Party+Systems

Herbert Kitschelt, “Party Systems,” in Carles Boix and Susan C. Stokes, eds., //The Oxford Handbook of Comparative Politics// (Oxford University Press, 2007), chap. 22, pp. 522-54. Caroline Chumo for SIS 700 November 1 **Purpose**: What are the components of party systems? What would it be like to treat party systems as an independent variable? **Thesis:** Party systems play a big role in political economy and public policy, but their theories are lacking. **Overview:** Kitschelt makes an analogy between Waltz’s international systems and party systems, reviews the major theories of party systems, and explores the equilibriums between the number of parties and the number of dimensions of competition, thinking beyond indirect exchange between voters and politicians and beyond conventional positional approaches. **Methodological approach:** Theory building, rational choice theory Analogy: A party is to a party system as a state is to an international system 3 levels of IR theory apply: individual behavior, organizations (parties), and party systems 1. Important aspects of party systems are: Number of players, distribution of resources, rules a. Goal: to predict competitor strategies and identify equilibria of those strategies b. Critical elements: number of competititors; currency/dimensions of competition c. Equilibria involve the number of sustainable players, their pay-offs and alliances d. Real life situations: maintaining government executives, resource extraction/allocation, maintaining/abandoning democracy e. Sample research questions: Does the structure of a party system as the sum of individual and party decisions causally determine outcomes in political processes? f. Assumptions about individuals: “electoral market” -- > voter judgment matters to politicians g. Assumptions about parties: parties are intermediaries to help individuals act collectively 2. Varieties of party systems: most theories assume “indirect” democratic accountability between voters and politicians (rewards to voters come long after the vote), but there are other models a. Critics argue these models are too positional, i.e. focus too much on policy positions b. Direct exchanges between voters and politicians are clientelistic – instant voter rewards c. Valence (as opposed to positional) approaches mean that parties compete to provide policy __better__ than their competitors. d. There is no easy correlation between indirect/direct exchange and valence/positional, but valence approaches tend to dominate e. The number of parties is no longer as interesting in the literature as the level of __fractionalization__ and __volatility__ among parties. f. Policy dimensions are durable, high cost cleavages between social groups. These divisions are mapped (many-to-few) onto parties. The dimensions of competition are how the divides are mapped. (e.g. US democratic party might represent both anti-war and pro-choice) g. Question: Can the number of parties predict the number of social divides? Both ‘divides’ and ‘competitive dimensions’ need to be conceptualized better. Some challenges to measuring the relationship between divides and competitive dimensions i. To simplify political processes, politicians want to reduce divides an therefore map new issues to existing divides ii. Politicians and citizens can only process so much information about divides. iii. Changes in social structures create new divides h. Competitiveness in party systems i. Occurs when elections are uncertain and uncertainty means large gains/losses ii. Intense closeness between candidates occurs when: marginal gains in support mean great gains in power; there are more than two top candidates and they are in alternative parties; voters perceive the outcome as close; some voters float, i.e. don’t have clear candidates; high stakes – large gap between costs and benefits iii. Competitiveness interacts with fractionalization/volatility but cause is debatable 3. Competitive statics (Kitschelt’s attempt to move beyond conventional party system theory) a. Most theories just look at indirect voter-candidate relations and positional campaigns; strategies vary just according to number of competitors and dimensions of competition b. Downs’s media voter theorem: too many assumptions, and equilibria are not easy c. Other models relax spatial-positional assumptions, integrating valence: i. Adams-Merril-Grofman: voters have non-policy preferences and will vote for candidates they identify with; problem: “non-policy issues’ is huge ii. Schofield: parties disperse over programmatic issues only with enough non-policy valence, e.g. credible candidates; problem: valence as an independent variable is too flexible – there is always some level of valence iii. May and others: preference heterogeneity is an assumption so candidates rely on policy activists to shape voter preferences; activists gain some political power; problem: activists may have other interests than getting the politician elected iv. Roemer: politicians may be interested in their policies and not just winning office; means that intra-party faction building creates distinction among party competitors v. Voters may think long term and vote for more extreme parties expecting compromise between extreme and moderate parties. vi. Voters may choose parties that are directionally compatible, e.g. they are on the right side of the aisle; gives parties incentive to disperse positions d. Agent-based modeling: a backlash against formal theory and historical narrative of party competition; uses computer simulations i. Assumes voters and politicians have little knowledge; parties change positions slowly and gravitate towards center positions; ii. Advantage: easy to introduce changes to model; Challenge: do models produce results that are observable? What if many model set ups lead to same result? e. Explaining the entry of new parties not understood well by formal models i. Informal literature talks about demand, and new parties satisfy unmet demand ii. A salient-issue theory has new interests trying to break down existing competitors
 * SUMMARY**