Are party systems more the result of social or institutional influences?

First published: ; Last modified:

This content was originally written in .
Footnotes were present in the original; sidenotes are used for comments.

Essay

Within established Western democracies, social influences have been the primary factor

What are the testable differences between the sociological and institutional accounts? Do they predict different things about the future? I presume they must, otherwise it hopefully wouldn’t be such a ferocious subject of debate, but it’s not obvious to me how they do.
determining the sorts of party systems which arose, including through their effect on institutional influences. Institutional factors currently play an important role in bounding the number of potential parties at a district level, but it is social cleavages which determine the actual number of nationwide parties, the lines along which they organise, and (historically) the institutional arrangements themselves. In non-Western nations where democracy emerged suddenly and exogenously, social influences still play a part in setting the dimensions of inter-party competition, but they did not historically contribute to determining the kind of electoral system which arose. For these countries, then, economic factors like level of development and extent of state subsidy play an important role in determining the features of party systems. In this essay, I will first define what I mean by a “party system”, before presenting the theoretical and empirical effects of electoral and institutional arrangements on the party systems which emerge. I then explore the historical role played by social divisions on determining characteristics of party systems, including indirectly by influencing institutional factors, and examine the data on how these social factors map onto party dynamics in young democracies. Finally, I conclude that party systems in developed democracies are mostly the result of historical social influences frozen in place due to institutional arrangements, whilst those in newly-democratic nations can be ascribed to a combination of current institutional factors and economic conditions.

At its simplest, a party system can be thought of as the set of all interactions between competing parties within a country or other political system (Sartori 1976, 44). Although different researchers have produced their own favoured typologies, a common theme is the classification of party systems based on the number and relative sizes of the parties they contain. Taking this approach leads to categories such as one-party dominant systems, two-party systems, and multiparty systems (Clark et al. 2012, 611). Quantitative analysis of the parties present can happen at both the district and national level, but the nationwide picture is what is usually considered when classifying a party system (Cox 1999, 27). This means that the extent of nationalisation of parties within a polity is an important explanatory variable for the resulting size of the party system. Apart from the number of political agents vying for votes, their relationship with citizens and thus manner of competition can also differ. In indirect policy exchange systems, voters delegate power to representatives because they expect them to produce the most beneficial long-term legislative outcomes, whereas clientelistic systems emphasise a direct responsibility of the representative to provide specific, immediate goods to each of their supporters. This affects whether parties compete by offering different programmes of government to voters, or by arguing that they would do a better job at delivering a given set of promises (including selective rewards and inducements) – positional and valence competition, respectively (Kitschelt 2009). Within Western nations, the relationship between voters and representatives is mostly indirect policy exchange, leading to positional competition. As a result, I primarily focus on variations in the number and sizes of parties, but will also touch on the other features of party systems when considering how well hypotheses generalise to newer democracies where the nature of competition is different.

There are compelling theoretical and empirical grounds for believing that institutional influences play a significant role in determining the variety of party system which emerges. Holding the amount of electoral support for each party equal, countries with more majoritarian electoral systems will tend to see fewer parties in their legislatures, purely because of the mechanical effect of how the electoral system translates votes into legislative seats (Clark et al. 2012, 644). Of course, electoral support for each party is not equal, because voters' incentives in choosing who to support is affected by the electoral system in use. Citizens are less likely to support candidates from small parties in majoritarian systems (since there is a high chance that their vote would be “wasted”), which in turn makes it less appealing for such groups to establish themselves as parties in the first place – further reducing the effective number of parties in a polity (Clark et al. 2012, 649). Electoral systems can also affect the strength of parties across the board: countries which employ candidate-centred

There’s an interesting point in footnote 3 here about rank orders of candidate-centredness.
systems such as the single non-transferrable vote or open-list proportional representation are more likely to see weaker, less disciplined parties than those using party-centred systems, because of the increased internal competition the former approach creates (McAllister 2016). Other institutional factors also play a role in influencing the extent to which parties are nationalised. Countries in which decision-making ability rests mainly with national government, and where there is a president with relatively strong powers compared to the legislature, are more likely to see a nationalised party system, because geographically localised parties would lack the power to bring about policy changes and thus lose electoral support (Clark et al. 2012, 657). Cox (1999, Table 2.1) provides extremely strong empirical support for the mechanisms by which electoral systems are theorised to influence party systems
Could explicitly mention Durverger’s Law and Hypothesis, that SMDP leads to bipartisanship and PR favours multiparty systems, respectively.
. By comparing the number of parties found in the upper and lower chambers of legislatures within the same country, the potential confounder of differing social structures can be controlled for, and the effect of differing electoral systems isolated. In all but one of the 15 cases examined, the direction of difference between the number of parties in the lower and upper houses matched the predictions, indicating that electoral systems do play an important role in determining the number of parties which emerge.

However, the evidence also shows a clear relationship between social factors and the number of parties, conditional on the electoral system. Specifically, within proportional electoral systems, the presence of one additional ethnic group is associated with a 1.24 increase in the effective number of electoral parties (Clark et al. 2012, Figure 14.10). In non-permissive majoritarian systems, additional ethnic groups have no significant effect on number of parties. It can be seen, then, that contingent on the bounds imposed by institutional influences limiting the number of possible parties, social structure does impact a country’s party system (Cox 1999, 9). Whilst institutional influences can describe why, for example, Durverger’s law that countries using single-member district plurality (SMDP) voting will have bipartisan party systems (Clark et al. 2012, 653), they are not sufficient to explain either why partisan divisions cut along the lines they do, or why some countries have fewer than the institutionally-implied upper bound of parties. To answer these two questions, we must employ a sociological approach investigating the cleavages, or durable divisions, within a country. In nations where an individual’s position on one dimension of personal attributes is strongly correlated with their position on another (say, native language and regional location), these reinforcing cleavages will lead to natural overlaps in identification, and therefore only a small number of political parties may emerge – even if a permissive electoral system meant there was the capacity for more (Stokes 1999). Furthermore, the precise social cleavages which are mapped on to competitive party divisions will depend on the extent to which voters change their partisan identification in response to parties’ movements along that dimension. Parties will only compete with each other

I’m a bit confused by Kitschelt’s point here.
on a certain dimension if citizens’ partisan positions are not entirely entrenched, another ultimately social factor (Kitschelt 2009). Thus, social influences play an important role in fixing the number of and boundaries between political parties, within the constraints imposed by institutional arrangements.

Crucially, within established Western democracies, these institutional arrangements are themselves socially driven. Countries with historically divided populations were more likely to develop consensus-based democracies with proportional electoral systems, which then permitted the continued existence of multiparty pluralism (Nohlen 1993, 25; in Cox 1997, 19). There is, therefore, an endogeneity problem for those claiming that institutional influences are the ultimate cause of differences in party systems in established democracies. The fact that in new African democracies – where the choice of electoral system was often taken by colonisers without regard for the nature of local social cleavages (Clark et al. 2012, 595) – there is a far weaker link between social factors and party systems supports this view: social cleavages shape the size of the party system within the bounds of institutional arrangements, and thus their role is limited if those institutions are created exogenously. Moreover, the poor economic conditions in these countries facilitate the existence of clientelistic voter-representative relationships, which leads to less positional competition between parties, and thereby weaker incentives for social cleavage-based partisan competition (Kitschelt 2009). In new democracies, then, social factors play a less significant role in shaping party systems.

So, to conclude, both social and institutional influences affect the nature of party systems. Within Western democracies, the gradual, endogenous development of electoral institutions reflected social divisions, and therefore both historic and current social cleavages play an important role in determining party systems. Institutional arrangements set bounds on the potential number of parties present, whilst social factors fix the actual number of parties and the lines along which they are organised. Because institutions in new democracies were imposed from outside, there was much less scope for social factors to influence electoral systems. As a result, institutions in those countries may not be well suited to the social structures, and impose binding constraints on the sorts of party systems which are able to emerge. This explains why there is a weaker link between social cleavages and party systems in new democracies.

References

Clark, W.R., Golder, M. & Golder, S.N. (2012). “Chapter 13: Elections and Electoral Systems” and “Chapter 14: Social Cleavages and Party Systems”, in Principles of Comparative Politics (2nd ed.). London: CQ Press.

Cox, G. W. (1997). Making Votes Count: Strategic Coordination in the World’s Electoral Systems. Cambridge: Cambridge University Press.

Hassan, M. (2013). Institutional factors affecting party systems in new democracies: endogenous or exogenous predictors? Democratization, 20(4), 668–692. https://doi.org/10.1080/13510347.2012.659022

Kitschelt, H. (2009). Party systems. In The Oxford Handbook of Comparative Politics (pp. 522–554). https://doi.org/10.1093/oxfordhb/9780199566020.003.0022

McAllister, I. (2016). Candidates and voting choice. Oxford Research Encyclopedia of Politics. https://doi.org/10.1093/acrefore/9780190228637.013.73

Sartori, G. (1976). Parties and Party Systems: Volume 1: A Framework for Analysis. Cambridge University Press.

Stokes, S. C. (1999). Political Parties and Democracy. Annual Review of Political Science, 2(1), 243–267. https://doi.org/10.1146/annurev.polisci.2.1.243

Tutor’s comments