to a shorter term study and examines the party systems of Latin America on an election by election basis. He shows that these nations have a high volatility of party systems and can swing greatly during very short period so he speculates that there might be more case by case issues that infiltrate each electoral cycle as opposed to the historic cleavages argument.
Kriesi’s argument can be seen as potentially supporting the competition based approach. The party system which develops in a nation is affected by the competitive environmental factors in a nation. For example if there is a party that can dominate the government, then the nation will more than likely remain a dominate single party system until another party can compete with the party in power and as a result change the political party system currently present in the nation. This can easily be seen in several of the nations in the Latin American study of Kriesi, where a nation can have widely different partisan structures over a short period in some cases nations completely changed their party system over the period of a single electoral cycle.
The other approach to studying why party systems differ is institutional in nature. This explanation argues that a country’s institutions are the leading explanatory force when discussing the difference. Duverger saw a immediate link between the British type of electoral system (an institutional example) and two-partism. Duverger’s Law explains that plurality voting is likely to reduce the number of parties in a system if voters are primarily connected to a national political system and that proportional representation essentially inhibits a move towards two-partism.
Ware gives other examples of electoral rules that shape the type of party system. Some of these include the structure of state institutions (presidential, parliamentary, etc.), the threshold rule, informal procedures, and the nature of the party system (federalist, specific constitutional rules, etc). One thing that both institutionalists and sociologists omit is the significance of the function of politicians as “entrepreneurs”, that they may actually create these lines of division within society. Schattschneider points this out in his challenge to the sociological model. He argues that among these divisions within society, some will become dominant in that they form the focus of division between parties and other cleavages become overridden and irrelevant to political competition. Party leaders exploit cleavages that provide access to government for their party. These leaders force order through their attempts to acquire and keep power and therefore politics influences the formation of party systems as much as society does.
Both Ware (1996) and Kriesi (1998) seem to believe that the sociological approach is best able to explain variations between party systems. Kriesi attempts to show that even if cleavage politics has recently been on the decline, there is still not an “end of structuration of politics by social divisions” (165). I am not convinced however that Kriesi was able to be fully persuasive in his analysis. If one basis their analysis of cleavages on Lipset and Rokkan’s work, they are not allowing for change that might occur in society. Do these lines of cleavages supersede everything else over time? Could there be other changes in a given country that might actually create these cleavages instead of the other way around?
In his article on Latin American parties, Coppedge (1998) argues that the sociological view is insufficient when studying the region because their party systems develop and change quicker than the social cleavages. He maintains that the institutional elements are the prevailing determinant of party system changes in the region. Coppedge assembled a widespread classification of Latin American parties, evaluating ideological clarity, polarization, fragmentation, personalism, volatility, and the mean left-right position of parties in eleven countries. In several ways this latter analysis is attractive, mainly because it corresponds to our concepts of left and right in politics.
However, I am curious as to just how exclusive Coppedge’s theory is to Latin America. Perhaps party systems of democratizing countries are similar to one another and therefore institutional elements would be a big factor in all of them and not just Latin America. The occurrence of leaders trying to use and manipulate institutional factors in order to maintain power occurs in many places, for example in Eastern Europe in the 1990s (Mair and van Biezen 2001). Political leaders from the old communist governments still tried to hold on to power and therefore built institutions in order to guarantee their own party’s safety as well as their own. While the sociological argument is interesting, there is no need in totally discounting institutionalism. Perhaps the integration of the two as Ware suggests would be feasible in explaining more, although at the expense of parsimony.
March 17, 2008
POL 628- Comparative Parties and Elections
Party System Differences
Dobell, W. M. 1986. “Updating Duverger’s Law.” Canadian Journal of Political Science 19: 585-95.
Dobell suggests that Duverger’s Law needs to be revised in order to account for the presence of minor parties in Canada and India. Most notably, Durverger’s Law only accounts for the death of minor parties. That is, Duverger suggests that voters will be unwilling to “waste” their votes by supporting minor party candidates where the electoral rules are such that these parties could never gain control of government. Presumptively, when these minor parties consistently failed in their electoral attempts, they would disband and/or be absorbed by one of the major parties. Thus, Duverger’s Law accounts only for the dissolution of parties; it says nothing about the formation of minor parties.
Dobell considers two possible reformulations of Duverger’s Law. The first, posited by Rae (1971), claims that “plurality formulae are always associated with two-party competition except where strong local minority parties exist” (586). Riker (1982) notes a second possibility that:
plurality election rules bring about and maintain two-party competition except in countries (1) where third parties nationally are continually one of two parties locally, and (2) one party among several is almost always that Condorcet winner in elections (588-89).
Rejecting each of these reformulations, Dobell concludes that minor parties exist in Canada and India for three reasons. First, voters do not consider the major parties to be acceptable alternatives to one another (592). Secondly, Leftist parties refuse to unite with the more liberal of the major parties or with each other and insist on pursuing elected office under separate party labels. Finally, minor parties exist “because of the frequent absence of more than one party possessing the plausible image of being a national party” (594).
Gross, Donald A. and Lee Sigelman. 1984. “Comparing Party Systems: A Multidimensional Approach.” Comparative Politics 16: 463-79.
In an attempt to facilitate comparison of party systems cross-nationally and over time, comparativists developed a set of numerical indices of different party system attributes. While this approach makes comparison easier, requiring only that the researcher compare one value to another, it greatly underestimates the complexity and diversity of party systems. Focusing on a single attribute increases the likelihood that party systems will appear similar when, in fact, they are not or that unexpected similarities will be ignored (463). Gross and Sigelman argue that a method for comparing party systems should take multidimensionality into account.
Party systems are structures of representation and articulation (464). Parties attempt to gain control of government by having their candidates elected to office. Once in office, parties serve an articulation function as well by enacting their policy or ideological program through government policy. A complete classification of party systems, then, should allow for comparison across various dimensions of representation and articulation. Gross and Sigelman posit a four-prong system of comparing party systems; party systems can be compared in terms of party fractionalization (the degree to which the party system is competitive), ideological fractionalization (the degree to which parties are distributed along the left-right continuum), the location of the ideological center of gravity (where does the ideological center of a party system fall along the left-right continuum?), and ideological polarization (the extent to which a party system has parties that are located at the extreme ends of the left-right continuum, with no or few parties occupying the center of the continuum). Accordingly:
every party system can be seen as occupying a position on the continuum representing each attribute. Moreover, the attributes are…separate from one another…Since we are dealing with four conceptually independent attributes, each party system is represented as a point in a four-dimensional space…[and] numerical quantities are assigned to the interpoint distances (467-68).
Taking this approach, forty-six party systems are compared. Data are taken from the 1979 Britannica Book of the Year; party systems that “contain more than purely token opposition” to the dominant party and that can be coded along the left-right continuum are included in the analysis (469). Using multidimensional scaling, the authors find that their four-part system of comparison is overly complex; rather, party systems can be compared along two dimensions, party and ideological fractionalization and polarization. Gross and Sigelman also find that “leftist-oriented systems tend to be more fractionalized and more polarized than do rightist-oriented systems” (476).
Inglehart, Ronald. 1984. “The Changing Structure of Political Cleavages in Western Society.” in Russell J. Dalton, Scott C. Flannagan, and Paul Allen Beck, eds. Electoral Change in Advanced Industrial Democracies: Realignment or Dealignment? Princeton: Princeton University Press, 25-69.
While the argument advanced by Inglehart is not directly relevant to this week’s discussion of party system differences, it is the counter to or revision of the argument advanced by Lipset and Rokkan (1967) and, thus, is included here. Lipset and Rokkan argue that party systems are based on cultural and class-based cleavages that developed given a country’s history. By contrast, Inglehart argues that developing party systems are shaped by the demands and pressures of and advances made by modern industrial societies and that existing party systems are being transformed by these same forces.
Inglehart asserts that traditional left-right cleavages, which involved conflict over the ownership of the means of production and distribution of income, have been replaced in advanced industrial nations by cleavages that tend to cut across this left-right dimension. The new cleavages are structured along a materialist-postmaterialist dimension. “The rise of postmaterialist issues tends to neutralize political polarization based on social class” (28). Postmaterialist issues include the environment, nuclear power, the women’s movement, the consumer advocacy movement, and the peace movement. Differences between traditional party cleavages and contemporary issue cleavages place the party system under stress; at some point, a restructuring will occur. This restructuring may come in the form of new political parties or the selection of new elites to take over existing parties.
While political party alignments may have been stable for generations and continue to shape voting behavior, traditional cleavages will not provide individuals with the initial motivation for becoming politically active. Those mobilized by postmaterialist issues tend to come from higher income groups, but are dissatisfied with society and open to social change. Additionally, they tend to side with the Left. When postmaterialist issues become salient, they may stimulate a materialist reaction, whereby the working class will side with the Right to reaffirm materialist interests, like economic growth and national security. Clearly, this alignment of the working class on the Right and the more affluent on the Left is just the opposite of what occurred when parties were organized along class-based cleavages. These claims are tested by surveying a representative national sample of publics in nine nations in combination with a sample of 742 candidates running for European Parliament.
Lipset, Seymour Martin and Stein Rokkan. 1967. “Cleavage Structures, Party Systems, and Voter Alignments: An Introduction.” in Seymour Martin Lipset and Stein Rokkan, eds. Party Systems and Voter Alignments. New York: Free Press, 1-64.
History causes party systems to differ from one country to the next. Lipset and Rokkan argue that history influences the development of social cleavages, which become political cleavages. These cleavages are ultimately expressed in the party system (2). Broad structural cleavages, based on culture, social class, and ideology, developed in the early part of the twentieth century along which alliances were formed. These cleavages resulted from both national and industrial revolutions. National revolutions forced individuals to choose sides based on territorial or culture identity, while the industrial revolution forced individuals to choose sides based on economic interests (14). The alliances that were formed along these cleavages reflected these territorial/cultural and economic distinctions. Territorial/cultural alliances were formed between people who were from the same region or who shared a common culture. Economic cleavages led to the formation of alliances among members of the same social class. Functional alliances were formed among individuals with a common ideology. These alliances were ultimately expressed in the party system via political parties. The impact of cleavages on the structure of the party system is the result of the ease or difficulty each coalition or alliance experienced in gaining representation within the political system.
Powell, G. Bingham, Jr. 1981. “Party Systems and Political System Performance: Voting Participation, Government Stability, and Mass Violence in Contemporary Democracies.” American Political Science Review 75: 861-79.
“The writings on democratic party systems concur in characterizing party systems with substantial extremist party voting as ‘weak.’ But theorists disagree on the desirable attributes of ‘strong’ party systems” (876). Powell examines competing explanations for determining the strength of a party system. The dominant view suggests “that a strong party system must be able to ‘moderate and channel participation of newly mobilized groups in such a manner as not to disrupt the system’” (861). Alternative conceptualizations argue that party system strength can be evaluated in terms of aggregative majoritarianism (the party system is dominated by one or a few parties), responsible majoritarianism (the party system is dominated by one or a few parties that are meaningfully different in terms of policy program), fractionalization (the party system lacks a dominant party or parties and parties do not have strong linkages to voters), and representation (the party system lacks a dominant party or parties but parties have strong linkages to voters and social groups).
Using these alternative views of party system strength and separating extremist from non-extremist party systems, Powell classifies the party systems of thirty democracies. He then compares them in terms of system strength and performance. System performance is measured by factors such as turnout, executive stability (if the executive’s tenure is ended prior to the next scheduled election for reasons other than death or retirement for health reasons), instances of riots, the number of deaths resulting from political violence, and whether democracy has been suspended.
Powell concludes that strong support for extremist parties is correlated with poor system performance. Additionally, there was no relationship between system strength and performance and death resulting from political violence or regime suspension, when controlling for environmental factors, like presence of ethnic minorities and level of economic development. Party systems where there are strong linkages between the parties and various segments of society demonstrate higher turnout. It is also interesting to note that representational systems experience less political violence and fractionalized systems experience the greatest amount of executive instability.
This week’s readings concern the classifications of different political parties and the partisan systems in which the parties reside. The classical approach to party classification is to divide the systems along the two party, or multiparty systems. This classification structure relies solely on the amount of political parties that were relatively competitive in the system. If there were two competitive parties then the system was considered a two party system and if the system contained more than two competitive parties then it would be classified as a multiparty system. This system of classification has been largely relegated to be used as a simple explanatory tool for an elementary understanding of comparative politics.
The large majority of the field has evolved beyond the classification structure that examines only the number of parties that are in the system to a classification structure that allows for a more specification and will allow for an increased ability to compare the different partisan systems.
One of the most widely accepted modern classification structures for organizing the systems of political parties is the Sartori Schema. Sartori uses two main characteristics of the partisan system in his schema to help classify these systems. He uses the level of party fragmentation and the ideological distance between the main political parties in the system. By the use of these two variables, Sartori is able to create four different classification structures for party systems. These categories are the two party systems, the polarized multipartisian system, the moderate mulitpartisian system, and the segmented multipartisan system.
To Satori and other authors of this week's readings the ideologies of the political parties involved is a determining factor in which type of partisan system will develop. It has been found that if the parties are located at opposite ends of a philosophical spectrum this will lead to direct competition between them and if a party develops inside the same part of the political spectrum then it will split the vote of the those that fit in the part of the spectrum and as a result there will be electoral losses.
Another great factor that affect which type of partisan system develops in a country are institutional factors in that country. It has been determined that whether or not a country has a parliamentary system or a presidential system will have an effect on the style of partisan system will develop. As well, the electoral system that is in place in the country will have a strong effect on the type of partisan system. If a system creates strong competition between the parties by establishing a system where each election is a zero sum game, then it can logically be assumed that a two party system will develop. In this style of system where there is a zero sum game because it can be assumed that there is a fixed number of seats, and by having a fixed number of seats each gained seat by a party is at the loss of another opponent party as well any gain by an opponent is costly to the party in that it looses seats. This fact where control is made by having a majority of the seats will lead to a consolidation of all the parties into a two party system.
The authors this week also discuss how the party system a country adopts can evolve over time and they offer several historical examples. In Italy, the top two parties that developed were a communist party and a more western party. It is with this system that there grew great cooperation with all of the parties that opposed the communists and the dominant party adopted a more moderate position on most other issues in an opportunistic model to help maintain there dominant position in the coalition. The authors also discuss how a partisan system can evolve and they use the newly democratizing nations as an example. The nations examined are those of Central and Eastern Europe who are coming out of the post communist era and those newly democratizing nations of Africa. The authors show that it is quite common for a multiple party system to develop and then as the nation gains electoral experience it is quite common for the system to evolve into one of fewer dominant parties or even a two party system or one where the system is focused around one dominant party that controls almost all of the legislative body.
- Gross, Donald A. and Lee Sigelman. 1984. “Comparing Party Systems: A Multidimensional Approach.” Comparative Politics 16: 463-479.
Gross and Sigelman view party systems as a multidimensional concept that incorporates several different aspects to compare party systems. They hold that “party systems are structures of representation,” an assumption they say is basic to every party system classification. They also contend “parties also exist to aggregate and articulate certain political points of view.” With these two points of view they focus on comparison among competition among organizations and also upon competition among “ideological-programmatic outlooks.” From these two assumptions they form four variables for classification of party systems: party fractionalization, ideological fractionalization, ideological center of gravity and ideological polarization. Party fractionalization summarizes the distribution of seats among parties. Ideological fractionalization summarizes the distribution of seats among ideological positions. Ideological center of gravity is the midpoint on the ideological scale among various parties in a system. Ideological polarization is a measure of deviation of parties from the ideological center of gravity. Using these variable they are able to compare party systems considering the four attributes. They utilize this measure to compare forty-six different party systems. They find that a two-dimensional representation of party systems is just as clear as a four-dimenstional model. Party fractionalization and ideological fractionalization have no significant empirical difference furthermore they find that ideological polarization is very important to differentiating between party systems.
- Molinar, Juan. 1991. “Counting the number of parties: An Alternative Index.” American Political Science Review 85: 1383-1391.
Molinar’s review of two indexes used to count parties finds that both of them are insufficient in differentiating between party systems. The effective number of parties index proposed by Rein Taagepera overstates the relevance of large parties in a system while the hyperfractionalization index or Kesselman-Wildgen index overstates the relevance of smaller parties. Molinar proposes a new index based of the effective number of parties index. The difference, he states, is the treatment of the winning party. The new model, called number of parties, counts the winning party, regardless of size, as one and weights the others by their contribution. In this way the problem of overstating the relevance of large parties in a system is overcome. The rationale of the model is that it considers the size of the winning party, the gap between the two largest parties, and concentration of minority parties in a system. This method effectively distinguishes between two-party, two-and-a-half-party, and three or more party systems better then the two review indexes.
- Taagepera, Rein and Matthew Soberg Shugart. 1993. “Predicting the Number of Parties: A Quantitative Model of Duverger’s Mechanical Effect.” American Political Science Review 87: 455-464.
Taagepera and Shugart posit and prove a rule that Duverger observed previously that one seat district lead to two-party systems and that multi-seat districts lead to multi-party systems. They calculate the effective number of parties in a district for the number of seats of the district and in a related model the effective number of parties based on the magnitude and seats in national assemblies. They find that the district magnitude is related to the effective number of parties. The variables of district magnitude and assembly size help to define the party system of a country.
- Blondel,J. 1968. “Party Systems and patterns of Government in Western Democracies.” Canadian Journal of Political Science 1: 180-203.
Blondel offers a categorization of parties based on the number of votes a party receives, the strength of the party, and the ideological position of the party. Six total groups of party systems are identified by this categorization. “On one extreme are the broadly based parties of the two party system countries” and on the other are multi-party systems with small ideologically divide parties. The notion of the two and a half party system is introduced and defined as a system with two large parties and one smaller party centre party also the dominate multi-party system is introduced.
- McKown, Roberta E., Robert E. Kauffman. 1973. “Party Systems as a Comparative Analytic Concept in African Politics.” Comparative Politics 6: 47-72.
McKown and Kauffman include a classification of Party Systems with the categories of no effective party, one-party, one-party dominant, and multiparty in an attempt to assess the usefulness of party system categorization as a variable in the comparative study of African Nations. What they find is that the number of parties that were identified as multiparty party systems have declined this category included two-party systems. Their overall findings are that the structure of the party systems in Africa have no bearing on political, economic, or social chance
The reason behind constructing a classification system comes from the desire to illustrate how parties evolved into their current structure as well as functioning as a guide to party systems (Ware 1996). In his explanation of the formation of parties, Duverger looks at conflict among parties and differentiates between metaphysical and technical dualism as probable causes for the structures of parties. Duverger shows that all parties have factions and when these sides cannot concede there is a propensity to move away from dualism and the parties divide or members leave one party and form another, which sets the stage for a multiparty system (Mair 1990).
Once the foundation has been laid regarding how parties have evolved to their current formation, one can assess which parties should be included when setting up a classification of party systems. The literature seems to be situated on discussion this inclusion or exclusion of parties based on specific standards. Sartori’s method of doing so includes the potential coalition building and blackmail conditions. If a party meets neither of these conditions, the party’s tactics come into play. He suggests that these tactics may increase party competition or change the direction of the current competition among parties. If this is the case, Sartori believes the party should be included in his classification (Mair 1990). That being said, excluding some parties but not others may hamper the accuracy of classification systems. Since classification systems are employed to illustrate why parties act as they do as well as trying to predict the behavior of these parties and party systems, problems may arise out of leaving some parties out. A particular party that was excluded may not be important at that particular time, but could gain momentum and in the future become important political actors. When excluding these parties in we are not able to observe everything in a changing party system.
After figuring out which parties should be part of the classification system, one can then decide on what standards should be utilized in classifying the actual parties. It is here that much contention is found in this week’s readings regarding these standards and ultimately the number of parties within the system. Blondel’s classification is mainly based on the numbers of parties in the system but he also involves the strength of the parties, support, ideology, leadership, and organizational structure (Mair 1990). Dahl’s theory is akin to Blondel’s, but he takes into account not only the number of parties and organizational structure, but the concentration of opposition within a party system as well. As opposed to Rokkan and Sartori, both Dahl and Blondel apply elements of the general behavior of the party and party system instead of purely numerical classifications. Sartori employs party fragmentation and ideological distance in his classification. Within each of these two divisions he evaluates differences within parties which entails the intensity in how parties spread out into society, ideology, stance the party has towards government leadership, and number of parties in each system. In other words, party fragmentation illustrates the size and number of parties.
In an attempt to synthesize these separate notions of how party systems should be categorized, Bogaards (2004) integrates several classifications for eighteen sub-Saharan African countries. In his analysis, Bogaards discovers that the classifications put forth by Blondel, Ware, and Sartori, produce comparable results. Utilizing Sartori’s classification, he is capable of characterizing four types of multiparty systems in the African region: dominant, dominant authoritarian, non-dominant, and pulverized. This classification of party systems steers him to the conclusion that dominant parties are more prominent in this region in Africa.
Bogaards’ findings help bring about a few interesting questions for future research. If the political structure is controlled prevalently by dominant parties, these parties are could very well cause problems for developing democracies. Here is an excellent example of why one should study party classification and/or why it is important in political science. Next, although Bogaards gives a convincing case for using Sartori’s classification, there is always a need for more research on classification systems. The standards used by Sartori may need to be changed a little in order to offset the revamping occurring within parties and party systems.
I also find the cutoff points used in vote and seat shares to be problematic throughout the literature. For instance, Ware proposes a minimal vote share of 40% and a minimal seat share of 45% for parties to be considered dominant (other authors propose their own cutoff points). These cutoff points are typically presented ad hoc without additional reasoning or justification. What is the reasoning behind a party with 39% of the vote being treated differently than one that has 40% of the vote? Furthermore, why exactly should vote shares matter to start with? As mentioned previously, why should parties that always gain a plurality or majority of seats but have vote shares below the proposed threshold be left out? The assumed effects of a party’s eminence or lasting control of government described by some of this week’s authors are clearly not a direct function of its vote shares.
And finally, several of the readings suggest the inclusion of a divided opposition, or a minimal distance between the largest party and the second largest when characterizing single-party dominance. Here again, some of the suggested separation points could be viewed as almost random. In other words, does opposition or party system fragmentation inevitably decide a party’s bargaining position? Would there be a difference for a party holding a majority of the seats if there were two opposed to eight other parties? Or do these numbers even matter?
Party System Classification
Ware (1996) recognizes four criteria that are useful for building a classification of party systems. First, party systems differ in the extent to which parties penetrate society (or how involved parties are in the lives of voters). The degree of penetration affects competition in terms of the ease by which new parties form. Weak penetration makes it easier for new parties to form. Conversely, a high degree of penetration means that most parties will concentrate on maintaining their following, simultaneously reducing the likelihood that these parties will become catch-all organizations and that new parties will develop. Secondly, the ideology of the individual parties must be taken into account when classifying party systems. Some party systems will exhibit a wide range of ideological leanings while others will have a relatively narrow range. Third, some party systems may include anti-system parties- parties that do not accept the premises of liberal democratic governance. If such parties gain control of government, they will attempt to “change the rules” (Ware 1996, 153). Finally, and most controversially (at least given this week’s readings), party systems can be classified according to the number of parties they contain, leading researchers to define party systems as two-party, multiparty, or something in between.
Ware’s criteria provide a good starting point for determining what should be considered when developing a typology of party systems. However, I wonder whether, in practice, they actually lead to a simplified understanding of reality. That is, I can imagine that trying to classify parties according to four criteria could lead to many different combinations of degrees of penetration, range of ideology, stance toward legitimacy, and the number of parties. Given this, I wonder whether there would be more than one or maybe two party systems per category. In the interest of parsimony, it might be better to select a single criterion for classification, although determining which one presents its own set of difficulties. Certainly, several of this week’s authors seem to suggest that the number of parties is the appropriate criterion.
Turning specifically to the issue of counting the number of parties in a given party system, there is little agreement as to how this should be done. That is, researchers disagree as to which parties should be included as members of the party system and which should be left out. For example, Blondel (1990) contends that the number of parties should be determined according to electoral performance. By contrast, rather than counting the number of effective parties, Bogaards (2004) notes that party systems in emerging democracies, like those in Africa, are best understood by using Sartori’s rules for counting (coalition potential and blackmail potential). Alternatively, Dahl (1990) that only competitive parties, those that are able to gain electorally at the expense of their opponent(s), should be included as party system actors.
Early attempts to classify party systems based on the number of parties led to parties systems being labeled things like two-party, two-and-a-half party, three-party, and multiparty (presumptively more than three parties). How one determines which parties to count (like in terms of electoral performance, for example) determines the number of parties one finds in a given system and, thus, how one classifies that system. For example, Blondel (1990) concludes that multiparty systems are those in which two parties obtain 2/3 or less of the national vote and two-party systems are those in which more than 90% of the national vote is split between two parties. Similarly, a three-party system would be one in which the national vote was more or less evenly among three parties (although this has never been witnessed in a liberal democracy). Rokkan (1990) finds evidence of a two-and-a-half party system which is understood as “two large parties running neck and neck for the majority point but generally thwarted in their endeavors by the persistence of one small above-threshold party” (311). This brief review illustrates Sartori’s (1990) observation that there is little agreement on how classification should work, with the result being that every researcher develops his own classification scheme, which does not serve to move political science forward.
As Sartori (1990, 316) notes, counting the number of parties is meaningless. That is, the number of parties in a given system matters, but simply counting may not allow us to understand exactly how it matters. Earlier research considered any party system having more than three parties as a multiparty system. Arguably, there are also types of multiparty systems. Sartori attempts to address this concern by considering ideology and party fragmentation in order to differentiate between types of multiparty systems. In doing this, Sartori identifies four different types of party systems- two-partism, polarized multipartism, moderate multipartism, and segmented multipartism.
Sartori’s method of classifying party systems goes beyond simply counting the number of parties; it indicates how the number of parties, the ideological distance between the parties, and the party composition of the legislature interact to form a particular type of party system. This interaction offers some suggestion as to why the number of parties is important to classifying party systems. At various points throughout the semester, we have criticized party classification schemes for being applicable only to Western European democracies. Sartori’s classification scheme may be especially useful in that, at least according to Bogaards (2004), it is applicable to party systems in emerging democracies outside of