Unpacking LogM: Toward a More General Theory of Party System Density

Abstract

The standard model of political party system density combines two traditions to explain why some countries have more political parties than others, one tradition that emphasizes social cleavages and another that emphasizes electoral institutions, especially district magnitude. Despite its considerable success, there are several reasons to be less than fully satisfied with the standard model. We examine two of these problems associated with the scope of strategic voting and the functional form of the specification used to test the model. In doing so, we contrast the standard interpretation with an organizational ecology model that accounts for what the standard model did so well, but also accounts for important anomalies it ignores. We reexamine some of the key analyses that have been used to test the standard model to assess the severity of its limitations and the utility of the rival organizational ecology account. P olitical scientists have traditionally explained the density of political party systems-number of parties-in two ways. The first, associated with political sociologists such as Grumm (1958), Lipset and Rokkan (1967), and Rose and Urwin (1970), emphasized social cleavages, arguing that parties represent preexisting regional, religious, or ethnic divisions. The second, based on the work of Lijphart (1990), Riker (1982, Taagepera and Shugart (1993), and Cox (1997), emphasized instead electoral institutions. Based on Duverger's (1954) observation that simple-majority, single-ballot institutions favor the development of a two-party system, these scholars explored how a range of electoral rules severely condition how many parties can meaningfully compete for legislative seats and executive offices. These approaches now have been combined into a single integrated or interTaagepera and Grofman 1985). A strong research program has developed around this interactive model exploring individual-and systemlevel properties of and interactions between social systems and electoral institutions (Despite this real success, there are several reasons to be less than fully satisfied with this standard model of party system density. After describing this model, we examine two key problems, contrasting the standard interpretation with an alternative organizational ecology model that accounts for what the standard model did so well, but also accounts for important anomalies it ignores. We reexamine some of the key analyses that have been used to test the standard model to assess the severity

Political Spaces, Dimensionality Decline and Party Competition

“…The most accepted models propose a direct link between district magnitude -i.e., the number of legislators elected in a district -and the number of political parties in a political system [7,36,43,44,48,49,57,58]. However, there are political systems in which these models fail to predict the number of political parties, especially where district magnitude is high [40]. This is due to the fact that issue-driven dynamics has been a missing factor in those models.…”

“…In addition, and more importantly, factors that influence strategic behavior may be pointless in a study of the impact of nonanticipatable events (e.g., a sudden dimensionality reduction) like ours (see e.g., [41]). (ii) We also assume that proportional representation mandates the configuration of government: When proportional representation is part of the governmental configuration rules, voters realize that voting sincerely for the preferred party does not preclude the possibility to get their preferences represented in the government formation process [13,40]. This rules out any possibility of strategic behavior of voters in our model.…”

“…On the one hand, many democracies in the world comprise multiparty settings (e.g., Continental European and Latin American countries) [40]. On the other hand, the number of dimensions affects the opportunities to build a distinctive policy profile.…”

The organizational ecology of ethnic cleavages: The nonlinear effects of ethnic diversity on party system fragmentation

“…Lowery and Gray, 1995;Nownes, 2004Nownes, , 2010Stretesky, Huss, and Lynch, 2012), electoral systems (Lowery et al, 2010), and political parties (Lowery et al, 2013). At present, it remains to be applied to social cleavages.…”

“…Noting that most work has employed the logged functional form of district magnitudewhich implies a nonlinear relationship to party system fragmentation-instead of the linear functional form, Lowery et al (2010) argue that the relationship between electoral system proportionality and party system fragmentation is nonlinear. Drawing from organizational ecology theories, they maintain that at higher levels of proportionality, competition for voters becomes too intense and the availability of resources too scarce to support additional political parties; thus, increases in party system fragmentation level off at the highest levels of electoral system proportionality.…”