Class Notes (in progress) for PS 441 (UM) and PS 155 (UCLA): Comparative Politics of Established Capitalist Democracies

Professor Robert (Rob) J. Franzese, Jr.

INTRODUCTION (1)
I. Intellectual History of Comparative Politics as a Field of Inquiry A. Traditional Study (through ca. 1950) 1. Configurative description 2. Parochialism 3. Formal Legalism 4. Some Empiricism, not very systematic, Absence of Theory & Methods 5. Non-comparative B. Continuing Methodological Problems & Debates 1. Case Study vs. Comparative Method 2. From Political Sociology to Political Institutionalism 3. Newer Debates a. Formal vs. Qualitative/Linguistic Theoretical Argumentation b. Quantitative vs. Qualitative Empirical Evaluation c. Positivism vs. Interpretivism C. Gabriel Almond & the Parsonian Revolution (mid 1950s) 1. Influence of Sociology 2. Introduces Central Question: What makes for (stable) (democratic) political development 3. Theory Socio-Political Structure 6 Homogenous Fragmented Political Role Structure Anglo-American Continental European Highly Differentiated Totalitarian Pre-Industrial (Traditional) Low Differentiation D. Seymour Martin Lipset, Political Man (1959) 1. Democracy implies a. Political formula/beliefs specifying which institutions (parties, free press, etc.) are legitimate b. 1 set of political leaders c. 1+ set(s) trying to replace them (legitimately) 2. Which systems exhibit these characteristics stably? 3. Democracy requires government & opposition, but which oppositions too much? Which cleavages too severe or irascible? 4. Factors facilitating stability a. Economic Development b. Class Structure c. 2-Party System d. Cross-Cutting Cleavages e. Federal, not Unitary, System f. Historical development of conflict resolution (?) Cleavage Structure 6 Cross-Cutting Reinforcing Class Stratification Intermediate Least Stable Highly Stratified Most Stable Intermediate Low Stratification E. Stein Rokkan: Cleavage Structure & Democratization Process (1967, 1970, 1983) 1. Four developmental processes a. Protestant Reformation b. National Revolution c. Industrial Revolution d. Communist Revolution CLEAVAGE TYPOLOGY National/Center Local/Periphery worker v. employer/owner primary v. secondary econ. Interests/Economy church v. state subject v. dominant culture Ideology/Culture F. Robert Dahl, Polyarchy (1971): The Conditions for & the Degree of Democracy POLITICAL SYSTEM DIMENSIONS Participation 6 High Low Contestation Polyarchies Competitive Oligarchies High Italy, Finland Switzerland, Ireland Inclusive Hegemonies Closed Hegemonies Low Hungary, Poland (pre-Dem.) Portugal, Spain (pre-Dem.) 1. Conditions for Democracy: A Long List a. Peaceful evolution of democracy c. Economic development b. Decentralized economy d. Economic equality Page 1 of 91

e. f. G.

Social homogeneity Elite pro-democratic beliefs

g. h.

Popular belief in democratic efficacy & sincere intentions of pol. adversaries Passive or supportive international conditions

Merkl, Almond & Verba: Political Culture 1. Peter Merkl, Modern Comparative Politics (1970) a. Common State of Democracy is Change, Instability, & Disequilibrium b. Culture (beliefs, attitudes) explain this best since they change observably (1) Descriptive at best (2) Tautological at worst 2. G. Almond & Sid Verba, The Civic Culture (1965) a. People have attitudes toward: (1) political system (2) input activities of citizens (3) output activities of government (4) themselves as political participants b. These configure into three archetypes of society Involvement in Pol. Sys. 6 High Low Knowledge of Pol&PolSys Participant Society Subject Society High [Not Considered] Parochial Society Low H. Gordon Smith, Politics & Society in Western Europe (1984) 1. Social Cohesion not directly or Necessarily linked to Political Cohesion 2. Governing Institutions structure the Incentives facing actors from Voters to Leaders 3. ===> Sociological Structure Works through Political Institutions to Determine Democratic Functioning I. Arend Lijphart: Consociationalism, Institutionalism, & Democracy 1. Social heterogeneity creates impetus to instability, but impetus modified or redirected by institutions 2. Original focus on institutional means of dividing power to diffuse disruptive conflict a. Minority representation b. Division of power: geographic, demographic, or functional 3. Then on elite behavior: Social Structure 6 Homogenous Plural (Heterogenous) Elite Behavior Depoliticized Democracy Consociational Democracy Coalescent Switzerland Austria Centripetal Democracy Centrifugal Democracy Adversarial United States Italy 4. Heterogeneity only a necessary condition for instability, absent consociation it becomes sufficient 5. In Democracies (1984), emphasis again on structure of decision-making process mediating between social cleavages & (in)stability ==> 6. Two ideal-types of democracy: a. The Westminster Model b. The Consensus Model J. Suzanne Berger, Gerhard Lehmbruch, Phillippe Schmitter: Corporatism & Interest Intermediation (1970s & 1980s) 1. Between the interests created by social structure & the polices implemented by governments are organizations which aggregate, mobilize, & press for policies: a. Parties b. Other actors... 2. Corporatist Policy-Making [DEFINE] credited with successful macroecon. management & thereby facilitation of democratic stability K. Giovanni Sartori: Parties & Party Systems (1976) 1. Parties do not simply reflect the underlying social structure ==> independent role of party systems for stability Party-System Polarization 6 Low High Party-System Fractionalization Bipartism Centripetal Competition Low United States United Kingdom Segmented Multipartism Polarized Multipartism High Ireland Italy 2. Also emphasizes key role of anti-system parties L. Comparative Party Research 1. Schumpeter (1942): defined party role 2. V.O. Key (1966): do parties set agenda or do people? 3. Pateman, Rose, Lawson: When Parties Fail? ===> 4. Strategies of Parties (Rokkan, Otto Kirchheimer): a. Mobilize a niche ==> b. Catch-all ===> 5. Probability of strategic success depends on strategies of other parties M. Origins of Modern Institutionalism 1. Douglas Rae, The Political Consequences of Electoral Laws (1971) a. Duvergers Law (DEFINE) & Beyond Page 2 of 91

N.

O.

Relations between voting rules and... (1) Voters incentives & actions (2) Policy-makers incentives & actions 2. Lawrence Dodd, Coalitions in Parliamentary Government (1976) a. Minimal Winning Coalitions (1) See, also, William Riker, The Theory of Political Coalitions (1962) (2) DEFINE b. Implications (1) MWCs are stable: Single-party majority or Multiparty (2) Emphasis on... (a) ...willingness to enter coalition and... (b) ...informational uncertainty in coalition bargaining... (3) ==> Re-emphasis on fractionalization & polarization c. Extensions: (1) Minority governments workings: Kaare Strom & George Tsebelis (2) Parliamentary behavior (a) Party discipline (b) Constituency 3. Modern Successors: a. Gary Cox, Making Votes Count (1997) b. Michael Laver & Ken Shepsle, Making & Breaking Governments (1995) Political Outputs & Outcomes (the present) 1. Traditionally little focus on what governments actually did (!) a. All about societal inputs & how political institutions produced governments from them, but b. Little or nothing about what policies were produced, i.e. So What? 2. Increasing focus on policies & outcomes produced by structure & institutions ===> a. Focus on Actors Opportunities, Objectives, & Constraints b. Empirical Analysis: attempt to link differences in policies & outcomes across time & countries to structure & institutions 3. The Keynesian Welfare State & Successor Policies become a central focus for modern comparative study a. Countries faced similar economic conditions & responded differently, Why? (1) Partisan differences? (2) Socio-economic-structure differences? (3) Institutional differences? b. Question was originally Does Politics Matter? Answer: yes; so new questions: How Does Politics Matter? G. Bingham Powell: Contemporary Democracies (1982) 1. One of the first systematic analyses of structure & institutions ==> Participation & Stability & Order 2. Fertile ground by then: rich in theory & assumption, empirical vacuum

b.

INTRODUCTION (2)
II. Start with a definition of the subject matter: What is an Established or Developed Capitalist Democracy? A. Leduc, Niemi, & Norris: Although elections & democracy are not synonymous concepts, the existence of free, competitive elections is invariably considered one of the critical features that define a nation as democratic (Inkeles 1991; Beetham 1994) (p. 4). B. Powell: 1. working political democracies[:] Citizens of these countries are able to organize & vote in competitive elections[, and] the national political leaders are held accountable to their citizens through electoral means (p. 1). 2. democracy[... :] political systems where representative leaders are chosen through competitive elections. The competitive electoral context, with several political parties organizing the alternatives that face voters, is the identifying property of the contemporary democratic process. ...competitive elections in which most citizens are eligible to participate (p. 3) 3. Case-selection criteria (p. 3): a. legitimacy of the government rests on a claim to represent the desires of its citizens [Monte Python... Supreme executive power...] b. the organized relationship that regulates this bargain of legitimacy is the competitive political election. Leaders are elected at regular intervals, & voters choose among alternative candidates. In practice 2+ parties with a chance of winning is minimal c. Most adults can participate in the electoral process, both as voters & as candidates for important political office d. Citizens votes are secret & not coerced e. Citizens & leaders enjoy basic freedom of speech, press, assembly, & organization 4. Although it is definitionally possible to have a democratic context in which power did not change hands (because of continuing support of the incumbents by a majority of citizens [e.g., Japan, Switzerland, Italy, Sweden for long periods]), such power changes seem a clear indication of importance of competitive elections (p. 6). C. Lane & Ersson: 1. West European Democracies[: ...] what sets these countries apart...is the legitimate operation of democratic political institutions within state structures that involve big government (p. 10). 2. A liberal democratic political order is combined w/ welfare state & mixed economy. The public sector plays a major role in economy of West European countries, allocating & redistributing resources by means of various institutions at different levels of government (p. 10) 3. open competition between political parties for the allegiance of the voter in free elections recruiting a legislative assembly that shares the exercise of power with government institutions in a national setting (p. 10). III. Notable Trends, Events, & Topics to Consider Page 3 of 91

A.

B.

Lane & Ersson: A Changing Scenario 1. 1950-65: End of Ideology, & Apathy (pp. 1-2) a. Attributed causes of perceived decline in conflict (1) Some saw a disappearance or reduction in ideological conflictharmony of views (2) Some saw very effective welfare society succeeding in providing security & thus reducing divisive consequences of socioeconomic cleavages b. The west was certainly thriving in this period (1) strong economic growth (2) stable law & order (3) legitimacy of established political institutions not questioned (a) either seen as evolutionary result on long process toward peaceful & prosperous democratic society (b) or as the result of recent (re-)establishment of same, hard-won by war 2. 1965-80: Rejection of Political Authority (pp. 2-6) a. Increased citizen activism in 60s turned scholars toward participation (1) demands for institutional autonomy in UK, Belgium, Spain (2) demands for increased decentralization in Sweden, France, Germany b. Affluence: (1) Some saw rising wealth as diminishing intensity of socio-economic cleavages relative to new post-materialist (Inglehart) issues. (2) Others noted that affluence itself can exacerbate economic cleavages through perception of relative deprivation (Stouffer, Gurr). c. Economic Hardship (Rising Inflation, then OPEC I (1973-4) & OPEC II (1979-80) & the responses thereto) & Increasing Dissatisfaction with Government (1) Corporatism: tri-partite bargaining over (particularly) macroeconomic management between organizations of labor, business, & government (Berger, Lehmbruch, Schmitter) (2) New & re-emerging parties & movements (Greens, radical left, radical right) & general protest parties d. Perceived Upshot (caveat not to exaggerate change or continuity): (1) changing voter alignments & behavior (2) changing party systems, patterns of support, & behavior (3) everything is now questioned (a) legitimacy of the governmentthough this is always relatively strong in developed capitalist democracies (b) efficiency of the Welfare State (c) social disorder becomes less unheard of if still not exactly common 3. 1980-90: Crisis of the Welfare State & Communism a. Slower growth, higher unemployment, more variant if not always higher inflation (1) certainty & optimism characterizing Keynesian macroecon. management gone (2) New Ideas: monetarism, supply-side economics, the Rational-Expectations revolution all (to varying degrees) reject policy effectiveness of macro mngmnt b. immigration & refugeesnew issue for many of these homogenous societies: consider Ireland, Italy, France, Germany c. Continued (for a time) growth of government, but economic slowdown produces huge deficit & debt issues in many countries ==> Crises of the Welfare State 4. 1990s: Integration, Unemployment, & Immigration a. The European Community (http://europa.eu.int/) (1) ECSCFrance, Germany, Italy, Belgium, Netherlands, Luxembourg (1950) (2) EC: Treaty of Rome 1957; EFTA 1960Austria, Denmark, Finland, Ireland, Norway, Portugal, Sweden, Switzerland, UK (3) Grows over time: (a) membership additions: i) UK, Ireland, & Denmark (1973) ==> 9 ii) Greece (1981) ==> 10 iii) Portugal & Spain (1986) ==> 12 iv) Austria, Finland & Sweden (1995) ==> 15 (b) scope: i) ECSC weak, loose, protectionist agreement on coal & steel quotas, standards, etc. tentative first step ii) Treaty of Rome: more-general trade-agreement, supposed internal free trade & common external trade policy, Common Agricultural Program iii) European Currency Union & the snake (exhange-rate convention) 1973 iv) VAT harmonization 1977 v) E.C.U. & the ERM: 1979 vi) The Delors Programme 1985 vii) Maastricht 1993 viii) Euro: Transition to a Single Currency: 1997-2002 b. High unemployment remains, slow growth, fiscal crisis; democracy stronger than ever, but its national locus more open to question LeDuc, Niemi, & Norris: 1. Three Waves of Democratization: late 19th to early 20th century; following WWII, & 1990s 2. Recent Democratizations a. Triple transformation in Eastern Europeparty systems, constitutional structures, & economic systems b. Collapse of last authoritarian regimes in West Europe: Greece, Portugal, & Spain c. End of Apartheid in South Africa d. growth and/or re-establishment of democracy in Latin America & Southeast Asia Page 4 of 91

N.b. these waves are not unidirectional ==> democracy is not secure or established once initially instituted Major Changes in Established Democracies also a. party systemstraditional patterns of conflict changing (Mair, ch. 3) b. attitudes & voting behavior of citizens (Dalton, ch. 13) c. rise of cynicism & legitimacy crises d. even changes in the foundations of representative institutions in some cases (New Zealand, Italy, Japan) e. Again, though, should not exaggerate these changes (1) Generally changes within the systemic confines of relatively liberal, relatively free-market democratic capitalism (2) And new & re-emerging democracies are not near established yet 5. N.b. these last changes (B.4.) + experience of E. Europe draw attention to electoral system as a manipulable, even engineer-able, variable C. Powell: 1. Electoral participation relatively consistent w/in nations over time, US not very exceptional in having seen much decline from late 60s-80s 2. Important incidents of breakdown of democratic order, even into violence: a. Northern Ireland (U.K.)violence flares up in late 60s (1) Catholic Minority in Mostly Protestant Ulster region of a Mostly Catholic island (2) Catholic minority demands greater political equality & economic improvementsdraws our attention to reinforcing cleavages b. U.S.race riots in late 60s through 70sagain draws our attention to disruptive potential of reinforcing cleavages (IV.D.2.c-e; Vietnam c. Various other occasionally violent problem areas (almost all have or have had their problems) (1) IrelandI.R.A. actions supporting Irish intervention in Ireland (2) Germanyfar-left Bader-Meinhof in late 60s, far right against immigrants in the 90s (3) Canadaperiodic separatist turmoil (4) Francestudent riots in late 60s (eventually commingled w/ other issues, esp. Algeria, bringing down regime), massive strikes (5) Spainseparatist struggles: Basque & Catalan (6) BelgiumFleming & Walloonlanguage, religion, & class IV. POWELL (Chapters 1 & 2) A. Working Political Democracies 1. Definition (I.B.1, +selection criteria, I.B.3) 2. Some Work Well a. Participate via elections b. Stable governments c. Discontent expressed through Democratic competition not violent conflict 3. Some Work Less Well a. Involvement characterized by turmoil not elected parties b. Government is unstable, unresponsive, or both c. Violence widespread--at worst the life expectancy of democracy itself is a matter of constant calculation. 4. Why? a. Social & economic environment? b. Political institutions & organizations? c. Beliefs & strategies of leaders/citizens? d. System of political parties? B. Social Structure ===> Institutions & Parties ===> Democratic Functioning C. Role of Parties 1. The link between social, economic, & constitutional setting & political performance a. Examining citizen partisanship, party strategies, & election outcomes (party factors) helps us understand relationship between economic development & voter participation or between electoral laws & government stability b. Not all effects of environmental conditions work through party system but many of them do 2. Independent Effects: The configurations of memory, organizations, & perception the system represents have direct effects once established D. Three focal Dimensions of Political Performance 1. Citizen Electoral Participation: (Voter Turnout = # voting/eligible population) a. Not only form of participation, but essential in forcing policy-makers to respond b. Symbolic--especially with increased secularization, legitimacy has come to rest on claim to respond to citizens preferences c. STOP HERE: What is the empirical puzzle? (1) Some Data (2) Why do people vote? (a) Generate some hypotheses (b) Tell the irrational to vote story (c) Generate more hypotheses 2. Government Stability & Effectiveness a. Definitions: (1) Stability: (Government Duration)--Some issues/problems in defining (2) Effectiveness: Def? (% of legislative support for executive) b. STOP HERE: Point out difference between definitions (theoretical, abstract) & operationalization (empirical, concrete) (1) Some data on duration (2) Generate some hypotheses (3) Some data on effectiveness?--or generate some alternative ideas on how to measure it c. Aside on party discipline 3. Political Order--Absence of turmoil & violence & maintenance of basic forms of democratic regime Page 5 of 91

3. 4.

V.

Distinguish observing order from attributing blame or credit for it; breakdown of order may reflect: (1) regime policy failures (2) bargaining intransigence by either party to another (3) efforts by either party to dispute to undermine democracy itself b. Democracies are unique in providing citizens (a.) resources, & (b.) freedom to use, (c.) expand it by organizing & mobilizing other citizens, & (d.) allowing them to use it to express discontent c. Democracy=a gamble that discontent can be channeled through legitimate (electoral & other) channels--an outbreak of serious, collective violence is a sign it is not working (whose fault is another question) d. Measures: Riots, Deaths from Political Violence, Suspension/Replacement of Regime (1) Some Data (2) Generate some hypotheses E. Relations between the Three Dimensions of Performance 1. Mutually reinforcing Hypothesis 2. Participation ==> (In)stability, & (In)effectiveness 3. Stability ==> Effectiveness & Order or ==> Corruption & Disorder 4. Powell finds a. Participation... (1) ... ==> Order (low violence, turmoil, & no regime )) (2) ... ==> Low government durability, b. otherwise the three generally unrelated Lane & Ersson, From Political Sociology to Political Institutionalism [This reviews some of the above] A. Theoretical Framework: We organize our [study]...in terms of cleavages, political institutions, & public policies (p. 11) B. Traditional Study of Politics: 5 Characteristics 1. Configurative description: just describe from A to Z everything political that exists or happens in some country 2. Parochialism: Western (and predominantly US) slant on identification of issues to be considered 3. Formal legalism: Excessive focus on the constitutional & legal details of a country 4. Absence of empiricism, methodology, & theory: all this description & legalism is conducted without any attempt to relate features & outcomes systematically 5. Non-comparative: despite the name, the field tended to focus on one country at a time (relates to previous problem) C. Two (continuing) Methodological Problems / Debates 1. Case-Study versus Comparative Methods 2. Political Sociology evolving into Political Institutionalism 3. To these, we should add a. Formal (i.e. mathematical) derivation of arguments vs. Qualitative (linguistic) derivation of arguments b. Quantitative (i.e. statistical) evaluation of evidence vs. Qualitative evaluation of evidence c. Positive theory as goal vs. Interpretation & Understanding as goal 4. These either/or divisions are somewhat arbitrary at least in that all involve at least some degree of the other; furthermore, several of these divisions are quite highly related; still, there remains some controversy around these divides D. A Little Intellectual History as an Introduction to the Subject 1. Gabriel Almond Comparative Political Systems (1956) & the Parsonian Revolution a. Talcott Parsons [Toward a General Theory of Action (with Shils, 1951), The Social System (1951), & Economy & Society (with Smelser, 1956)] brought Webers new systematic approach to the study of social behavior (Sociology) to the forefront of political science (and social science more generally) Homogenous Fragmented b. Almonds 2x2: POLITICAL CULTURE --------> POLITICAL ROLE STRUCTURE: Highly Differentiated Anglo-American Continental European Low Differentiation Totalitarian Pre-Industrial c. Almond argued that Anglo-American political systems were stable b/c there was high degree of consensus on political means & ends & because political roles of individuals & institutions were clearly differentiated & delineated. (Interesting as a theoretical conjecture; empirically, he neither defined nor measured either independent or dependent variables. Still, a huge advance over what preceded. 2. Seymour Martin Lipset (Political Man 1959): conditions necessary for democracy in societies & organizations (p. 9) (p. 19) a. Democracy implies... (1) ...a political formula or body of beliefs specifying which institutionspolitical parties, a free press, & so forthare legitimate (accepted as proper by all); (2) ...one set of political leaders in office; and (3) ...one or more sets of recognized leaders attempting to gain office (p. 45) b. Lipset is interested in explaining which political systems exhibit these conditions stably & why; i.e. stability of democracy not government stability within democracy c. Democracy implies government & opposition; thus democracy cannot exist without cleavages; question which cleavages are too much and/or which factors may serve to mitigate conflict arising out of cleavages d. Factors Lipset links to Stable Democracy: (1) Economic Development produces greater income, economic security, & widespread educationall considered conducive in that they form the terms of the class struggle permitting the less well-off to take a long-term [and optimistic] view (2) An appropriate class structurelarge middle class (3) An appropriate party system2-party rather than multi-party: 2-party competition fosters integrative rather than divisive politics (4) Cross-cutting cleavagesterritorial rather than proportional representation help in this regard (5) Other political system featuresfederal rather than unitary statesconcentration of power frequently a threat to democracy (6) A historical development of conflict resolution Page 6 of 91

a.

3.

4.

5.

6.

7.

Lipsets 2x2: Cleavage Structure --------> Cross-Cutting Mutually Reinforcing Class Stratification: Highly Stratified Intermediate Least Stable Low Stratification Most Stable Intermediate Stein Rokkan: Cleavage Structure & Democratization of National Societies (1967, 1970, 1983) a. Four fundamental processes in the democratic transformations of Western Democracies (1) Reformationdivided Catholic/Protestant (2) National revolutionbreaking local & regional affiliations (Germany & Italy mid to late 19th century (3) Industrial revolutionreplacing diffuse agrarian & religious loyalties with class lines (4) Communist revolutioninternational fault lines b. Rokkans 2x2: Cleavage Typology National/Center Local/Periphery Interests/Economy worker v. employer/owner primary v. secondary economy Ideology/Culture church v. state subject v. dominant culture Robert Dahl (Polyarchy 1971): The conditions for democracy a. the degree of democracy (polyarchy) is given along two dimensions: participation & contestation b. Dahls 2x2: Dimensions of Political Systems Participation High Low High Polyarchies (Democracies) Competitive Oligarchies Contestation Italy, Scandinavia Switzerland, Ireland Low Inclusive Hegemonies Closed Hegemonies Hungary, Poland (pre-Dem) Portugal, Spain (pre-Dem) c. The laundry list of conditions for democracy (1) peaceful evolution of democracy transferring legitimacy from old to new regime (2) a decentralized economy (avoiding concentration of economic power) (3) high economic development (4) economic equality (5) social homogeneity (6) elite beliefs that support democracy as legitimate means of resolving conflict (best when authority structures are the same in most institutions of society) (7) beliefs in the effectiveness of democracy & in the sincere objectives of even political adversaries (8) either passive or supportive international conditions Peter Merkl; Gabriel Almond & Sid Verba & Political Culture (Modern Comparative Politics 1970 & The Civic Culture 1965) a. Merkl: common state of affairs is instability or disequilibrium; contends that cultural (beliefs, attitudes) approach explain that better b/c they change observably (survey research). [Critics would note this hardly explains why they change => descriptive, tautological] b. Almond & Verba: a key to democracy is the manner in which citizens relate to the political system. (1) People have attitudes toward: (a) the political system (b) input activities of citizens (like voting) (c) output activities of government (policies & programs) (d) themselves as political participants (2) They identify three archetypes of society in terms of how citizens relate to politics in these four ways (a) Parochial Society: low knowledge of & involvement in political system (b) Subject Society: high knowledge of but low involvement in political system (c) Participant Society: high knowledge of & involvement in political system (3) Argued that peoples attitudes, (1), defined public orientation toward political system, (2), producing a certain sort of politics Gordon Smith (Politics in Western Europe 1984): Social & Political Cohesion: a. main point here was that social cohesion was not directly or necessarily linked to political (government) cohesion. b. The critique came at a time when social-structuralism had run its course & culturalism was losing steam c. ==> begins focus on governing institutions & how they structure incentives facing political actors from individual voter to president Lijphart & Consociationalism or Consensus Democracy: a. Social heterogeneity creates an impetus toward instability, but this impetus can be modified or redirected by political institutions adapted to an environment characterized by this potential for disruptive conflict (or not) b. Produces a focus on the capacity of political institutions & political leaders to respond to stimuli from society c. At first, focus was on institutional means of dividing power so as to diffuse disruptive conflict (1) Minority representation or other means of safeguard? (2) Division of power along geographic, demographic, or topical bases? d. Then focus increasingly on elite behavior, yielding: Lijpharts 2x2: Regime Typology Homogenous Society Plural Society Coalescent Elite Behavior Depoliticized Democracy (Switz.) Consociational Democracy (Austria) Adversarial Elite Behavior Centripetal Democracy (U.S.) Centrifugal Democracy (Italy) Lijphart: Only bottom-right unstable (heterogeneity a necessary condition for instability, absent consociation it becomes sufficient) e. Later, in Democracies (1984), the emphasis is again on the structure of the decision-making process mediating between social cleavages & (in)stability, He now derives two ideal-types of democracy: (1) The Westminster Model: (2) The Consensus Model (a) one party, bare majority governments (a) Executive power-sharing (b) fusion of legislative & executive power, & cabinet dominance (b) Separation of powers, formal & informal (c) asymmetric bicameralism (c) Balanced Bicameralism & minority representation (d) Two-party System (d) Multi-party system Page 7 of 91

e.

8.

9.

10.

11.

12. 13.

14.

(e) One-dimensional party-system (e) Multi-dimensional party system (f) plurality electoral system (f) Proportional representation (g) unitary & centralized territorial government (g) Territorial & non-territorial federalism, decentralization (h) unwritten constitution & parliamentary sovereignty (h) Written constitution & minority vetoes Lehmbruch, Berger, Schmitter: Corporatism & Interest Intermediation a. Interest Intermediation: between the interests created and arrayed by the social structure & the policies implemented by government lie organizations which aggregate & mobilize the former to press the latter for various policies b. Focus especially in early period on labor unions & their role in tripartite bargaining: government, employers (capitalists), & labor c. Corporatist decision-making mechanisms credited w/ successful regulation of macroeconomy & thereby facilitating dem. stability d. Recent innovations have begun to shift the focus toward the role of firms as employers & allocators of capital Giovanni Sartori: Parties & Party Systems (1976) a. Parties & party systems do not simply reflect the cleavage structure of society b. Sartoris 2x2: Party System Low Polarization High Polarization Low Fractionalization Two-partism (U.S.) Centripetal Competition (U.K.) High Fractionalization Segmented Multipartism (Ire., Switz) Polarized Multipartism (Italy) Again the bottom-right are the most unstable. c. Key role of anti-system parties Comparative Party Research a. V.O. Key (1966): do parties set agenda for voters or do voters decide between parties? b. Schumpeter (1942): defined the role of parties (1) interest articulation, aggregation, policy-making, & system legitimation (2) parties present voters w/ choice b/w ideologies represented in manifestos & implemented by govts. in office; voters evaluate c. Pateman, Rose, Lawsonparties increasingly failed in that: Lawsons When Parties Fail definitive statement of why stability ended d. Strategies of Parties (Rokkan, Kirchheimer): mobilize a niche or manufacture a catch-all coalition (1) probability of success for a party strategy depends on the social structure, institutional rules, & strategies of other parties (2) highlights strategic behavior of political actors; but also tends to explain change by actor mistakes & stability by actor wisdom e. New directions: why parties at all? Post-materialist values stress independence & critical of system==>crisis for parties Dodd & Coalitions in Parliamentary Government (1976) a. Argues that multiparty parliaments==>coalition governments; coalition governments unstable; so multiparty systems unstable=false b. Introduces notion of Minimal-winning coalitions; argues that MWCs are stable (1) Single-party governments are usually MWCs, so theyre usually stable (2) Multi-party governments are stable too if theyre MWCs c. Formation of MWCs is a function of parties willingness to enter coalitions & informational uncertainty in bargaining with potential coalition members d. Minority governments: how they work & the conditions under which they work well (Strom, Tsebelis) e. Parliamentary behavior: party discipline a key factor related to electoral system & government structure Budge (1983) & Issues: relates party strategies & soc. struct.: elites choose issues over which politys social structure will divide its citizens Political Outputs & Outcomes: a. traditionally very little said about what governments actually did (!)all about societal inputs & political institutions structuring how those inputs manifested themselves in governments, but what about policies? [corporatism exceptional in this regard] (1) Opportunity, objectives, & constraints (2) Finally some discussion of topics which concretely affect citizens lives in democracy (3) An attempt to link variation in policies across countries & over time to differences in state structure & the like b. The Keynesian Welfare State becomes an important subject for study (1) countries faced similar economic problems but responded differentlywhy? (a) partisan differences? (b) socio-economic structural differences? (c) institutional differences? (2) countries differ in their redistributional efforts, tax burdens, spending allocations & amounts, etc. Same set of why questions... c. The question was originally: does politics matter? That has become irrefutable; the question now is: how does politics matter? G. Bingham Powell & Democratic Performance: a. One of 1st systematic & empirical studies seeking to test theories about causes & consequences of Stability, Participation, & Violence b. The ground was fertile by this point: many theories, little systematic evidence

INTRODUCTION (3)
VI. Claus Offe: Competitive Party Democracy & the Keynesian Welfare State, in Contradictions of the Welfare State MIT Press 1984. A. Why were classical democratic & economic theorists so concerned that capitalism & democracy could not mix? (p. 179) 1. Mill: the egalitarian threats of mass society & democratic mass politics ... would necessarily lead to tyranny & class legislation by the propertyless, uneducated majority 2. Marx: [Democracy is...] a political form that would exacerbate social contradictions by withdrawing political guarantees from the socially dominant & giving political power to the subordinate 3. De Tocqueville, Hamilton, etc.virtually anyone who considered the matter 4. The 20th century suggests strongly that their concerns were unwarranted or exaggerated. What were they so worried about? B. A Simple Model: 1. Let the economy be described by: Page 8 of 91

C.

D.

E.

F.

yi = yi(t) where yi is the income/output of the ith person and t is the tax rate levied on each individual 2. Suppose the tax rate, t, is levied on everyone & the total proceeds are redistributed evenly, then the income net of tax & redistribution for individual i is: yi - tyi + tA3yj/n 3. Thus, anyone for whom tA3yj/n-tyi>0 receives more transfers than they pay taxes & therefore should favor redistribution. a. But this is anyone earning less than the average income in the society. b. So, since income is distributed rightward skewed (more poor, fewer but very rich), redistribution always has a majority. 4. How much redistribution? Considering that Myi(t)/Mt<0, we can solve... a. Result: provided voters know that increasing redistribution lowers societys total output, at least beyond some point, then the median voter desired amount of redistribution increases in the difference between median & mean income. 5. The classicals thought it was worse: they doubted the masses would understand that redistribution lowered total income. In any event, the concern certainly seems well-founded logically, so why havent the fears been realized? How is it that capitalism & democracy not only can co-exist, but seem very much to go together (not to assign a causal direction)? 1. The Leninist Tradition, Bourgeois Democracy: There is no tension because the state, democratic or otherwise, serves interests of capital a. On the one hand, we have the privileged position of capital b. On the other, we have that bourgeois democracy serves primarily to deceive the masses 2. The pluralist-elitist democratic theorists a. Democracy subordinates all social arrangements to mass approval: if the masses wanted something different they could, via their control of the state & its instruments of control, establish it. b. The fact that they dont implies that capitalism is actually desired. 3. Offe & Other Theorists of the Class Compromise, Postwar Settlement, or Keynesian Welfare State a. continued compatibility of capitalism & democracy...emerged historically due to appearance & gradual development of 2 mediating principles: mass political parties & party competition, & the Keynesian Welfare State (KWS) (p. 182) b. These institutions mix the logic of authority & the logic of the market: mixed economy & oligopolistic competition in politics infuses some of the logic of authority into the market & some of the logic of the market into c. ==> state not Rousseauian (absolute, patrimonial, pedagogic), market not Smithean (absolute, impersonal, devoid of power relations) The role of competitive party democracy in facilitating joint capitalism-democracy 1. Weber: the transformation of class politics into competitive party politics is not only a change of form but of content. a. Organizations such as the trade unions, but also the Social Democratic Party, constitute a very important counter-balance against the direct & irrational mob rule typical of purely plebiscitary peoples (p. 184) b. Bureaucratized parties with their charismatic & demagogic leaders would reliably bulwark against aimless mass rage 2. Luxemburg: the division of labor within working-class organizations between leaders & mass constituencies deradicalizes their political impact because leaders are professionals, have a stake in the system, etc. 3. Michels: organizational necessity translates mass radicalism into oligarchic, almost pro-system, pragmatism at the top 4. Offes summary: the very dynamic of...the organizational form contains, perverts & obstructs class interest & class politics in ways...leading to opportunism (Luxemburg), oligarchy (Michels), & the inescapable submission of the masses to the irrational impulses of the charismatic leader & his demagogic use of the bureaucratic party machine (Weber) (p. 184) Offe: Three Major Effects of the will of people being transformed through competitive parties striving for office 1. Deradicalization of the ideology of the party (both/all sides) a. It must maximize votes b. It must be prepared to govern with other parties (coalition, divided government) 2. A bureaucratized & centralized party organization a. The modern party must: (1) Collect material & human resources to wage electoral battle (2) Disseminate propaganda & information on the party positions, issues, candidates, etc. (3) Monitor & explore the political market (4) Manage internal conflict b. The party leaders develop a stake in maintaining & expanding the organization itself c. This has two consequences (1) Socio-economic composition of leadership increasingly diverges from that of membership &, more so, its electoral constituency (2) De-activation of the rank-and-file 3. Especially in party systems which favor catch-all parties (Kirchheimer), party competition produces increasingly heterogenous socioeconomic composition of a partys supporters a. Dissolves collective identity b. More ideological deradicalization 4. All three effects can be seen to contribute to functionality of democratic capitalism: a. Provides a virtual guarantee that the structure of political power will not deviate far enough from the structure of socio-economic power so as to make the distribution of each type of power incompatible with the other (p. 187) b. Fact: no competitive party system has ever yielded a distribution of political power which would alter logic of capitalism Offe concludes competitive party system is integral to democracy-capitalism coexistence, and moves on to its problems as he sees them. [As in all neomarxist theory, theres going to be some inherent tensions somewhere & impending doom, here its the...] Declining Party System 1. The party system... [by which he means: Territorial representation, Party competition, Parliamentary representation] 2. ...increasingly bypassed by other practices & procedures of political participation & representation (p. 188) a. New social movements [as a neomarxist, any deviation from pure class-based politics is a decline in some way] b. Corporatism: interest intermediation via socio-political organizations which are not electoral competitors c. Repression [he perceived a trend reining in democracy, not clear empirically founded then & certainly not now] 3. The problems for democracy & capitalism produced by these three trends [all of which are contestable Id say]: Page 9 of 91

Social Movements (1) often define membership by ascriptive traits (2) often seek autonomy not representation by which market & democratic status can be improved or protected (3) often political & esp parliamentary representation considered unnecessary because whats sought is not govt. action but inaction b. Corporatism (1) functional, not territorial representation (2) representation is fused with implementation & thus corrupted c. Authoritarianismobvious implications if true 4. The decline of the party system is likely to lead to the rise of less constrained & regulated practices of political participation & conflict, whose outcomes may then have the potential of effectively challenging & transcending the institutional premises of capitalist social & economic organization (p. 191). a. [Possibly it happened the opposite way? Challenges of capitalist & social organization led to increasingly less constrained practices of participation & conflict?] b. [Either way, another large theoretical & empirical project has been outlined by this line of inquiry] G. The resolution of the tension between the democratic principle of equal mass participation & the economic principle of unequal & private decision-making power has also rested heavily on the institutionalization of the KWS 1. The effects, rightly or wrongly, attributed to the KWS are broadly: a. An unprecedented & extended economic boom favoring all advanced industrial democracies b. Transformation of the radical or even revolutionary class conflict into peaceful, institutionalized, moderate politics 2. The Class Compromise [implicit foundations of postwar class peace]: a. Labor accepts the logic of profitability & markets as the guiding principles of resource allocation, distribution, & investment (growth) in return for... b. Capital accepts guarantee of minimal living standards, trade-union & liberal democratic rights, avoidance of mass unemployment (sharing of adjustment costs) & allowing real incomes to rise roughly in line with labor productivity 3. The Political Lessons of Keynesianism: a. Economics is a positive-sum game (no fixed-pie fallacy), therefore b. Each class must take the interests of the other into consideration (1) workers must acknowledge importance of profitability because their employment security & real-income growth depend on it (2) capitalists must accept the need for wages & welfare-state expenditures because these secure a healthy, well-trained, happy workforce, along with social peace & economic stability on which profits also depend 4. The Purpose & Role of the Keynesian Welfare State a. The primary role is to ameliorate the impact of economic vicissitudes of capitalism upon those least able to bear them; this has ancillary benefits for capitalists as well (1) one contention is that without the KWS, there would be a much higher level of industrial conflict (2) another is that this conflict would have both economic & political costs b. Crucial Functions (to be) Performed by the KWS (1) remove the most basic needs of the working class from the arena of class struggle & industrial conflict (2) provide the means for working class to pursue their interests collectively & therefore at least moderately effectively peacefully (3) making production more regular & predictable by removing these sources of disruptive conflict, and (4) foster this stability plus possibly to further growth potential by providing built-in stabilizers & public goods (5) [The last paragraph on 195 to first on 196 provides a summary of the functioning ideal of the KWS.] 5. The Inherent Tensions in the KWS are Manifested in its Unintended Consequencesa trade-off a. [The core ideas are offered on pages 198-99.] VII. Robert J. Franzese, Jr., Introduction, The Political Economy of Over-Commitment: A Comparative Study of Democratic Management of the Keynesian Welfare State, Ph.D. Dissertation, 1996.

a.

Page 10 of 91

Consolidated Central Government Revenue as a Fraction of GDP

0.6
Consolidated Central Government Revenue

in Developed Democracies in the Postwar Era

Shaded bars separate countries; each bar runs from 1950-1995.

NE BE NO SW DE NZ IR FI GR AL CA SP PO

0.5
as a Fraction of GDP
FR

0.4
GE

UK IT JA

AU

0.3 0.2 0.1 0.0

US

SZ

All data are from the IMF International Financial Statistics, various issues and computer tapes.

Figure I.1

Consolidated Central Government Expenditure as a Fraction of GDP

0.6
Consolidated Central Government Expenditure as a Fraction of GDP

in Developed Democracies in the Postwar Era

Shaded bars separate countries; each bar runs from 1950-1995.

BE

NE IR DE GR FI NO PO

0.5 0.4 0.3

US JA

SW NZ

FR IT GE UK AU

SP CA

AL

0.2 0.1 0.0

All data are from the IMF International Financial Statistics, various issues and computer tapes.

SZ

Figure I.2

Page 11 of 91

Consolidated Central Government Revenue Plus Expenditure

The Overall Size of the Public Economy

1.2 1.0
FR

in Developed Democracies in the Postwar Era

Shaded bars separate countries; each bar runs from 1950-1995.

NE BE IR NO PO SW NZ

as a Fraction of GDP

0.8
GE IT

UK

DE AU FI GR

0.6
US JA CA

SP

AL

0.4 0.2 0.0

All data are from the IMF International Financial Statistics, various issues and computer tapes.

SZ

Figure I.3

Transfer Payments as a Fraction of GDP

0.30
Transfer Payments as a Fraction of GDP

in Developed Democracies in the Postwar Era

Shaded bars separate countries; each bar runs from 1950-1995.

NE BE FI SW NO GR IR SP SZ

0.25 0.20
GE

FR AU IT CA UK JA PO AL DE

0.15
US

0.10 0.05
NZ

0.00
All data are from OECD National Accounts Volume II: Detailed Tables, various issues and data diskettes. "Transfer Payments" are the sum of items 30-32 on Table 6: Accounts for General Government.

No Data Available

Figure I.4

Page 12 of 91

Gross Public Debt of Consolidated Central Government

120
Gross Public Debt as a Percent of GDP
Shading Separates Countries; Data are from 1950 to 1990

As a Percent of GDP

100 80 60
US JA

BE IT UK

IR AL NZ NE SW

CA FR GE

AU

40 20 0

DE GR FI

NO

SP

SZ

All data are from IMF International Financial Statistics, various issues and computer tapes, augmented by data from OECD sources and from Roubini and Sachs (1989a, 1989b).
Figure I.5

Government Employment-Share in Developed Democracies

0.35 Government Employment As a Share of Total Employment DE 0.30 FR UK CA AU BE 0.20 US GE 0.15 JA 0.10 GR IT IR NE PO SP AL NZ SZ NO SW

0.25

FI

0.05
Shading separates countries; each shaded bar runs from 1950-1998.

0.00

Page 13 of 91

Unemployment in Developed Democracies in the Postwar Era

25
Shaded bars separate countries; each bar runs from 1950-1995.

SP

20
Unemployment Rate (%)
IR

15
UK FR BE CA DE IT FI AU GR NO JA SW SZ PO NE AL NZ

10

US GE

0
The data are the internationally comparable unemployment figures compiled by Layard, Nickell, and Jackman (1991) from OECD sources.

Figure I.7

R e a l - p e r - C a p i t a - G D P G r o w t h - R a t e s in t h e P o s t w a r E r a
15

Real per Capita GDP Growth Rate (%)

10

-5

-10 1951 1953 1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987

Year
C irc le m a rks t h e m e a n ; t h e b o x o u t lin e s a + / -1 s t a n d a rd d e v ia t io n r a n g e a ro u n d t h e m e a n ; t h e lin e s e xt e n d t o t h e e xt r e m a . T h e d a t a a r e f r o m t h e S u m m e rs a n d H e s t o n P e n n W o rld T a b le s M a c h V.

Figure I.8

Page 14 of 91

Inflation in Developed Democracies in the Postwar Era

30
Shading Separates Countries; Data are from 1950 to 1990

Consumer Price Inflation Rate (%)

25 20 15 10 5 0 -5
US GE JA IT

UK FI

GR IR

PO SP NZ AL

CA FR AU BE DE NE

NO

SW SZ

All data are from IMF International Financial Statistics, various issues and computer tapes.

Figure I.9

Result of Outcomes at T-1

The Cycle of Political Econom y

Exam p les of the Elements at Each Stage:
(A) Interests:
Sectoral Structure of Economy Income Distribution Age Distribution Trade Openness

(A) Interest Structure of the Polity and Econom y

Result of Outcomes at T0

On to T+1
Ele ctio ns

Elections:
Electoral Law Voter Participation

Action at Time T0
Ac tor s

Government Formation:
Fractionalization Polarization

(B) Partisan Representation in Government

t en rnm n ve tio Go rma Fo

Policy:
Fiscal Policy Monetary Policy Institutional Adjustment

Government Termination:
Replacement Risk

(C) Outcomes:
Unemployment Inflation Growth Sectoral Shift Debt Institutional Change

(C) Political and Econom ic Outcomes

ent ernm ) Gov ination ( m Ter

Exogenous Factors

Figure I.10

Page 15 of 91

Pol icy

Partisanship

No n-G ov ern me nta l

(B) Representation:

SOCIO-ECONOMIC STRUCTURE & POLITICS (Part 1) (L&E ch. 2; P ch. 3; GLM, ch. 9)
Introduction: Politics is the means by which conflicts (of opinion, of interests, of views) are reconciled (not necessarily resolved) so that policies may be decided upon & implemented. In liberal-democratic countries the goal is that this be managed in as orderly, efficient, & peaceful a manner as possible while remaining democratic. Analysis of socio-economic cleavage structures, then, is our inroad into understanding the fundamental, enduring, & potentially prominent schisms in society across which such conflict arises. VIII. Our analysis of socio-economic cleavage structure has four goals Definition of concepts Introduction of empirical facts, two parts 1. variation & similarity in broad, cross-sectional features within & across countries 2. change & continuity over time C. Arguments relating socio-economic cleavage-structure to outcomes 1. Party Systems 2. Democratic Performanceparticipation, stability, & violence IX. Definition of a cleavage A. L&E: a division on the basis of some criteria of individuals, groups, or organizations [b/w] whom conflict may arise (p. 53) [emphases added] B. Rae & Taylor: Cleavages are the criteria which divide members of a community or subcommunity into groups... (p. 53). A typology: 1. Ascriptive or traitrace, caste, ethnicity, language 2. Attitudinal or opinionideology, preference, class, religion 3. Behavioral or actthose elicited through voting & organizational membership C. [Id say R&Ts first two are sufficient; lets call them: 1. Ascriptive Cleavages 2. Preference/Value Cleavages] D. Characteristics by which cleavages differ from each other: 1. criteria by which membership on a side is determined: objectivity v. subjectivity, e.g. 2. exclusivity of the groups so divided 3. nature of the disagreement: policy differences v. general belief & value differences 4. salience 5. latent v. manifest E. If an abstract typology is difficult, perhaps more concrete enumeration will suffice (Daalder): 1. class or sectoral interests: worker/employer; traded/non-traded; private/public 2. religion: Catholic/Protestant; Christian/Non-Christian; religious/secular; fundamentalist/moderate 3. geographical conflict: urban/rural, center/periphery 4. nationality or nationalism: ethnic, nationalist, & regionalist separatism F. Multifarious cleavages, basically infinite in principle, the question of which cleavages matter is then best seen as entirely a practical one of explanatory power v. parsimony. Lane & Erssons choices [paraphrased] are a reasonable set: 1. Religious 2. Ethnic/Linguistic 3. Class/Economic 4. Regional/Geographic G. Some other definitions: 1. Powell: A set of attitudes that divides the nations citizens into major political groups. a. Sometimes... purely political: they reflect opinions about [particular] policies [or policy areas] & are developed, sustained, & organized by political leaders committed to such policies... b. more frequently, alignment of large groups is based on deep social, economic, & cultural divisions in the fabric of society...[e.g.] occupation, religion, language, race, custom, & geography... (42) c. focuses on two: (1) ethnic (cultural) divisions (2) economic divisions (3) one question has always been the relative weight of these two (a) [Pink Floyd: With/Without/& wholl deny/its what the fightings all about from Us & Them off Dark Side... ] (b) but avoid any false exclusivity: Both ethnic/cultural & economic always matter though relative weight may vary (c) [Any hypotheses as to why & how the relative weight might vary?] 2. GLM: implies much more than mere division, more even than outright conflict, b/w 2 sets of people... a. fundamental cleavage...has 3 quite specific connotations [GLM are reviewing Lipset&Rokkan here] (1) involves a social division that separates people who can be distinguished from one another in terms of key social characteristics such as occupation, status, religion, or ethnicity...Cannot be defined at the political level alone (210) (2) the groups involved in the division must be conscious of their collective identity...& be willing to act on this basis (3) must be expressed in organizational terms...usually achieved as a result of the activities of a trade union, church, political party, or some other organization that gives formal institutional expression to the interests of those on one side of the division (p. 211) b. following Lipset & Rokkan, focus is on four: (1) Center-Periphery (2) Church-State (3) Rural-Urban (4) Class Structure A. B. Page 16 of 91

Some Key Concepts: cleavage; cleavage structure; salience/intensity; latent v. manifest; cross-cutting v. reinforcing; cleavages in structure v. in consciousness; modernization/social-mobilization; fractionalization & polarization; ethnic, national, party (etc.) identification or consciousness; class, religious, ethnic (etc) voting; post-industrialism; post-materialism; electoral volatility XI. Modernization Hypotheses (Moving on to Powell) A. Key features of development: 1. Growth of affluence 2. (Economic) sectoral-structure change 3. urbanization 4. education, literacy, mass media 5. greater social & occupational mobility B. Classical Hypotheses: Countries with modernized social structures & developed economies & democracies are expected to have higher participation, more stable & effective governments, & less political violence than countries with more traditional social structures & underdeveloped economies & democracies. Why? 1. Modern Person hypothesis (Lerner, Inkeles) & value harmonization 2. ratio of economic development to socio-political mobilization (Huntington) 3. level of economic development at time of introducing political competition 4. pace-of-development hypothesis 5. the affluence hypothesis 6. the government-resources hypothesis 7. the increasing-interdependence hypothesis 8. education, [etc.] C. Powell: Impact on political performance? [generate some hypotheses] 1. impact on Participation a. greater information, education, & psychological involvement of modern person b. more extensive & specialized group structures, more effective organization (esp. parties & related group organizations) 2. impact on Government Stability/Effectiveness: seems logical that more developed democracies would be more stable & effective, empirically does not seem to be the case. Why? Is something missing? 3. impact on Social Upheaval/Violence a. see B above b. the relationship between development & social upheaval (riots) seems different than that between development & political violence (1) former curvilinear with early stages of development appearing worst (2) latter relatively linear, with political violence generally declining in development (3) are the patterns different among democracies? [discuss] XII. Ethnically/Linguistically/Religiously Divided societies A. problem of which divides are relevant B. problem of determining whether E/L/R source of conflict or economic differences which coincide therewith are at the root C. supposedly difficult to manage because not compromisable or bargainable issues D. Expected relation with... 1. participation = unclear, could go either way 2. government stability & effectiveness = negative 3. social upheaval & violence = positive E. Powell finds... 1. negative relationship with electoral participation 2. strong negative relationship with stability & effectiveness 3. complicated relationship with upheaval & violence a. no greater tendency to riot apparent b. however, given social upheaval, appears more likely to become violent & involve deaththe latter frequently occurring as governments attempt to restore order XIII. Economically Divided Societies A. Back to Aristotle: economic inequality produces political instability & a large middle class was the key to developing out of this 1. little doubt that occupational divisions & the inequalities that usually accompany them are a convenient basis for constructing political organizations (p.47) 2. still less question, Id say, that theyre a natural basis for potentially disruptive conflict B. Whats less clear is the mechanism by which economic inequality produces political instability 1. how are such inequalities so frequently tolerated in traditional societies & become issues in modernizing & modern ones? 2. n.b. its usually the abrogation of traditional peasant rights which triggers revolutions, doesnt appear to be the awareness of inequality per se 3. [concepts: relative deprivation, perceived v. objective inequality, justice] C. Expectations: Small population, high development, ethnic etc. homogeneity, & low inequality reduce riots & deaths D. Findings: Yes to population & development, some evidence for ethnic homogeneity, weak evidence for economic inequality. Why do you suppose? XIV. N.b., Powells findings are based on a sample of Democracies, we are interested particularly in Developed Democracies ==> we should reconsider some of his results in this smaller sample [Pretty Pictures Time] XV. Powell does a very nice job of summarizing the chapterread last 3 paragraphs carefully XVI. Topical Review & Summary of Material from L&E ch. 2, Powell ch. 3, & GLM ch. 9 A. Reinforcing (coinciding) Cleavages & Cross-cutting (not coinciding) Cleavages 1. Definitions: Page 17 of 91

X.

B.

C.

XVII.

a. Reinforcing: two cleavages are reinforcing to the degree that they divide society into the same two groups of people b. Cross-cutting: two cleavages are cross-cutting to the degree that their divisions of society do not coincide 2. N.b. this is a property of the relation between one cleavage (dividing line) & another 3. [Examples & So What?] Pattern of Cleavages & the Number of Parties 1. The cleavage structure of a polity could be described in various dimensions a. fractionalization of society (# of sub-groups, possibly overlapping) b. polarization of society (distance between groups) c. degree to which cleavages reinforce or cross-cut d. relative salience of the various cleavages 2. These factors combine to suggest the number of political parties which may evolve to represent various issues a. these features tell us the number of relevant disjoint subgroups b. assuming political parties evolve to represent groups of people with distinct interests, we should expect that to be at least suggestive of the number of parties which evolve 3. Social-structural data which would enable us to operationalize the abstract concepts in IX.B.1.a-d & process them into the number of disjoint subgroups has not been attempted & would be no simple task (though I think it might be feasible). However, subjectively, my impression is that the relationship is not as strong as might be expected: a. USmany disjoint subgroups, effectively 2 parties b. Italyrelatively homogenous, many parties c. Denmarkrelatively homogenous, many parties d. [Any guesses why this might be?] DiscussionEthnic/Religious/Linguistic/Cultural Divisions & Functioning Democracy: Why do you suppose polities, societies, & economies in some country-times appear able to function smoothly despite them & others not? 1. History? a. History of oppression & repression & internal & external war (1) US (2) UK/Ireland (3) Israel, former Yugoslavia (4) France, Germany, France & Germany b. History of homogeneity or heterogeneity (1) Germany v. Switzerland (2) Italy v. Canada 2. Reinforcing cleavages? a. US v. Switzerland b. Finland v. Spain 3. Relative size of the factions? a. Too small to be a threat b. So large it must be dealt with c. Hypothesis: The 15-40% range have most severe problems? 4. Institutional structure? a. Regional autonomy (Switzerland, Canada) b. Parliamentary representation (Netherlands, Belgium) c. Functional representation (Austria, Netherlands) d. Checks & balances (US)] SOME DATA & STATISTICS:
ethind .77 relind

AL AU

BE

DE FR GR IR JA NE NZ SP SZ US CA FI GE IC IT LU NO PO SW UK

Page 18 of 91

91.3
IC DE IT NO PO GR

AU GE

SW NZ

AL

BE NE LU

SP FI

vpart

IR JA FR

UK

CA

US

48.2 .01 ethind 91.3

IC DE LU NO SP PO GR FI IR JA FR BE IT AU NZSW

SZ

.77

AL GE

NE

vpart

UK

CA

US

48.2 .04 relind 91.3

IC LU NO PO GR FI IR CA SP AU SW AL GE DE BE NE IT

SZ

.61

NZ

vpart

UK FR JA

US

48.2 5.4848

SZ

12.3833 lpop

Page 19 of 91

91.3

NZ

AU AL

SW GE

IC IT NE

BE

DE LU NO

PO GR

SP FI

vpart

IR

UK JA FR

CA

US

48.2 8.2239 lrgdpc

91.3
AL SW NZ AU GE BE IT NO SP GR FI IR CA JA DE NE

SZ

9.44328

PO

vpart

UK FR

US

48.2 .32675

SZ

.45 gini

91.3

AU NZ GE

AL SW

BE IT NE NO SP GR FI IR CA FR JA

DE

PO

vpart

UK

US

48.2 48.4375

SZ

108.6 edsec

Page 20 of 91

91.3

AL NZ IC NE IT LU BE

AU GE DE NO PO SP GR FI IR

SW

vpart

CA JA

UK FR

US

48.2 10.1

SZ

17.2 age65o

91.3
BE

AL

SW NZ

AU GE IC DE NE IT NO

LU

SP GR FI IR FR

PO

vpart

CA

UK

JA

US

48.2 38

SZ

99 ethdom

Number of obs = 21 R-squared = 0.8833 -----------------------------------------------------------------------------vpart | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lpop | -3.336463 1.212639 -2.751 0.019 -6.005463 -.6674632 lrgdpc | 1.848647 6.740736 0.274 0.789 -12.98761 16.68491 ethind | .0623695 7.192933 0.009 0.993 -15.76917 15.89391 relind | 12.3443 7.94188 1.554 0.148 -5.13566 29.82426 gini | -111.8817 44.40408 -2.520 0.028 -209.6144 -14.14894 edsec | -.2808857 .1271328 -2.209 0.049 -.5607032 -.0010682 age65o | 1.203603 .7275164 1.654 0.126 -.3976493 2.804856 US | -17.27092 6.455974 -2.675 0.022 -31.48042 -3.061414 SZ | -46.59973 6.659673 -6.997 0.000 -61.25757 -31.94189 _cons | 144.9013 57.75213 2.509 0.029 17.78975 272.0129 ------------------------------------------------------------------------------

Page 21 of 91

60

LU

UK

SZ US

dgov80

IC NE NZ GR AL IR NO DE PO JA GE AU FI SW

SP

CA

BE FR IT

12.3 .01

.77 ethind

60

LU

UK

SZ

US

dgov80

SP IC NZ GR IR NO PO BE FR IT FI DE AU JA SW

CA NE

AL GE

12.3 .04

.61 relind

60

LU

SZ

UK

US

dgov80

IC NE NZ GR AL IR FI NO DE SW AU PO BE

CA

SP

GE

JA FR IT

12.3 5.4848 lpop

12.3833

Page 22 of 91

60

LU

UK

SZ

US

dgov80

SP NZ GR

IC NE

CA

AL IR AU PO FI SW DE JA BE FR IT GE NO

12.3 8.2239 lrgdpc 48

SZ US

9.44328

UK

dgov80

CA NE NZ

SP

GR AL GE IR SW NO DE JA BE FR PO AU FI

12.3 .32675

IT

.45 gini

48

SZ

UK

US

dgov80

SP

CA NE

NZ GR AL GE IR AU SW NO FI DE PO BE FR JA

12.3 48.4375

IT

108.6 edsec

Page 23 of 91

60

LU

US

SZ

UK

dgov80

IC

CA

SP NE GR

NZ AL IR FI JA PO BE FR IT AU DE GE NO SW

12.3 10.1

17.2 age65o

60

LU

US SZ

UK

dgov80

CA

SP NE NZ GR AL GE SW AU FI IR

IC

NO DE JA PO

BE FR IT

12.3 38 ethdom

99

Number of obs = 21 R-squared = 0.6162 -----------------------------------------------------------------------------dgov80 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lpop | -1.208241 1.905755 -0.634 0.538 -5.360524 2.944042 lrgdpc | -21.06676 11.06763 -1.903 0.081 -45.18106 3.047533 ethind | 24.58645 10.87294 2.261 0.043 .8963491 48.27655 relind | 28.05811 12.91752 2.172 0.051 -.0867556 56.20297 gini | -48.43518 70.82302 -0.684 0.507 -202.7453 105.8749 edsec | .220791 .2004675 1.101 0.292 -.2159903 .6575722 age65o | 1.553841 1.179984 1.317 0.212 -1.017124 4.124805 US | 18.08332 10.5959 1.707 0.114 -5.003161 41.1698 _cons | 192.649 94.03041 2.049 0.063 -12.2256 397.5237 ------------------------------------------------------------------------------

Page 24 of 91

3.59

UK

lattopp

FR GE SP

GR IT IR AU AL NO PO

SZ

BE

US

-2.14849 .01

NE

CA

.77 ethind

3.59

UK

lattopp

FR GE SP

GR IR BE AU

IT

SZ

US AL

NO PO

-2.14849 .04 relind 3.59

UK

CA

NE

.61

lattopp

FR GE SP

SZ IR AU

GR

IT

BE AL

US

NO PO

-2.14849 8.17752

NE

CA

12.3833 lpop

Page 25 of 91

3.59

UK

lattopp

FR GE SP

GR IR

IT BE

SZ

AU AL

US

NO PO

-2.14849 8.2239 lrgdpc 3.59

UK

NE

CA

9.44328

lattopp

FR GE SP

SZ IT BE AU AL NO

GR IR US

-2.14849 .32825

CA NE

PO

.45 gini

3.59

UK

lattopp

FR GE SP

SZ

GR IT IR AU AL NO PO

BE

US

-2.14849

CA

NE

38 ethdom

99

Number of obs = 16 R-squared = 0.6063 -----------------------------------------------------------------------------Page 26 of 91

lattopp | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lpop | .6046398 .3707692 1.631 0.142 -.2503556 1.459635 lrgdpc | -3.45209 1.985412 -1.739 0.120 -8.030458 1.126277 ethind | 1.977078 1.950378 1.014 0.340 -2.520503 6.474659 relind | 1.822315 2.361272 0.772 0.462 -3.622789 7.267419 gini | 3.157216 13.30192 0.237 0.818 -27.51707 33.8315 edsec | .0401198 .033673 1.191 0.268 -.0375303 .1177698 age65o | .7656431 .2847183 2.689 0.028 .1090814 1.422205 _cons | 9.059294 16.14223 0.561 0.590 -28.16475 46.28334 ------------------------------------------------------------------------------

Page 27 of 91

How to read these tables of regression output: 1. The output refers to an equation where the name in the top-left, here lattopp, is the dependent variable & the rest of the names in the first column are independent variables, lpop, lrgdpc etc. 2. The second column are the estimated coefficients on those independent variables. The last such coefficient, on _cons, is the coefficient on the constant, otherwise known as the intercept. Coefficients tell you how much the dependent variable tends to move for a 1-unit increase in the independent variable. The sign, therefore, tells you the direction of the relationship & the size tells you the magnitude (keeping in mind the substantive scales of independent & dependent variables). So, the above estimated equation is: lattopp = 9.06 +.605(lpop) -3.45(lrgdpc) +1.98(ethind) +1.82(relind) +3.16(gini) +.040(edsec) +.766(age65o) All data are for the 1980s: lattopp is the average number of political attacks & oppressions lpop is the natural log of the size of the population lrgdpc is the natural log of real GDP per capita ethind is an ethnic fractionalization index: the probability that two persons chosen at random will belong to different ethnic groups relind is a religious fractionalization index: the probability that two persons chosen at random will belong to different religious groups gini is the GINI index of income inequality edsec is the percentage of population in relative age group having attained at least secondary education (high school) age65o is the percentage of the population over 65 years old 3. The third column is the standard error. It tells you the precision of the estimated relationship: with how great a certainty do increases in the independent variable lead to movements in the dependent variable? You can read these as loosely something like the +/- number that comes with every survey. E.g., when the percent of the population high-school educated or better goes up by 1, the natural log of political attacks & oppressions go up by about .04 [coefficient on edsec] plus or minus .034 [standard error on that coefficient]. 4. The next column are the t-statistics. Standard errors should at the very least be smaller than the coefficient for us to lend any credence to the coefficient estimate. Wed prefer them to be no larger than half as large as the coefficient. The t-statistic is simply the coefficient divided by the standard error. So, loosely, larger t-statistics imply more precise relationships (tighter relationships, not necessarily larger ones). We look for t-stats approaching or exceeding about 2 (this is a very crude rule-of-thumb, please dont ever tell a statistician this). 5. The next column is the probability, under certain assumptions of having estimated a coefficient this far or farther from zero if the true relationship were zero. We tend to look to these to see if we can say with any statistical certainty that there is a relationship (positive or negative). We like these p-levels to be approaching or smaller than 0.10 (again, this very rough statement would make a statistician cringe). 6. The last two columns are the 95% confidence interval. Again, exceedingly crudely, but good enough for now, we can say that, under certain assumptions, we are 95% confident that the true relationship is between these bounds. 7. Finally, this is multiple regression. Its just like the scatter-plots for two variables, only we are looking for the relationship between a set of possible independent variables & the dependent variable. Thus, each coefficient is the effect of X on Y, holding all else constant or the effect of X on Y after having netted out all relationships between the other xs & Y. Questions?

Page 28 of 91

Summary: XVIII. Definitions of Cleavages A. L&E: a division on the basis of some criteria of individuals, groups, or organizations [between] whom conflict may arise (p. 53) B. Powell: a set of attitudes which divides the nations citizens into major political groups 1. Sometimes purely political: they reflect opinions about policies or & are developed, sustained, & organized by political leaders committed to such policies... 2. more frequently [they are the] alignment of large groups...based on deep social, economic, & cultural divisions in the fabric of society...[:] occupation, religion, language, race, custom, geography (p. 42) C. GLM: [cleavage] implies much more than mere division, more even than outright conflict, between 2 sets of people. Cleavages have 3 properties: 1. Social divisions on the basis of identifiable characteristics (e.g., occupation, status, religion, ethnicity); cannot be defined at the political level alone (e.g., not parties) 2. Groups involved must possess a group consciousness & be willing to act on that basis 3. The cleavage must be expressed in organizational terms (e.g., unions, action groups, etc.) [We emphasize this less as a criteria for being a cleavage] XIX. Typologies of Cleavages: A. We distinguish 2 Broad Cleavage-Types: 1. Ascriptive or Trait Cleavages: the basis for division are inherent characteristics of the individuals involved, theres no element of choosing to belong to some group: e.g., race, caste, ethnicity, language 2. Functional, Preference, or Value Cleavages: these are adopted views or, at least to some degree, chosen affiliations: e.g., class, preferences, ideology, & possibly religion B. We identify 4 Common Cleavage-Bases: 1. Religious: Christian/Non-Christian, Catholic/Protestant, Fundamentalist/Moderate, Religious/Secular 2. Ethnic/Linguistic: Race, Language, Cultures 3. Class/Economic: worker/employer, manual/skilled-labor, traded/non-traded, private/public-sector 4. Regional/Geographic: Urban/Suburban/Rural, Center/Periphery, Regionalism XX. Cleavage Characteristics Potentially Important to Political Outcomes: A. Salience: the importance of the dividing line & its associated conflicts to those involved. (Some refer to dormant cleavages as latent & active cleavages as manifest.) 1. Argument: cleavages are more likely to be more detrimental to democratic stability & social order the more salient are the issues involved. B. Divisibility: the degree to which the issues over which the groups divided by the cleavage are easily bargainable or compromisable as opposed to absolute (0-1). 1. Argument: cleavages are more likely to be less detrimental to democratic stability & social order the more divisible are the issues involved. C. Relative Power of the groups divided by the cleavage 1. Argument 1: cleavages are more likely to be more detrimental to democratic stability & social order the more closely is power distributed among the groups. 2. Argument 2: cleavages are more likely to be more detrimental to democratic stability & social order as the power of ascendant groups approaches that of dominant groups. D. Fractionalization & Polarization: Fractionalization refers to the number of cleavages operating (manifest); polarization refers to the distance separating the parties preferred policies & outcomes 1. Argument: cleavages are more likely to be more detrimental to democratic stability & social order the more fractionalized & polarized the cleavage structure E. Cross-Cutting v. Reinforcing: Cleavage structures (a set of cleavages) are cross-cutting to the degree that each divides the nation differently into different groups & reinforcing to the degree that each divides the nation into the same groups 1. Argument: cleavages are more (less) likely to be more detrimental to democratic stability & social order the more they reinforce (cross-cut) each other. XXI. Modernization & Democratic Performance: A. Key Features of Development: growth & spread of affluence; economic- & social-structural change; urbanization; education, literacy, & the spread of mass media; greater social & occupational mobility B. Arguments Relating Development to Democratic Performance: 1. modernization enhances political participation, government stability, & social order a. education & the modern person hypothesis (Lerner, Inkeles) b. modernization produces value harmonization c. affluence allows greater fulfillment of social demands, partly via increase of government resources d. the increasing interdependence hypothesis 2. modernizations political participation, government stability, & social order effects differ: its complicated a. Ratio of Economic Development to Socio-Political Mobilization (Huntington) b. Level of Economic Development at Time of Political-Competition Introduction c. Pace-of-Development Hypotheses C. Powells Expectations & Findings (sample of all democracies ca. 1980) 1. Participation: a. Modernization implies greater information, education, & psychological involvement of modern person, & more extensive & specialized group structures, more effective organization, so: Hypothesis: Modernization ==> Greater Participation b. Finding: Modernization ==> Lesser Participation 2. Government Stability & Effectiveness:

Page 29 of 91

XXII. A.

B.

C. D.

XXIII. A.

B.

XXIV. A.

a. Hypothesis: Modernization ==> Greater Government Stability & Effectiveness b. Finding: Modernization ==> No Apparent Effect on Government Stability & Effectiveness 3. Social Order & Violence: a. Hypothesis: Modernization ==> Greater Social Order & Less Violence b. Findings: (1) Social disorder greatest in early stages of modernization, less at undeveloped & developed stages (2) Violence, contrarily, strictly declines (increases) in modernization (underdevelopment) Ethnically/Linguistically/Religiously & Economically Divided Societies & Democratic Performance: Powells Expectations & Findings Ethnically/Linguistically/Religiously Divided Societies 1. Preliminary Problems: Which divides matter? Whether E/L/R divides cause conflict or economic differences coinciding therewith do? 2. Argument: E/L/R divisions are less divisible (compromisable or bargainable) and, thus, greater prevalence & saliency of E/L/R divisions worsens democratic performance ===> participation could go either way, government stability & social order should decline as E/L/R divisions rise 3. Findings: a. Weak evidence that participation decreases with E/L/R fractionalization b. Strong evidence that government stability declines with E/L/R fractionalization c. Greater E/L/R fractionalization produces no greater tendency to riot, but, given some degree of social upheaval, it appears more likely to become violent the greater is E/L/R fractionalization Economically Divided Societies 1. Classical Arguments: (date all the way back to Aristotle) a. Economic inequality produces political instability b. Large middle class essential to the defusing of this potential 2. Powells Arguments: a. No doubt that occupational divisions & the accompanying inequalities usually accompanying them frequently provide bases for political organization [still less question, I add, that they are a natural basis for potentially disruptive conflict] b. Whats less clear are the mechanisms by which these divisions affect democratic performance 3. Hypothesis: Economic inequality reduces participation, government stability, & social order 4. Powells Findings: Weak evidence of any relationship between inequality & democratic functioning Powells findings refer to the entire sample of democracies; we are focusing on developed democracies. Our findings in that sample differ somewhat. [See previous handouts for scatterplots & regressions.] Further Considerations: why have some divided societies functioned democratically, stable, & peaceably & others not? 1. History of oppression & repression, & of internal & external war: US, UK/N.Ire.; Israel; former Yugoslavia; the Franco-German, Franco-Spanish, German-Danish, Italian-Austrian, Italian-French border regions 2. History of homogeneity or heterogeneity: compare reactions to massive immigration in Germany v. Switzerland; Italy v. Canada; etc. 3. Institutional Structure: a. Regional autonomy (Switzerland, Canada, US, Germany, Australia, Italy, Spain: federalism or special relationships with peripheral groups) b. Parliamentary representation of minorities (Netherlands, Belgium: highly proportional systems) c. Functional representation (Austria: proportz, Netherlands: pillarization) d. Checks & balances (US); constitutionally ensconced liberties (all developed democracies) Patterns of Cleavages & the Number of Political Parties Argument: The cleavage structure suggests the number & types of social groups which parties could evolve to represent; i.e. social structure provides a basis for political organization suggesting that the fractionalization & polarization of society may be mirrored by the fractionalization & polarization of the party system. Evidence: 1. To measure adequately the effective number of relevant social groupings would be a daunting empirical task: only the crudest of attempts have been made so far. 2. However, my subjective impression is that the correlation between the number of social groups & the number of parties is weak at best: US: many groups, 2 parties; Switzerland: many groups, many parties; Italy: relatively homogenous, many parties; UK: relatively homogenous, 2 parties. 3. Why? Probably because between social structure & party system intervenes an extremely powerful set of variables: electoral law & institutions Lipset & Rokkans Diachronic Model of Cleavage-Structure Formation & Freezing Two stages, the Process of Nation-Building & the Industrial Revolution, each bringing two fundamental conflicts, form the underlying cleavage structure. The nature of their resolution forms the party system. 1. Nation Building I: Center v. Periphery a. Those seeking to standardize laws & consolidate power in the nation-state (center) against those striving to maintain local powers & privileges (periphery) b. The struggle can end in: (1) secession (e.g., Ireland from UK), (2) absorption of periphery & its gradual fading as distinct (e.g., Bretton or Occitan in France), (3) local autonomy (e.g., some Spanish & Italian regions), (4) or retention of diffuse, persistent tension (e.g., Germany, except Bavaria which is more 3) c. Only the last two are likely to result in this cleavage being reflected in the party system 2. Nation-Building II: Church v. State a. conflict between (Catholic) Churchs claim of rights & privileges, esp. its supremacy in certain moral areas, versus state consolidators; central issue is often education Page 30 of 91

B.

C.

XXV. A.

B. C. D.

E. F.

Resolution depends centrally on the nations history & role in the Protestant Reformation & later secularist movements: (1) Where Protestant movements allied with the state & won, party system does not usually reflect the Church-State cleavage (e.g., UK, Scandinavia) (2) Where Catholicism continued to dominate, the Church-State cleavage tended to persist (e.g., Latin Europe) (3) Where secularist gained most momentum (e.g., France), the party system often evolved anti-clerical elements as well 3. Industrial Revolution I: Urban v. Rural a. conflict between traditionally dominant rural interests & new commercial & industrial classes b. These have almost universally faded (basically because urban interests won), but... (1) ...in some places agrarian parties emerged, & these often persisted if strategists allowed the party flexibility from its early aims & purposes, (e.g., Sweden, Finland) (2) ...the split has seen rebirth in reverse as urban decline began, (e.g., US, UK) (3) ...some rebirth also in conflicts over agricultural protectionism (e.g., esp. Europe & Japan) 4. Industrial Revolution II: Labor v. Employers a. increased concentration of production, & the accompanying increased organization of labor & employers as a group, led to almostinherently-organized conflict b. Resolution occurred by two patterns: (1) Where workers rose & the bourgeoisie adopted an accommodative strategy, Socialist parties arose as the representatives of labor (e.g., UK, Sweden) (2) Where workers rose & the bourgeoisie adopted a stonewalling and/or repressive strategy, Communist parties arose as the representatives of labor (e.g., France, Italy, Germany, Spain) Argument: The pattern of how the conflicts arose in each of these countries as democracies & how they had been or were being resolved were being at the time mass democracy arrived was frozen. I.e., the cleavage structure is frozen in the party system by the mid-20th C because: 1. The underlying conflicts persist & the groups involved have developed collective identities 2. Major new political entities typically can arise only w/ large increases in suffrage & universal suffrage was mostly completed by then. 3. Political rules are made by & so favor established parties 4. Established parties follow electoral strategies to isolate their supporters from outside appeals Perceived increases in electoral volatility in the 1970s led some to seek explanations for unfreezing: 1. Social-Structural Changes in Developed Democracies a. Sectoral-Structure: industrial decline, massive agricultural decline, massive service-sector rise b. Erosion of Class Boundaries (1) general education & more wide-spread higher education (2) changing work modes: relative decline of manual labor (3) increasingly widespread affluence (though relative disparities may have widened too) 2. Changes in Patterns of Individual Voting Behavior: Decline of Structural Voting a. Decline of class voting & of religious voting but still recognizably different. b. Arguments: suggested reasons for the former are listed above, to which add: (1) secularization & decreasing coherence of views among religious & secular alike (2) increased individuality & political sophistication 3. Changes in the Issues Concerning those Individuals: Dealignment v. Realignment a. Realignment: new parties & changing support among existing parties (1) Some see a New Politics movement (a) new, highly educated, young middle class (b) distinctive new values & issues: environment, feminism, etc. (c) new-left parties: greens, new communists; new-right: anti-government, xenophobic, protest (2) GLM suggest this may be exaggerated (a) these new parties are still marginal electorally (b) parliamentary and/or electoral necessity implies increasing association of the new lefts & rights with old ones (3) Conclusion: Realignment is very limited, some new dimensions in a still-recognizable left-right divide. Great electoral stability remains across left-right blocks; instability is within blocks. b. Dealignment: non-partisan allegiances & no allegiances. Supposed evidence: (a) declining party identification, (b) rise of new parties & party-system fractionalization, & (c) increased electoral volatility. All of which are questionable. Additional Key Concepts & Considerations Statistical Interpretation: 1. How to interpret multivariate regression analysis 2. What are Lorenz Curves & GINI indices? 3. What is the Alford Index? 4. What is Structural Voting? Explanatory Power versus Parsimony Trends: Secularization & (Economic) Sectoral-Structural Change Questions regarding Development & the Saliency of Different Cleavages: 1. Does development increase the saliency of functional cleavages relative to ascriptive ones? 2. Does rise of post-industrial society & post-materialism decrease saliency of economic cleavages? 3. Did rise of welfare state & more-recent plateau-ing or receding of the welfare state increase or decrease the saliency of economic cleavages? 4. More broadly, are new cleavages replacing old cleavages? Income Equality correlates (positively) with unionization & structural voting. Why? Revisiting the classic Cost-Benefit Model of Voting: vote if p(V0-V1)+Bs > Cost(Voting)

b.

Page 31 of 91

SOCIO-ECONOMIC STRUCTURE & POLITICS (Part 2) (Dalton in LNN & Dalton ch.7-8)
Summary of What We Have So Far on Socio-Economic Structure & Politics: Basic Notion & Definition of Social Cleavage & Cleavage-Structure Typologies of Social Cleavages: 1. Ascriptive v. Functional Cleavages 2. Religious, Ethnic/Linguistic, Class/Economic, Regional/Geographic C. Characteristics of Social Cleavages & Cleavage-Structures Potentially Important to Key Elements of Democratic Performance 1. Salience of the Divide (latent manifest) 2. Divisibility (Bargainability or Compromisability) 3. Relative Power of the Groups Involved 4. Fractionalization & Polarization (of the Cleavage Structure) 5. Cross-Cutting v. Reinforcing Nature (of the Cleavage Structure) D. Concepts in Theory-Building: Parsimony & Explanatory Power E. Modernization & Democratic Performance; Key Hypotheses 1. Modernization Increases Participation, Govt Stability, & Social Order a. Education & the Modern Person Hypothesis b. Value-Harmonization Argument c. Increasing Density of Interactions (Increasing Interdependence of Individuals in Society) Argument 2. The Impact of Modernization Depends on Other factors a. Ratio of Economic Development to Social Mobilization Argument b. Economic Development-Level at the Time of Democratization Arg. c. Pace of Development Argument F. Why do some democracies function well & others not? Answer in this material emphasizes that it has to do with the nature & severity of underlying potential conflicts as expressed in social-structure & the way they are expressed as outlined above. This omits other important parts of the answer, some of which we will emphasize later. For now, note: 1. History: of oppression & repression & of homogeneity or heterogeneity 2. Institutional Structure: Regional Autonomy, Parliamentary Representation of Minorities, Functional Representation of Minorities, Guaranteed Rights G. Social-Cleavage Patterns & the Party System: Basic Notion is that the party system, i.e., the number, sizes, & ideological positioning of parties, must reflect, at some level, the underlying social-structure. But the relationship is moderated by: 1. Institutions operating between social structure & the party system (most centrally, electoral institutions). 2. Party & other elite strategies, & history (i.e., loosely, inertia) H. A Brief Introduction to Regression Analysis Dalton, Russell J. Political Cleavages, Issues, & Electoral Change, in LeDuc, Niemi, & Norris, Comparing Democracies, Sage (1996), & Citizen Politics, 2nd ed., ch. 7-8, Chatham House (1996). XXVII. Broad Theme: A. Two prominent changes in the link from socioeconomic structure to voting behavior 1. Traditional cleavages (esp. class & religion) transformed & weakened as predictors of individuals electoral choices 2. Changing bases of ideological conflict: the rise of post-materialism B. These are producing a general increase in issue-based voting 1. Less based on socially pre-determined & structured competition 2. A new calculus of electoral decisions a. Individualistic b. Varying issue positions & weights thereupon XXVIII. The post-war era through late 60s to 70s A. Party competition & voting patterns structured around social divisions within a polity 1. Review Lipset & Rokkans Diachronic Model & the Social-Cleavage/Party-System Freezing Hypothesis 2. Expected to persist because... a. Such cleavage structures long-lasting: the underlying conflicts persist & the groups have developed collective identities b. Universal suffrage largely completed: new parties tend to form (only) when new groups enter electorate c. Political rules are made by & so favor the established parties d. Established party strategies: often try to isolate their supporters from outside appeals e. Alignment with underlying social-structural groups advantageous to parties & voters [How? See B below.] 3. Why were these divisions so potent? a. Represented & (re-)produced deep ideological divisions (1) Class: conflict over the nature of politics & economics, over the very organization of society (2) Religion: basic value-systems (right & wrong) in conflict b. Social groups enabled parties to institutionalize a basis for support (1) Labor unions & firms provided organizational & people support (2) Ditto for churches B. What does the (reputation for) alignment of parties with social-structural groups do for voters & parties? 1. The groups so divided provided a social & political reference & a source of info for their members (voters) 2. They provide organizational structure, people, & political ally for parties XXIX. General decline in sociologically determined vote [n.b. from figure on pp. 172, 183-4, usually said to be post-1968 or so, but seems pretty much uniform since the war to me] A. Simple Class-Voting Decline Page 32 of 91 XXVI. A. B.

1. 2. 3.

B.

C.

XXX. A.

B.

C.

XXXI. A.

Define class voting: Blue/White collar divide appears to be decreasingly relevant as a predictor of party vote Explanations for specifically this cleavage decline: a. Class division better defined by degree of job autonomy & authority b. Rise of the middle class salatariat & affluent blue-collar worker c. Educational divide: skilled v. unskilled labor, human-capital rich v. poor d. Broad sectoral: public v. private; traded v. non; service v. industry v. agriculture e. Life-style differences: industrial v. yuppie 4. Even so, a general decline in the ability of such socioeconomic status indicators to predict the vote Religious Voting Decline 1. Define religious voting, two types: a. Denominational voting b. Religious v. secular 2. Explanations for specifically this cleavage decline: a. Secularization (1) Decreasing church memberships & attendance (2) Ethical individualitye.g. US Catholic Church b. Church/state or secular/religious conflict is largely resolved 3. N.b., esp. the religious/secular voting divide appears to have persisted to a much greater degree than simple class-voting 4. Daltons claim rests more solidly therefore on fact that the division is of relevance to a declining proportion of the population Other Key Social Divisions 1. Region a. Occasional, dramatic flare-ups b. Despite that, general decline in regions predictive power as well 2. Urban/Rural & Center/Periphery a. Differences diminishing due to transport & communications revolutions b. But note the still-present agricultural/industry/service divide 3. Race/Ethnicity a. Prominent exception to general declining relevance of socioeconomic structure to ones vote-choice b. Many societies are still homogeneous or nearly so, less relevant there obviously General explanations (hypotheses) for decline of socioeconomic structure as a predictor of individuals votes The underlying conflicts have increasingly been resolved 1. Nieuwbeerta (1995) finds that strength of class voting declines in the size of the welfare state 2. [Note: this would also explain the persistence of race/ethnicity & relative decline of other cleavages] 3. Dalton dismisses this, though, noting: a. Periodic recessions bring economic issues back to front [doesnt this weaken his other claims?] b. Persisting poverty, homelessness [yes, but do these segments of the population vote? It varies.] c. Crime [not exactly clear to me how this fits with any of the above cleavages per se] d. Large differences remain on abortion, homosexual rights, & other moral issues [yes, but its exactly in this respect that the religious/secular divide has persisted] 4. [As is clear from my comments here, I think the case for the relative resolution of underlying conflicts argument could be made more strongly. Notice that degree to which the underlying conflicts have been resolved might explain relative decline of some types of structural voting & relative persistence or rise of others] Parties have broadened or have sought to broaden their appeals across cleavage lines 1. Some evidence of platform convergence on broad socio-economic issues [e.g., industrial nationalization], but recently the evidence indicates a reversal of that trend: increasingly polarized politics. Perhaps, the polarization is diminishing again? (Pragmatic left winning in US, UK, France, Germany?) Remains to be seen. 2. Dalton notes that voters, party leaders, & political experts still perceive large and/or clear partisan differences which would imply... a. ...that its not that voters dont perceive or are unclear about partisan differences, nor that parties no longer express such differences, but rather... b. ...that the socio-economic status of the voter serves less as a determinant of how the voter will interpret & react to the different signals sent by parties & received by voters. Daltons preferred explanation emphasizes the structural & institutional changes which have produced a declining relevance of relatively fixed social characteristics for contemporary electoral politics 1. These social cleavages still very relevant to those deeply enmeshed in them, but thats fewer & fewer. 2. An increasing proportion of the population is characterized by: a. Fragmentation of life spaces [?] b. Less participation, membership, or involvement in stable & bounded social structures [?] c. Lifestyles becoming increasingly individualized & diverse [?] 3. [How about social fluidity & individual mobility across structured divides? Notice how this explains relative decline of class & less decline in religion & no decline in race/ethnicity.] 4. Dalton acknowledges that parties have contributed to & reinforced all this: a. By seeking to accommodate these changes in structure & institutions within the existing party structure they have eroded their core, established images/reputations b. But they do not wish to commit yet to any strong stand on the new issues. New Bases of Political Cleavages, i.e. Post-Materialism Define/Examples of Post-Materialism: Page 33 of 91

How did/do they arise? 1. The rise of new issues benefitted from the decline of old ones: the open space argument 2. Vice versa, the new issues cut across old divides, weakening their ability to divide: the emergence of cross-cutting issues argument 3. Meanwhile, citizens everywhere have increasingly demanded more opportunities to participate more actively in policy-making: the general further democratization of society & politics argument 4. Finally, notice that the new issues appeal exactly to those groups increasingly outside of traditional structured & institutionalized cleavages: young, new middle class, educated, non-religious, [others?] 5. [Other hypotheses (?): affluence?, education?, (cross-marginals), esp. regarding environmentalism are the problems increasing?] C. Before we go flying off proclaiming new divides, new cleavage patterns, though, we need to keep in mind: 1. Not all of these issues are so terribly new, even if perhaps there current prominence might be unprecedented. 2. Compared to the structured ideologies underlying class & religious cleavages, the post-materialist framework is (still) diffuse & imprecisewhat beside their novelty links these concerns? XXXII. So what is the impact of all this change? A. Decline in long-term partisan predispositions ==> 1. Rise short-term factors like issues & image? 2. Rise performance-based voting, like economic voting? 3. Rise candidate-centered politics [define, n.b. US always more so perhaps? Why?] 4. Rise in issue-voting? B. Issue-voting: 1. Define Issue-Voting: 2. Harder to study systematically because: a. Issues vary across elections & across individuals b. Multiplicity of issues at all times 3. Issues underlying old cleavages still around; new issues are added thereto C. Takes some time for, & its an uncertain process by which a new basis, for partisan competition can arise. 1. Groups must organize to represent & mobilize the interested. 2. Parties must establish positions, reputations, & images on those interests (in an uncertain environment). D. New Politics is orthogonal to old politics divides: Left v. Right, Religious v. Secular (or Cath. v. Prot. etc.) 1. The new divide is, maybe, sustainable society & libertarian values v. conservative social values & structured life choices. 2. [Is this another dimension? or is it more like a replacement/new manifestation of religious v. secular?] XXXIII. Summary of Material from Dalton on Decline of Structural Voting, etc. A. Lipset & Rokkans Model: 1. Two process: nation-building & industrial revolution 2. Produce two central conflicts (S.E. cleavages) each: a. Center v. Periphery & Church v. State b. Urban v. Rural & Labor v. Capital (Class) 3. The way these conflicts resolved historically determines how they are reflected in the party system 4. These resolutions are then frozen in the party system because: a. Underlying conflicts persist b. Universal suffrage had been largely completed c. Political rules made by & so favor established parties d. Established party strategies to insulate their voters e. The alignment of socio-economic structure with the party system & party ideologies serves both voters & parties: essentially, provides info & reference point for voters, provides organizational & electoral ally for parties B. Decline in Class Voting [Define]. Why? 1. Class needs to be redefined to reflect modern job structure: a. Rising middle-class both in lower-end white-collar jobs & upper-end blue-collar b. Educational (Skill-level) divide more than blue/white collar c. Broad sectoral conflict (across classes): public/private, trade/non-traded, service/industry/agriculture replacing class conflict 2. Even redefined, though, evidence indicates a general decline in class voting C. Decline in Religious Voting [Define]. Why? 1. Rising Secularization a. Declining church attendance (i.e., regularized, organized, structured religion) b. Church/State & secular/religious conflict largely resolved in many places 2. Among those professing a faith, religious voting remains strong; its more that the number to whom this applies is declining D. Other Structural Voting: 1. Regional: occasional, dramatic flare-ups, but generally decline in regionally distinct voting patterns 2. Urban/Rural & Center/Periphery: declining in most places, perhaps due to improving transport & communications, &, in many places, decreasing difference in urban/rural living standards [US exceptional] 3. Race/Ethnicity: where society is heterogeneous, these voting-pattern differences have persisted or even intensified. However, many countries remain basically homogenous ethnically. E. General Explanations for Decline of Structural Voting 1. The degree to which the underlying conflict has been resolved explains the relative decline of each type of structural voting. 2. Parties have broadened or sought to broaden their appeals across such S.E. cleavage lines 3. Dalton emphasizes: a. Increasing socio-economic mobility of individuals in modern society [i.e., the degree to which this applies to each cleavage explains its relative decline or persistence] Page 34 of 91

B.

F.

G.

b. Increasing individualization of society: largely via education & the above-mentioned mobility New bases of Political Cleavages: Post-Materialism 1. Define: emphasis on non-material, i.e. non-economic & non-structural in the above sense, issues 2. Examples: environment, gender equality, alternative lifestyles, etc. 3. Why do the new cleavages rise relative to the old ones? a. Rise of new stems from decline of old: the open space argument b. But, vice versa, the new cleavages tend to cut across the old ones, weakening their ability to divide. c. [Note the ambiguity about causality: what causes what here?] d. Increasing demand from increasingly sophisticated & individualistic populations to participate: the further democratization of society hypothesis e. A version of the open space argument: as old issues become increasingly resolved, at least in the perception of some voters, the relative value of addressing other issues increases for those voters. Impact of the Decline of Structural Voting & the Rise of New Politics 1. Decline in long-term partisan predispositions ==> a. Rise of short-term issues & image? b. Rise in performance-based voting [e.g., economic voting]? c. Rise in candidate-centered politics? d. Rise in issue-voting? [define] 2. New Politics is orthogonal [i.e., perpendicular] to old politics ==> two-dimensional politics? Is this just a rebirth of the old twodimensional politics? Used to be Economics & Religion; now its Economic & Social Policy? 3. [Extensions? Do you suppose these changes will have any effect on the stability over time of policies enacted by governments? How might it impact the sorts of policies which will make the agenda & be implemented? Broadly, an important question is how these changes might affect the incentives facing elected policy-makers.]

ELECTORAL SYSTEMS (Part 1) (Blais & Massicotte in LNN, Powell ch. 2, GLM ch. 11)
ELECTIONS & ELECTORAL SYSTEMS--Gallagher, Laver, & Mair + Lane & Ersson, pp. 226-40 + Powell, ch. 4, pp. 54-73 XXXIV. Importance of Elections A. Practically 1. Produce parliaments--legislative & executive policy-makers 2. Determined who becomes part of political elite 3. Bearing on the formation of governments a. Direct, especially in presidential systems b. Indirect, particularly in systems characterized by coalition governments 4. Focal point for activity for: Parties; Citizens (often only activity)--input mechanism; Interest groups B. Symbolically 1. Legitimization of the political system 2. Citizen means of participating, expressing opinion, evaluating 3. Give citizens feeling of exercising choices (even if individually little weight) XXXV. General Matters of Importance Regarding Elections A. Suffrage Expansion 1. Universal male typically by WWI 2. Universal female typically by WWII except: a. BE, FR, GR, IT just after b. PO, SP, Switz. 1970s (as late as 1991 in one Swiss canton) c. Are there any hypothesis on late & early expanders to full female suffrage? [Show figure] 3. Voting age generally reduced from 21 to 18 in postwar period...Why? B. Suffrage Restrictions 1. Generally citizens only (but UK/IR) 2. Prisoners & mentally ill usually excluded 3. Otherwise--all 18+ for most part & usually anyone who can vote can run 4. Registration: in most places governments responsibility, in some places individuals (Implications?) C. Turnout--Generally higher in Europe, declining since the late 1970s (Implications?) D. Timing of Elections: Endogenous v. Exogenous Election-Timing 1. Generally the incumbent government can call an election when it wants, subject to: a. Must be an election within X years (usually 4 or 5) b. Sometimes must call an election if fails a vote-of-confidence 2. Exceptions a. Presidents, Where directly elected, are usually fixed term b. FR: Parliament elects at Presidents discretion c. US: Legislature fixed terms & elections as well d. NO & SZ: Fixed four-year election interval e. SW: Election every 3 years, but government may call before that--still would be an election in third year (dont restart clock) 3. Implications? E. Other Elections 1. All countries also have at least some local elections, the offices so-elected vary greatly in their practical importance 2. EU every 5 years Page 35 of 91

Directly elected Presidents, (importance in *s): Austria (*), Finland (*), Iceland (*), Ireland (*), Port (*), France (*), US (**)--others have appointed (usu. by legislature) presidents--more figure-headish typically but can have some importance(e.g., Italy, *) F. Other Voting--Referenda 1. Most frequently used in Switzerland (nearly the referenda in the world occur in Switzerland) a. ==> voter fatigue? b. ==> democracy by referendum? 2. Others employing it: a. Italy: Approximately 1 year--key ones on Divorce, Abortion, Election Law b. France: President may call one c. Most other places, at discretion of parliament & very rare d. California 3. Issues over which referenda tend to called a. Issues that cut across party lines b. Constitutional Issues c. Constitutional amendments require referenda in Denmark, Ireland, & Switzerland; optional in France & Italy d. Major & fundamental changes in nations place in the world (1) NATO or EC membership (2) Neutrality/Allegiance e. Moral/Ethical questions--especially divorce & abortion in Catholic countries f. Why do you suppose theres a rising use of referenda? XXXVI. Types of Electoral Systems A. Definition of electoral systems: the mechanisms that turn votes cast by people on election day into seats to be occupied by deputies in the parliament. The electoral system is what converts the choices of voters into a legislature. GLM (p. 274) B. Where do electoral systems come from? 1. Determined by the political elite of the day, some of whose motivations may be partisan GLM (p. 274) 2. Their designs reflect the constitution-makers values, their expectations regarding the consequences of various arrangements, their often laboriously negotiated compromises Powell (p. 66) 3. Cultural/Historical Legacy: See Powell Table 4.3 (p. 67) 4. Generally not frequently tinkered with for electoral advantage, despite obvious opportunities a. France & Greece (especially the latter) exceptional on this b. Germany's famous 5% threshold c. Recent Italian & New Zealand changes d. French IVth to Vth transition 5. [If such a potent political engineering tool, why do you suppose it is so rarely manipulated?] a. Why would those in power change system that put them there? b. Parties uncertainty about future electoral position c. Difficult to change (constitutions usually require super-majorities) d. Relatively obvious opportunism when employed, may trigger negative voter-reaction e. Sometimes imposed from abroad (e.g. Germany, Japan, Italy, Aust.) C. Key distinctions between types of electoral systems 1. Primary Distinction: Proportional Representation (PR) vs. Plurality/Majority (P/M) systems a. Former stresses representation & concept of proportionality b. Latter stresses decisiveness & accountability model 2. Other key features [define each] a. District magnitude b. Degree of candidate vs. party voting (preference voting) c. Number of tiers d. Electoral formula (within the PR-P/M divisions) e. Thresholds f. Constituency pattern XXXVII. Plurality/Majority Systems A. Historically, plurality was the common system (through the 19th Century) B. Single-Member Plurality (SMP) a.k.a. first past the post, winner take all (UK, US, Canada, New Zealand--pre-1993 change) 1. Merits a. Simplicity--for voters, parties, & all involved b. Also argued that: (1) Produces majorities, & therefore decisiveness & accountability (2) Since one rep./MP per district, fosters MP-constit. bond 2. Criticisms a. Unrepresentative/Winner may be disliked by a majority b. Encourages strategic voting c. Anti-small party, & therefore, possibly, anti-minority (political & social minorities) 3. [Aside: plurality need not be conducted in single-member districts, multiple member, winner take-all is possible, tends to be even more disproportional & even more likely to produce elected majorities; e.g., US Presidential Electoral College, India used to have multiplemember districts] C. Majority Systems: 1. STV: The alternative or single-transferable (majority version) vote Page 36 of 91

3.

In use in Australia (some in France); used to be used in Illinois; was used more widely in US at one time Voters rank candidates; candidate with majority wins; if no majority, drop lowest contender & transfer his/her votes to those voters second choices; continue until someone has a majority c. Therefore, a Majority system 2. Multiple-Round Majority Balloting: e.g., the French (Vth) system(s), many US local elections (called run-off elections in US) a. French Vth Parliament: Simple vote; if no majority, eliminate & candidate with less than 12.5% of the vote; have another vote-plurality candidate then wins (often-called a plurality-plurality system) b. French Vth President: Simple vote; if no majority, drop all but top 2; second election will produce a majority winner (a pluralitymajority, or run-off system) c. [Describe French party systems of IVth & Vth; What do you suppose the impact of the Vths electoral process & the introduction of a strong President has been on party systems & party behavior?] 3. Merits & Demerits of Majority Systems a. Largely the same as SMP, but: b. Slightly more choice usually preserved because more parties usually persevere (why?) c. Less simple (and French version requires two trips to the polls) XXXVIII. Proportional Representation (PR) Systems A. The key feature of PR is the Multi-member district 1. Cannot divide one seat proportionally, so PR requires multiple seats per district 2. In fact, the proportionality of the result [define] tends to be proportional to number of seats per district (magnitude) B. Key types of PR: 1. List systems 2. STV systems (Ireland & Malta are only current users) C. List Systems 1. Each party lists a number of candidates (usually equal to the number of seats available in district) 2. List systems vary by a. Formula b. Number of tiers c. Degree of preference voting d. Thresholds e. District magnitudes 3. Formulas: Two Basic Types a. Largest Remainders (Quotas) Methods (Hare, Droop) b. Highest Averages (Divisor) Methods (dHondt, Sainte-Lague, & Mod. SL, Imperiali--Defunct Italian system) c. See Blais & Massicotte Tables 2.1 & 2.2 for examples d. Basics: (1) Highest Averages: divide votes for each party by a series of divisors, allocating seats one-at-a-time then dividing that partys vote by the next divisor, each stage awarding seat to the party with the most votes (so-divided) (2) Largest Remainders: divide the total votes in the district by the number of seats (Hare Quota) or number of seats +1 (Droop Quota). Thats a quota (Q). Each Q votes for a party buys it a seat. When no party can purchase any further seats, remaining are allocated to parties with the largest remainders, one for each until gone e. Relative Proportionality (roughly) (1) Imperiali > Sainte Lague & Hare > Droop > Modified Sainte Lague > dHondt (2) But, district magnitude is a considerably more important determinant 4. Tiers: a. Especially in small to moderate district magnitudes (e.g., 6 the average in Spain), considerable disproportionality can remain (1) Option 1: Larger districts (Finland, Portugal, Luxembourg > 12 avg., Netherlands & Israel whole country is one district) (2) Option 2: Higher tiers in electoral process to iron out deficiencies b. Fixed vs. Variable 2nd-Tier allocation (1) Fixed: DE (20%), IC (20%), (25%), NO (5%), SW (11%), GE (50%) Fixed Number of Second-Tier Seats: Set aside to be allocated so as to move district-level results as close as possible to national-level proportionality... (e.g. Vote Shares: Red=30%; White=40%; Blue=20%; Green=10% District-level Results in Seats: Red=25%; White=42%; Blue=19%; Green=5%) ==> Allocate set-aside seats to try to make 2nd row match the first ==> Larger 2nd-Tier proportion seats set-aside produces more proportionality (2) Variable (Austria, Belgium, Greece) Variable Number of Second-Tier Seats: All votes for all parties which were not used in winning seats at district level are pooled together at regional or national level & another round of PR allocation s undergone (not necessarily by the same formula) 5. Thresholds: a. Primary features designed to limit proportionality & in particular limit small parities. Why? (1) Self-interest on the part of larger parties (2) Concern that unmitigated proportionality ==> proliferation of small parties ==> difficulty forming stable governments (3) Aim to exclude extremists (e.g., Germanys 5% rule) b. Examples: (1) Germanys famous 5% rule (2) Sweden 4%; Aust. 4% or 1 seat; Netherlands .62% (almost meaningless) (3) Greece--PASOK has manipulated it relentlessly loaded or reinforced PR--as high as 17%, removed for 1989-90 (3 eles), 3% by New Democracy in 1990 ==> Very good example of electoral-law manipulation Page 37 of 91

a. b.

Preferential vs. Non-Preferential List: Who decides which of the parties listed candidates gets the seat the party has won? a. Non-Preferential or straight list (1) Relatively rare--FR (1986), GE, IT (since 1994, for PR seats), PO, SP (2) Party orders its candidates & their allotted seats go in order b. Preferential List--Many variations (1) IT (until 1994): Voters give up to 3-4 preferences, voters prefs. decide who gets seats--can choose partys default ordering though. This system received much blame in Italy for the clientelistic politics & corruption (vote-buying) that prevailed (2) Finland--Voters obliged to choose one candidate (3) Switzerland & Luxembourg--As many preference votes as seats, can cross party lines in preferential ordering (panachage) (4) Denmark--Party discretion as to how to list (5) In some cases, party default is very hard to override though a nominal preference option exists--BE, AU (pref. introduced in 1971, restrictive, changed in 1992 purportedly to more effective choice), NE (parties will usually demand any preferenced-in candidate cede his/her seat to the party order), NO, SW D. Single-Transferable-Vote (PR version) 1. Very Rare (relatively new): Ireland, Malta, & North Ireland (since 1972) 2. Aims at proportionality, but does not assume opinions organized on party lines 3. Mechanics: a. Voters rank candidates listed (therefore relatively small magnitudes are required) b. Droop Quota calculated (1) Anyone over quota is elected & remaining votes allotted to 2nd preferences (e.g. 100 first preferences, quota=75==> elected, 25 votes transferred to 2nd preferences in proportion to 2nd preferences of these 100 voters) (2) Continue until no-one exceeds quota, then eliminate candidate with the fewest votes, transfer his/her votes, & continue (3) Continue until number of seats in that district are allocated. 4. Merits (argued) a. More information on voter preferences revealed b. Not constrained by party lines c. Votes cannot harm a favored candidate ==> no inventive to vote strategically d. Allows voter input at polls on which tendencies within the party to expand/contract (via ranking) 5. Demerits (argued) a. May weaken party discipline [Aside: is party cohesion a good or a bad thing?] b. May induce vague positioning from candidates--as much incentive not to be disliked as to be liked almost c. Disproportionality because of small district magnitude (too complicated to have large lists of candidates to rank) d. GLMs read of Evidence: (1) The Ireland case seems to support a & b, but Malta does not (2) IR & MA not much different than others in practice on c (3) ==> GLM are advocates of STV-PR XXXIX. Assessing the Impacts of Electoral Systems A. The Simple Standard Story 1. Pluralist/Majoritarian ==> Disproportionality, but largest two parties take all or nearly all of the seats & thus the system produces stable majority governments 2. PR ==> Proportionality, but parties proliferate & thus system produces coalition governments, fractionalized & polarized legislatures, & thus unstable governments 3. Obviously, it is not quite so simple, but the broad outline is roughly correct (i.e., is supported by the evidence) B. Many other questions, however 1. Which affords better constituency representation? 2. Which offers better access to political & social minorities? 3. Redistricting/gerrymandering opportunities & incentives? 4. [etc.--see Powell, Blais, & Massicotte,...] C. G, L, & Ms Assessment on six possible effects of electoral systems 1. Proportionality--absolutely no doubt PR ==> more; in fact, tight relation with District Magnitude (see pictures) 2. The number of parties in legislature (parliament usually) a. Effective (size-weighted) vs. raw number of parities Taagepera & Laakso (n*=Effective number parties, n=raw number parties, Pi=Party is share of seats/votes): n* = [3i(1/Pi)2]-1 b. Number of parties in legislature or the number of parties contesting elections? Any way you slice it, PR ==> more parties. c. Why? (direct, mechanical effects) Non-PR ==> big bonus to large parties ==> fewer parties, esp. fewer effective parties, in legis. (indirect, psychological, strategic effects) Non-PR ==> strategic voting & strategic entry or not of parties d. Some counter-evidence (1) Several countries: Belgium, Denmark, Germany, Norway had multiparty before PR (2) The number of parities in Austria reduced after 1919 switched to PR (3) Malta nearly pure 2-party but a PR system (4) ==> PR does not always cause proliferation of parties, more a necessary than a sufficient condition (5) GLM: PR systems will give parliamentary expression to a multiparty system if other factors, such as the number of political or social cleavages, cause voters to create one I the first place, but PR does not by itself bring a multiparty system into being e. Most fractionalized parliaments: Belgium, Denmark, Finland, Italy, Netherlands, & Switzerland all have PR 3. Coalition or Single-Party Government? a. Again no doubt the broad outlines of the simple story are correct: Single-party-majority governments in 10% of cases in PR while 60% of plurality/majority Page 38 of 91

6.

4.

5. 6. 7.

Again, many exceptions (1) Minority governments have occurred in UK & elsewhere in P/M systems (2) Single-party governments have occurred in Germany & elsewhere in PR systems (3) four key parties in France (usually compete as 2-party coalitions) c. Tradeoff: Clarity of responsibility vs. accuracy of electoral message Constituency Representation a. Could argue that single representative per district facilitates constituent service b. Could counter that multiple representatives/district helps ensure at least one of your political persuasion to approach c. What little evidence available shows no discernible relationship Backgrounds of Parliamentarians: [What features of various electoral system do you suppose might impact the probability of female and/or minority candidates being elected? Evidence is that PR raises female representation in parliament.] Gerrymandering possibilities & incentives: Obviously gerrymandering much more effective in Plurality/Majority than PR. GLM state that differences in economic performance show little relation to differences in electoral systems. This is misleading. a. Economic policy varies quite a lot by electoral system, esp. insofar as they produce different types of governments b. Some evidence that economic performance varies by electoral system too

b.

Blais, Andr & Louis Massicotte, Electoral Systems, in LeDuc, Niemi, & Norris, Comparing Democracies, Sage (1996). XL. Describing the Various Types of Electoral Systems A. Plurality/Majority/Proportional Representation B. Magnitude C. Tiers D. Thresholds E. Selection of Candidates XLI. (Positive) Political Consequences A. Psychological & Mechanical Effects 1. Psychological: a. P/M/PR & the number of parties b. Electoral System & ideology / cohesion c. Electoral System & strategic voting d. Obvious impact in plurality elections: Gunther (1989) found supporters of small parties less likely to vote for them in districts of smaller magnitudemake sense? whats the logic here? 2. Mechanical a. vote-seat proportionality b. number of parties (raw v. effective number) c. Duvergers Law d. Lijphart finds: Plurality ==> about 2.0 effective parties , Majority ==> about 2.8, PR ==> about 3.6 e. threshold effects too f. Ordeshook & Shvetsova find: relationship between # parties & ethnic heterogeneity increases with district magnitude g. presence/absence of parliamentary majority: Lijphart: Plurality ==> majorities 93% of cases, Majority ==> 50%, PR ==> 20-30% depending on threshold XLII. Normative Debateinformed by Postive Theory, Evidence, & Debates Some Data: Raw Correlations of Variaous Socioeconomic Conditions & Political Outcomes
| lpop lrgdpc ethind relind gini edsec lmag vpart prop enpp dgov80 psupg80 npgov80 attopp --------+----------------------------------------------------------------------------------------------------------------------------lpop| 1.0000 lrgdpc| 0.2121 1.0000 ethind| 0.2919 0.4089 1.0000 relind| 0.3915 0.5090 0.4680 1.0000 gini| 0.0742 -0.3997 -0.0288 -0.2382 1.0000 edsec| 0.0173 0.5804 0.1929 0.0905 -0.2585 1.0000 lmag| -0.1310 -0.0565 -0.4859 -0.0592 0.0261 -0.1147 1.0000 vpart| -0.3360 -0.1393 -0.4735 -0.1867 -0.1405 -0.0934 0.3422 1.0000 prop| -0.1885 0.1740 -0.0715 -0.1855 -0.2574 0.0654 0.4178 0.0582 1.0000 enpp| -0.2744 0.0873 -0.3626 -0.0874 -0.0724 -0.0256 0.6698 0.0398 0.4062 1.0000 dgov80| 0.2125 0.1954 0.5897 0.5222 -0.0760 0.1199 -0.2862 -0.5699 -0.3684 -0.1180 1.0000 psupg80| -0.1414 -0.1086 0.2782 0.1825 -0.0981 -0.2837 0.0666 -0.2439 0.2347 -0.1192 0.2400 1.0000 npgov80| -0.2504 0.1249 -0.0328 -0.2107 0.0856 -0.0468 0.3536 -0.0776 0.8342 0.3345 -0.3579 0.3393 1.0000 attopp| 0.2792 0.0093 0.1424 0.1441 0.2152 -0.0428 -0.2643 -0.1296 -0.2204 -0.2986 0.4303 0.0928 -0.2025 1.0000 lpop: natural log of population lrgdpc: natural log of real GDP per capita ethind: ethnic fragmentation index relind: religious fragmentation index gini: GINI index of income inequality edsec: index of primary & secondary school enrollment lmag: natural log of electoral district magnitude vpart: voter participation rate prop: proportionality of legislative seat distribution to vote distribution enpp: effective number of parliamentary parties dgov80: average duration of governments (in months) in the 1980s psupg80: average percent of seats in parliament supporting the government in the 1980s npgov80: average number of parties in government in 1980s lattopp: natural log of the number of political attacks & oppressions in 1980s

Page 39 of 91

Determinants of the Proportionality of Electoral Outcomes Number of obs = 21 R-squared = 0.6983 -----------------------------------------------------------------------------prop | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lmag | 2.568204 .7283494 3.526 0.004 .994701 4.141707 lpop | -1.886149 .9376368 -2.012 0.065 -3.91179 .1394918 lrgdpc | 2.941667 4.083722 0.720 0.484 -5.880679 11.76401 ethind | -5.946238 6.171297 -0.964 0.353 -19.27851 7.386038 edsec | -.0141931 .0961284 -0.148 0.885 -.2218658 .1934797 US | 13.95252 5.210637 2.678 0.019 2.69562 25.20941 SZ | 4.030246 5.213686 0.773 0.453 -7.233238 15.29373 _cons | 80.45612 31.88155 2.524 0.025 11.58023 149.332 -----------------------------------------------------------------------------Number of obs = 23 R-squared = 0.6303 -----------------------------------------------------------------------------prop | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lmag | 2.394217 .681835 3.511 0.003 .9556706 3.832763 lpop | -.9292101 .5964194 -1.558 0.138 -2.187545 .3291249 ethind | -7.55057 5.362998 -1.408 0.177 -18.86551 3.764366 US | 12.91222 4.860203 2.657 0.017 2.658084 23.16635 SZ | 6.514613 4.677849 1.393 0.182 -3.354785 16.38401 _cons | 96.9738 5.687898 17.049 0.000 84.97338 108.9742 -----------------------------------------------------------------------------Determinants of the Effective Number of Parties in Parliament Number of obs = 21 R-squared = 0.2939 -----------------------------------------------------------------------------enpp | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lmag | .4464824 .3087327 1.446 0.172 -.220494 1.113459 lpop | -.1250728 .3974454 -0.315 0.758 -.9837014 .7335557 lrgdpc | .925385 1.731008 0.535 0.602 -2.81423 4.665 ethind | .5329234 2.615889 0.204 0.842 -5.118361 6.184208 edsec | .0055997 .0407469 0.137 0.893 -.0824286 .093628 US | -1.144733 2.208684 -0.518 0.613 -5.916305 3.62684 SZ | 1.341764 2.209977 0.607 0.554 -3.432601 6.116129 _cons | -4.88439 13.51395 -0.361 0.724 -34.0795 24.31072 -----------------------------------------------------------------------------Number of obs = 21 R-squared = 0.2581 -----------------------------------------------------------------------------enpp | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lmag | .5087292 .285378 1.783 0.095 -.0995395 1.116998 lpop | -.2890912 .324563 -0.891 0.387 -.9808809 .4026984 lrgdpc | 1.046867 1.603379 0.653 0.524 -2.370654 4.464388 ethind | 1.081365 2.262387 0.478 0.640 -3.740798 5.903528 edsec | -.0026652 .0366468 -0.073 0.943 -.0807759 .0754456 _cons | -3.897634 12.48691 -0.312 0.759 -30.51286 22.71759 -----------------------------------------------------------------------------Number of obs = 21 R-squared = 0.2460 -----------------------------------------------------------------------------enpp | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lmag | .4435446 .2316949 1.914 0.073 -.0452888 .9323781 lpop | -.2552405 .2989749 -0.854 0.405 -.8860225 .3755414 lrgdpc | 1.237125 1.12866 1.096 0.288 -1.144139 3.61839 _cons | -5.798764 10.07684 -0.575 0.573 -27.05904 15.46151 -----------------------------------------------------------------------------Number of obs = 21 R-squared = 0.1745 -----------------------------------------------------------------------------enpp | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lmag | .4554715 .2272422 2.004 0.059 -.0201518 .9310948 _cons | 2.907758 .5454194 5.331 0.000 1.766182 4.049333 ------------------------------------------------------------------------------

Page 40 of 91

Determinants of the Number of Parties in Government Number of obs = 21 R-squared = 0.7386 -----------------------------------------------------------------------------npgov80 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lpop | -.0857474 .1697331 -0.505 0.623 -.455564 .2840692 lrgdpc | .126545 .7444957 0.170 0.868 -1.495572 1.748662 ethind | .0155078 1.114686 0.014 0.989 -2.413185 2.4442 edsec | -.0084333 .0173481 -0.486 0.636 -.0462315 .0293649 lmag | .0064405 .1415198 0.046 0.964 -.3019047 .3147857 enpp | .5247365 .1179966 4.447 0.001 .267644 .781829 US | .1011615 .9493274 0.107 0.917 -1.967245 2.169568 SZ | .7775238 .9534556 0.815 0.431 -1.299877 2.854925 _cons | .5166641 5.778226 0.089 0.930 -12.07301 13.10634 -----------------------------------------------------------------------------Number of obs = 21 R-squared = 0.6960 -----------------------------------------------------------------------------npgov80 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------enpp | .5604526 .0849834 6.595 0.000 .3825802 .738325 _cons | .0025308 .3487783 0.007 0.994 -.7274705 .7325321 -----------------------------------------------------------------------------Determinants of Voter Participation Number of obs = 21 R-squared = 0.8380 -----------------------------------------------------------------------------vpart | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lpop | -4.246587 1.577331 -2.692 0.021 -7.71827 -.7749051 lrgdpc | 13.62715 6.147855 2.217 0.049 .0958134 27.15849 ethind | .3665997 9.440726 0.039 0.970 -20.4123 21.1455 edsec | -.3292375 .141427 -2.328 0.040 -.6405162 -.0179588 lmag | 2.831868 1.509369 1.876 0.087 -.4902306 6.153967 prop | -.4603597 .4176456 -1.102 0.294 -1.379591 .4588722 enpp | .0927027 .9852923 0.094 0.927 -2.075911 2.261316 US | -10.9058 9.83255 -1.109 0.291 -32.54709 10.7355 SZ | -41.28559 7.886355 -5.235 0.000 -58.64333 -23.92784 _cons | 64.44762 58.44081 1.103 0.294 -64.17974 193.075 -----------------------------------------------------------------------------Number of obs = 21 R-squared = 0.8379 -----------------------------------------------------------------------------vpart | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lpop | -4.235734 1.446679 -2.928 0.012 -7.361094 -1.110374 lrgdpc | 13.77331 5.398993 2.551 0.024 2.109492 25.43712 edsec | -.3283281 .1299159 -2.527 0.025 -.6089943 -.0476618 lmag | 2.838158 1.356566 2.092 0.057 -.0925245 5.76884 prop | -.4568824 .3620822 -1.262 0.229 -1.239114 .3253487 US | -11.0263 8.544266 -1.290 0.219 -29.48507 7.432461 SZ | -41.02204 6.447957 -6.362 0.000 -54.95201 -27.09208 _cons | 63.03857 52.37633 1.204 0.250 -50.11361 176.1907 ------------------------------------------------------------------------------

Page 41 of 91

Determinants of Government Durability Number of obs = 21 R-squared = 0.7844 -----------------------------------------------------------------------------dgov80 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lpop | -.3782034 2.813301 -0.134 0.896 -6.865687 6.10928 lrgdpc | -4.091615 9.845209 -0.416 0.689 -26.79471 18.61148 ethind | 15.2347 12.93021 1.178 0.273 -14.58243 45.05183 edsec | .1928452 .2382477 0.809 0.442 -.356555 .7422454 lmag | 1.058475 2.369944 0.447 0.667 -4.406626 6.523576 prop | .0176297 .5973344 0.030 0.977 -1.359826 1.395085 vpart | .0248729 .4148886 0.060 0.954 -.9318619 .9816077 enpp | -.605193 2.177049 -0.278 0.788 -5.625478 4.415092 psupg80 | .2608685 .2983566 0.874 0.407 -.4271431 .94888 npgov80 | -4.543667 3.260021 -1.394 0.201 -12.06129 2.973955 US | 19.76192 14.06015 1.406 0.197 -12.66086 52.18469 SZ | 24.22542 19.60525 1.236 0.252 -20.98437 69.4352 _cons | 36.34008 90.46144 0.402 0.698 -172.2644 244.9445 -----------------------------------------------------------------------------Number of obs = 23 R-squared = 0.5847 -----------------------------------------------------------------------------dgov80 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------ethind | 9.006614 10.86274 0.829 0.419 -13.91175 31.92498 psupg80 | .6395543 .2689653 2.378 0.029 .0720871 1.207021 npgov80 | -4.939279 1.973245 -2.503 0.023 -9.102461 -.7760958 US | 22.63304 11.39794 1.986 0.063 -1.414514 46.68059 SZ | 11.60115 11.95508 0.970 0.345 -13.62187 36.82416 _cons | -.8834984 14.97185 -0.059 0.954 -32.47134 30.70434 -----------------------------------------------------------------------------Number of obs = 23 R-squared = 0.5680 -----------------------------------------------------------------------------dgov80 | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------psupg80 | .7023434 .2558337 2.745 0.013 .1648566 1.23983 npgov80 | -5.21179 1.928714 -2.702 0.015 -9.263868 -1.159712 US | 26.66925 10.21622 2.610 0.018 5.205769 48.13274 SZ | 13.71925 11.57712 1.185 0.251 -10.60337 38.04186 _cons | -1.972604 14.78407 -0.133 0.895 -33.03278 29.08757 -----------------------------------------------------------------------------Determinants of Political Attacks & Oppressions Number of obs = 16 R-squared = 0.6223 -----------------------------------------------------------------------------lattopp | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lpop | 1.421063 .5419046 2.622 0.039 .09507 2.747056 lrgdpc | .3458066 1.710578 0.202 0.846 -3.839827 4.53144 ethind | -3.110833 2.794551 -1.113 0.308 -9.948853 3.727188 relind | -2.244102 2.512261 -0.893 0.406 -8.391384 3.90318 gini | 9.470207 15.16302 0.625 0.555 -27.63238 46.57279 edsec | .0360057 .0400388 0.899 0.403 -.0619657 .133977 lmag | -.3206208 .2987685 -1.073 0.324 -1.051681 .4104394 US | -3.668315 1.954821 -1.877 0.110 -8.451588 1.114959 SZ | 3.554796 2.143322 1.659 0.148 -1.689725 8.799317 _cons | -22.13167 16.97318 -1.304 0.240 -63.66356 19.40021 -----------------------------------------------------------------------------Number of obs = 16 R-squared = 0.5978 -----------------------------------------------------------------------------lattopp | Coef. Std. Err. t P>|t| [95% Conf. Interval] ---------+-------------------------------------------------------------------lpop | 1.515946 .4647737 3.262 0.011 .4441763 2.587716 ethind | -2.959762 2.430442 -1.218 0.258 -8.564372 2.644847 relind | -2.342026 2.101011 -1.115 0.297 -7.186966 2.502914 edsec | .0315434 .0312592 1.009 0.342 -.0405405 .1036272 lmag | -.3256798 .2626285 -1.240 0.250 -.9313023 .2799428 US | -3.735173 1.734805 -2.153 0.063 -7.73564 .2652943 SZ | 3.346048 1.882843 1.777 0.113 -.9957944 7.687891 _cons | -15.94199 5.526296 -2.885 0.020 -28.68565 -3.198325 -----------------------------------------------------------------------------Page 42 of 91

ELECTORAL SYSTEMS (Part 2) (Lijphart)

XLIII. A. Four basic properties of electoral systems Electoral formula 1. Majoritarian (plurality/majority) & P.R. 2. Different forms of proportional representation District magnitude Electoral threshold Assembly size Ancillary properties Ballot structure Malapportionment Presidentialism Apparentement Basic methodology Unit of analysis 1. electoral systemsets of essentially unchanged election rules under which on or more successive elections are conducted 2. Elections under the same electoral system are regarded as repeated observations of the operation of a single electoral system Dependent variables 1. Disproportionality 2. Degree of multipartism 3. Production of (parliamentary/legislative) majorities Independent variables: properties of the electoral system Strategies of empirical evaluation 1. Comparable cases (within-country, longitudinal): [advantages & disadvantages] 2. Cross-sectional comparison: [advantages & disadvantages] Summarizing the conclusions: Of the dependent variables, disproportionality is the best explained by electoral system properties The strongest explanitory factor across all the dependent variables is effective threshold, a combination of district magnitudes & legal thresholds The impact of of electoral system on multipartism is more modest (but there) than that on disproportionality Ditto for the effect of some of the other independent variables relative to effective thresholds Types, Patterns, & Trends The two key features are formula & district magnitude 1. Broad formula categories a. Majoritarian (1) Plurality (2) Two-ballot (3) Alternative vote b. PR (1) Largest remainders (2) Highest averages (3) Single transferable votes (STV) c. Semi-proportional (1) Cumulative vote (2) Limited vote 2. District magnitude a. Rae (1967), presaged by Hogan (1945) & Horwill (1925), demonstrated that this the key factor b. Considerable variation in magnitude exists across electoral systems which enables systematic study of its effects c. Distinct sub-components: (1) average magnitudeincreases proportionality, multipartism, reduces majorities (2) number of districtsditto, if lesser extent (3) assembly sizeditto if lesser extent d. Complication: there may be two or more levels of districts (tiers) 3. Legal thresholds, & at which tier applied, also important & variable 4. Magnitude & Legal Threshold jointly determine the effective threshold Four minor, but not necessarily negligible, features 1. Ballot structure a. categorical (within party only) v. ordinal (cross-party voting possible) b. Raes hypoth: ordinal ballots allow for more parties, ceteris paribus, but found to contrary 2. Malapportionment a. Irregular voting populations per district b. May therefore systematically favor certain parties c. Empirically tends to take the form of rural over-representation 3. Presidentialism a. Shugartif pres directly elected & pres&leg elections simultaneous, should tend toward bipartism b. Evidence favorable, does it extend, however, to semi-presidential systems? Page 43 of 91

B. C. D. XLIV. A. B. C. D. XLV. A.

B.

C. D.

XLVI. A. B. C. D. XLVII. A.

B.

Apparentement a. Possibilty of linked lists b. Should favor small parties & thereby circumvent some disproportionality effects of other features C. Patterns 1. Majoritarian systems a. Plurality more common than majority (France two-ballot & Australia alternative-vote systems) b. Almost exclusively single-member districts (no multi-member plurality districts left) c. Given single-member districts, number of districts is invariably very large d. Since natural exclude small parties, legal thresholds are exceedingly rare e. Tend to work against smaller partiesunless geographically concentratedthereby produce... (1) disproportionality (2) tendency toward bipartism (3) legislative majorities 2. PR systems a. Far the most common system (about 3/4 of all democracies) b. dHondt divisor formula is far the most common within PR systems c. Proportionality: dHondt & Imperiali least, then mod St. Lagu & LR-Droop & STV then LR-Hare d. Magnitude is even more important however D. Effective thresholdsroughly determined by district magnitude (or exactly legal threshold if latter is higher) 1. Three problems in determining it a. There is a lower threshold (lowest vote share which could produce a seat) & an upper threshold (lowest vote share which assures a seat) b. Such thresholds also depend on formula & the number of parties competing c. Number of parties, magnitude, etc. & therefore effective threshold can vary from district to district 2. Roughly equal to the larger of a. Legal threshold or b. Approx.: Teff = 50%/M (where M=magnitude) c. Except in plurality, single-member where its 35% by assumption E. Other central issues 1. Two- & multiple-tier systems a. Increasingly used, & usually accompanied by legal thresholds b. In terms of proportionality effects, 2nd tier dominates if enough seats at that level c. [What about in terms of other effects?] 2. Intermediate systems a. Japans SNTV/LVmore limited the vote, more it approximates Irelands STV b. Greeces reinforced PRfour tier system with increasing thresholds ==> considerable large-party bonus 3. Apparentement F. More general patterns 1. All electoral systems are fairly proportional as a result of chosen district magnitudes 2. The reason for two-tier districting is generally to combine close constituency contact of small district magnitudes at lower level with proportionality at higher; indeed lower-level magnitudes much lower 3. Legal thresholds usually applied in large upper levels; rarely do legal thresholds raise effective thresholds to the level of those systems without legal thresholds 4. Large countries have larger assemblies (duh); cube-root rule G. Links among the dimensions 1. Teff much higher in majoritarian than PR, but not systematically higher in less proportional PR formulas 2. Assembly size somewhat positively correlated with PM/PR dichotomy, again not so with various PRs 3. Assembly size & Teff slightly positively correlated, not within PR systems alone though H. Trends 1. Changes from PR to PM or v.v. extremely rare (France, very recently Italy & NZ) 2. More broadly, changes in systems (as defined by Lijphart) within country about 2.5/ctry over postwar era; range from 1-6/ctry 3. Which countries change: a. No change: US, Can, Fin, Switz, Bel, Ire, Lux, Port, Sp, UK b. Moderate changes: Austria, India, Jap, Austral, Costa Rica, Ice, Neth, Den, Ger c. Major Changes: Fra, Gre, Israel, Malta, Nor, Swe, & recently It & NZ 4. Large (>20%) changes in assembly size also rare 5. Other Trends: a. Toward two-tier b. From dHondt to more prop. PR systems c. Raise/install legal thresholds d. [note way first two counter-acted by last] XLVIII. The dependent variables A. Disproportionality 1. Various summary statistics devised for measuring deviation between seat allocation & vote shares 2. Lijpharts preferred: LSq = [3(vi-si)2 ].5 3. Notes a. There are differing notions of proportionality, some of which are embodied in different elect systems Page 44 of 91

4.

b. National proportionality depends on a range of things beyond district-level formulas & magnitudes [such as...] c. These measures are of absolute not relative disproportionality 4. For many purposes not raw disproportionality that matters but relative tendency to favor large parties 5. [interesting alternative: regress party seat shares on vote shares, compare coefficient to one] B. Party system 1. Key distinction is two-party v. multi-party 2. More generally, number of parties a. But what to do about widely varying size of parties (some negligibly small, but how negligible?) b. Solution is effective number of parties: (1) Effective number of elective parties: Ne=1/3vi2 (2) Effective number of parliamentary parties: Np=1/3si2 3. Ne always larger than Np, and/but highly correlated 4. Ne & Np conceptually & theoretically different things a. Ne affected entirely by the psychological (expectational) effects of electoral systems b. Np affected by both psychological & mechanical 5. Other key property is generation of parliamentary majorities a. Possible electoral outcomes (exhaustive, not exclusive list) (1) earned majority (2) manufactured majority (3) natural minority (4) artificial minority b. Well measure two things (1) Tendency to produce parl. majorities where electoral majority absent (ManMaj) (2) Tendency to produce parl. majorities (whether manufactured or earned) (Maj) c. [interesting fact: US has had 100% congressional majorities, only 8.7% manufactured; UK has had 92% majorities, all manufactured. What produces the difference do you suppose?] 6. A key problematic in all of this: what counts as a party? XLIX. So what? A. Proportionality is one of the aims, not necessarily most central, of democratic society; intrinsically interesting B. Proportionality is also the hypothesized link between electoral system & party system C. Empirical relationships: 1. In full 27-country sample: a. Corr(LSq,Ne) = -.11 b. Corr(LSq,Np) = -.45 ** c. Corr(LSq,Maj) = +.58 ** d. Corr(LSq,ManMaj) = +.63 ** 2. In PR-systems sample: a. Corr(LSq,Ne) = -.02 b. Corr(LSq,Np) = -.29 * c. Corr(LSq,Maj) = +.42 ** d. Corr(LSq,ManMaj) = +.41 ** D. Why is relationship with number of parties as weak as it is, esp. with Ne? [dont exaggerate this question, the relationship certainly exists] 1. Bi-directional causality [Ne==>+Disprop, Disprop==>-Ne] 2. Psychologicalparties dont ever have to leave electoral arena 3. Multiple other factors involved here (e.g., geographic concentration of support) 4. Somewhat of a statistical artifact: systems performing oddly (producing too many or too few parties relative to designers aims) will tend to be the ones changed. Lijphartd definition of an electoral system therefore will tend to over-count the amount of country-time in which relationships running in odd directions are occuring. E.g., US plurality, single-member produces around Np=Ne=2; partly because of this, never changed. Thats one case in Lijpharts counting. Greeces less than proportional PR system was still producing too many parties for leaderships taste, so it was changed several times raising the bar. Thats several cases by Lijphart;s counting. In actuality, the US 1 case is around 20-25 elections, each Greek case is very few elections. L. The Comparable Cases Method A. Synonyms: the comparative method, method of controlled comparison, most similar cases design, natural experiments, etc. B. The basic idea is to try to approximate a controlled experiment. 1. How: Select cases for comparison which are alike in all (independent-variable) dimensions except in regard to the one (or as few as possible) factor the effect of which you wish to determine. Any variation in the dependent variable, then, may be attributable to that single varying independent variable. 2. Advantage: to the degree you succeed in so isolating factors, you can be certain you have reliable results. 3. Disadvantages: a. No guarantee youll ever find any such perfect experiment or even a very good one b. This social science. You cannot control the entire environment of conditions under which the variation occurs, nor can you ever be certain all remaining factors are irrelevant. Cant even be sure you could list all relevant factors, so something will always be outside of your view & possibly therefore varying across your cases without your knowledge. To the degree youve missed something, the limited number of cases youll have for comparison becomes that much more extreme a limitation on the reliability of your results. LI. Controlled comparisons: within country changes in a single dimension of electoral systems A. How good is this as a natural experiment? 1. [What sorts of factors does this control for?] Page 45 of 91

2. [What sorts of factors might be left out?] Changes in electoral formula (Table 4.1): there were seven 1. All seven produced changes disproportionality as wed expect 2. Only 3 of 7 produced changes in the expected direction in the effective # electoral parties (eNep), though the magnitudes of these changes were such that average change was in right direction 3. 6 of 7 produced changes in the expected direction in the eff. # parliamentary parties (eNpp) 4. Only two of these produced any observable changes in majority production: a. Norway: 2 of 2 under dHondt, 2 of 9 under mod. St.-Lag. b. Sweden: 1 majority generated, but under mod. SL not dHondt 5. Conclusions: a. Overwhelming to strong support for Formula==>Disproportionality & Formula==>eNpp b. Weak to no support for Formula==>Majorities & Formula==>eNep C. Controlled comparisons of major changes (20%+) in Effective Thresholds (Teff) & Assembly Size (AS) 1. Four major changes in Teff: a. all 4 produced changes in disprop., eNep, & eNpp in expected direction b. magnitude of Teff change doesnt appear related to magnitude of eN changes though c. Only one relevant change in majority generation: from 22% in Norway 1953-85, to no majority in the 1989 election (1 case) under the more proportional system (as expected, but not much evidence) 2. Nine major changes in AS a. 8 of 9 right direction on disprop. b. Only 3 of 9 on eNep & 4 of 9 eNpp c. Regarding majority generation: 3 unchanged, 3 of 6 in right generation on both types of majorities, 2 of 6 in wrong direction on both, & 1 of 6 split 3. Conclusions: a. Not much evidence on Teff, but most or all of it points in the right directions b. Assembly size appears weaker, especially beyond its proportionality effects (which are strong) D. General conclusions from the 20 instances of major change in a single dimension 1. Predictabilty of longitudinal (within country over time) changes in proportionality & party system on the basis of major changes in one of the three key features of electoral systems (formula, Teff, AS): a. Changes in proportionality almost invariable occur as predicted: 19 of 20 (95%) b. Changes in eNpp usually occur in the predicted direction: 15 of 20 (75%) c. Changes in eNep occur as predicted only slightly more than half the time: 11 of 20 (55%) d. Changes in Majority Generation occur slightly less than half the time: 9 of 20 (45%), when they do, a bit over half the time they occur in the right direction 5.5 of 9 (61%) 2. Relative importance of the factors: a. Teff appears very strong across the board as an explanitor (not too much evidence yet though) b. Formula perhaps a bit less but still strong c. AS does well for disproportionality but otherwise a weak explanitor/predictor of changes E. A refining question: 1. Why eNpp so much more responsive to electoral system than eNep? Why eNep appears basically unaffected in fact in this longitudinal analysis? a. Takes time for expectational effects to come on line; politicians presumably know the expected effects of electoral systems, voters have to work them out b. Historicitytwo (effective) party system wont become 3 over-night, etc. c. Uncertainty regarding electoral support for various potential new parties, or over who is going to lose out under new rules, & whose support is now vulnerablewinners curse leads to excessive net entry of parties as rules change d. Spurious/endogeneity: when do electoral rules change? Isnt it likely that the same conditions which trigger electoral law change are likely to be producing party system changes? e. Periods of time being compared are usually so long that its questionable whether these are reliably controlled casestoo much else is also changing. On other hand, this too much else ought to net out on average across some number of such comparisons. 2. One strategy which partially addresses these is to compare last election under old system with 2nd or 3rd under new rather than all elections under old with all under new. Drawback is that a single election is a less reliable indicator of the functioning of the electoral system than an average over some number of elections. [V(X)/N V(Xbar)] 3. The results of this test... a. ...confirm that disproportionality strongly determined by electoral system b. ...strengthen the degree to which eNpp appears to be impacted by the electoral system c. ...weakens the degree to which eNep appears to be impacted by the electoral still further d. ...Lijphart doesnt offer any comment on how the majority-generation conclusions are impacted F. Examining 11 cases where 2 changes made in the same direction reinforces these conclusions further G. Within system changes (i.e., those of less than 20% in Teff and/or AS) show little consistent impact on any of the dependent variables, which result Lijphart claims strengthens his case for treating the systems as he defines them as internally homogeneous LII. Bivariate & Multivariate Analysis A. Bivariate Correlations: 5 dependent variables with... 1. ...electoral formula in Table 5.1 (p. 96) 2. ...Teff in Table 5.2 (p. 99) 3. ...AS in Table 5.3 (p. 101) B. Multivariate by regression in tables 5.9 & 5.10 (pp. 108-9) across all systems & table 5.11 (p. 112) in PR systems only B. Page 46 of 91

C.

LIII. A.

B.

C.

D.

LIV.

Basic conclusion: 1. Some support for expected effects of all three dimensions 2. Teff the main factor (n.b. that it subsumes the PR/PM distinction which is itself a very strong predictor) 3. Disproportionality is the dependent variable most completely explained by electoral system The impact of four ancillary properties of electoral systems Ballot structure: categorical (within-party voting only) v. ordinal (potentiality of cross-party voting) 1. Douglas Raes hypothesis that ordinal ballot allowed voters mandates to be distributed across a larger number of parties, therefore might cause micro-fractionalization & contribute to greater eNep 2. Rae himself found no support for that hypothesis: my theory is absolutely wrong 3. Lijphart extends the hypothesis to eNpp & disproportionality a. ...to eNpp because relatively obviously fewer (more) parties competing in elections should mean fewer (more) winning seats b. ...to disproportionality because, for a given set of electoral rules, more parties produces more disproportionality 4. Lijphart finds that... a. ...these hypotheses are supported in higher Teff systems (8%+) but opposite in low Teff b. ...categorical produces consistently more manufactured majorities than ordinal, ceteris paribus c. ...controlling for Teff, a switch from categorical to ordinal appears to lower the frequency of manufactured majorities by 14-5% Malapportionment: differing numbers of voters per representative across districts 1. Gallagher: since this leads directly to over- or under-representation of some voters, hypothesizes that it contributes to disproportionality 2. Lijphart finds no empirical support for that hypothesis. Explanation: a. Malapp. Highly correlated with single-member plurality, once control for that (or Teff more generally), no relationship between Malapp & disprop is found b. Impact of malapp depends very heavily on the geographical distribution of party support relative to the distribution of voters/repfavors some smaller parties like Scottish National Party, Plaid Cymru, Australia National Party while it disfavors others 3. Better questions, then, might be where does malapp arise, who benefits from it, & why? Presidential government & elections 1. Shugart & Carey: when president is powerful, elected by plurality rather than majority (why?), & elected simultaneously with legislature, it provides an impetus toward bipartism 2. Lijphart finds that, among systems with similar Teff, presidential systems are... ...than non-presidential a. ...less disproportional... b. ...smaller eNep & eNpp... c. ...higher frequency of manufactured & earned majorities... [with one exception: US high earned, yes, but few manufactured, why?] 3. However, these conclusions based on only US & Costa Rica 4. Lijphart attempts to extend consideration to all popularly elected (powerful or not) presidents, finds no empirical support for that broader set Interparty electoral links: 1. At least three modes by which such links can be achieved a. apparentement: (1) parties overtly & explicitly linked lists (2) SZ, IS, Neth, SW 1948, NO 1945 & 1985 b. Transferable votes (STV or AV): (1) parties can urge voters to list some other partys or partys candidates second (2) Australian & Irish parties often do; Maltese usually do not c. French 2-ballot majority implicitly allows for something quite like this (and quite often used so) 2. Hypothesis: since favors small parties, should reduce disprop & increase (reduce) eNep, eNpp, (manufactrued & earned majorities) 3. Only consistent effect found is on disprop., others may be there but not much evidence Grand summary of findings: tables 6.2 & 6.3, though perhaps understate strength of conclusion in favor of the elect formula & AS links eNpp. Remains broad conclusion that Teff the main factor & Disprop. The most completely determined

Table of Correlations: LSq ENEP ENPP LSq 1.00 -0.11 ENEP -0.11 1.00 ENPP -0.45 0.91 ParlMaj 0.55 -0.52 ManMaj 0.59 -0.30

ParlMaj ManMaj -0.45 0.55 0.91 -0.52 1.00 -0.67 -0.67 1.00 -0.49 0.83

0.59 -0.30 -0.49 0.83 1.00

PARTY SYSTEMS (Part 1) (Mair in LNN, GLM ch. 7-10, Powel ch. 5, L&E ch. ch. 3-5)
LV. Patterns in Party Politics (GLM ch. 7) A. Intro: 1. Party System: character of political competition in a country, described in terms of the relative strengths & policy positions of its parties (GLM, p. 151) 2. Though each system is in some ways unique, many similarities in party systems across countries B. Seven West European Party Systems Described: common basis in left/right divide defines the competition 1. UK a. Description: (1) Major Parties: the players Page 47 of 91

2.

3.

4.

5.

6.

7.

(2) Relative Strengths: (3) Describe the types of government & general pattern through time of them b. Responsible Party Government (1) ideological decision sharply defined for voters (2) cabinet government with relatively unchecked power Sweden a. Description: (1) Major Parties: the players (2) Relative Strengths: (3) Describe the types of government & general pattern through time of them b. Key differences from the UK (1) Soc Dems much more successful than Lab (2) Soc Dems not quite monop of leftsmall but persistent comm (left) party (3) Moderates quite far from monop non-soc opposition (a) Liberalscenter-type libs (b) Center Partyagrarian (c) Christian Democratsnewer (4) Relative strength recently of Ecology Party & New Democracy (rt, protest) Germany a. Description: (1) Major Parties: the players (2) Relative Strengths: (3) Describe the types of government & general pattern through time of them b. Key differences from the UK (1) relatively strong position of the FDPFDP more conservative than UK Liberalssec. opp. to Cath. pol. rather than lib. opp. to sec. conserv. (2) CDU/CSU a Christian Democratic party not a secular conservative party (a) gen. more receptive to social justice concerns, thus somewhat more centrist on econ. policy (b) gen. more concerned with moral issues, thus more conservative on social policy, esp. that on which major churches have a stance (c) CSU is the more Catholic & socially conservative of the two (3) SPD was one of most radical in 19th C, one of most moderate now (a) excluded for its extremism until 66, after Bad Godesberg commitment to free-mrkt & NATO (b) since 66 increasingly de-ideologized, increasingly pro-system, & toward consensual politics (4) TRIANGULAR POLITICS: CDU/CSUFDPSPD, how do they ally on different sorts of issues? c. Changes in the 80s & 90s (1) Greensright around threshold since 80s, possible impacts on governmental coalition patterns? (2) 1990 Unification: 12 mill. new, presumably volatile voters, originally mostly to CDU & FDP benefit, recently turned against them. Why? (3) Greens allying with related social movements, becoming more effectively organized to compete (4) Die Republikanernot yet a parliamentary player, but disturbing coupled with recent social strife (5) Politikverdrossenheitdisillusionment with politics Netherlands a. Description: (1) Major Parties: the players: PvdA, CDA, VVD (2) Relative Strengths: 1/3+1/3+1/5+ smaller, periodic players (3) Describe the types of government & general pattern through time of them b. Triangular + small party complications Italy (pre-1994) a. Description: (1) Major Parties: the players: MSILibRepDCPSDIPSIPCI+host of smaller (2) Relative Strengths: DC 1/3, PCI 1/3 (3) Describe the types of government & general pattern through time of them b. More fractionalized & polarized than most other systems (nearest comparison was French IVth) c. Apparent Changes (1) Major Parties: the players: ANLegaForzaCentroPSIGreensPDSRifondazione, + Rete (2) Relative Strengths: roughly even blocks (3) Describe the types of government & general pattern through time of them (series tied: 1-1) France a. IVth Republic: PCF (1/4)PSF (1/5)Rad (1/10)Cath MRP (1/4)Gaullists (1/5)Poujadists (1/10) + smaller b. Fifth Republic: PCFPSUDFRPR, + smaller (1) Two-bloc system: what facilitated the change? (2) Shift in balance of strength within each bloc (3) Front National, Generation Ecologie, Les Verts Spain a. early volatility & flux, ill-defined loose alliances b. early domination by Suarez UnionDemCtr ==> broad Ctr-Rt & Ctr-Lft coals Page 48 of 91

C.

LVI. A.

B.

C.

D.

E.

LVII. A.

B.

collapses about 82, replaced by PSOE domination, rt, in fragmented disarray Emerging system: PCE+PSOE+smaller lefts vs. PP (sec cons) + PC (peoples coal, loose lib-cath-cons coal) + bunch of regional parties Uniformity & Diversity 1. Major commonality is tendency toward left-right competition, occasional exceptions dominated by center 2. Core Differences a. location of liberals: (1) Egalitarian (center-left) liberals: e.g., UK Liberal Democrats (2) Libertarian (center-right) liberals: e.g., German FDP or Italian PLI b. whether major left or, more rarely, right party can govern alone c. strongly structured v. loose/fragmented governmental alliances d. Degree of party discipline 3. The Nature of Left-Right Competition a. Left working-class parties: common historical origins of both major strands (Soc, Comm) b. Right middle/upper-class parties: more variegated origins & politics (1) religious v. secular conservatives (2) rural/farming v. industrial/service (3) some rightist cultural linguistic, subcultural, regional c. Compare development of Welfare State across configurations of left/right competition d. Complications arising: New Left & New Right & new issues 4. Conflict within Left & Right a. Socialist/Communist & Right/Far-Right more usually bitter enemies than allies b. Systems with much electoral strength at extremes tend to produce centrist multiparty coalitions c. Growing Disillusionment with Politics==>anti-party parties (1) sources of this trend? (2) effects of this trend? 5. Other Dimensions of Party Politics a. Other cleavage lines (1) religion (2) cultural/ethnic/linguistic (3) post-materialist/materialist b. Within the broad rubric of Left-Right Competition: social & economic dimensions Party Families (GLM ch. 8) Definition/Grouping Characteristics 1. Origins 2. Links they form among & between themselves, within & across countries 3. Similarity of (broad) policy stances Families of the Left 1. Social Democrats (shrinking support or unchanged) 2. Communists (shrinking support in most places, recent Italy exceptional but perhaps not comm. any more) 3. New Left (growing in most places, social democrats usually moving to absorb eventually) 4. Greens (growing or unchanged support in most places) Families of the Center & Right 1. Christian Democrats (shrinking or unchanged support in most places, until very recently) 2. Secular Conservatives (growing or unchanged support in most places, until very recently) 3. Liberals (growing support in most places, but some places the CDs or Sec Cons. moving to absorb) 4. Agrarian/Rural (shrinking or unchanged support) 5. Far Right (growing support in some places) Other 1. regionalist, nationalist (unchanged or growing support) 2. miscellaneous, protest (growing support in most places) Patterns of Partisan Competition among Families 1. Christian-Democrat-led Right a. left strong, united (AU, GE) b. left strong, divided (IT) c. left weak, united or divided (BE, LU, NE, SZ) 2. Secular-Conservative-led Right a. left strong, united (UK) b. left strong, divided (Fin, Ice, Fra) 3. Fragmented Right, Left strong, united (Den, Nor, Swe) Party Systems & Structures of Competition, ch. 3 by Peter Mair in LeDuc, Niemi, & Norris, eds., Comparing Democracies Summary: ...the whole notion of a party system is centered on the assumption that there exists a stable structure of partisan competition [in elections & for control of government]. Structures of competition can be seen as closed (and predictable) or open (and unpredictable), depending on the patterns of alternation in government, the degree of innovation or persistence in processes of government formation, & the range of parties gaining access to government (p. 84). Brief Intellectual History of Party-System Classification Schemes 1. Duverger (1954): the number of parties, esp. 2-party v. multi-party systems Page 49 of 91

c. d.

C.

D.

2-party systems argued to foster/enhance: (1) single-party government (2) accountability (3) alternation in government (4) moderate, centripetal competition b. Multiparty systems argued to... (1) ...foster coalition governments (2) ...prevent voters from gaining a direct voice in government formation (3) ...not ensure alternation in government (4) ...sometimes favor extremist, ideological confrontations between narrowly based political parties c. Although exceptions are fairly common, broadly speaking, all of this seems empirically true 2. Dahl (1966): the competitive strategies adopted by the parties at the electoral & the legislative level a. Strictly competitive systems b. Cooperative-competitive systems c. Coalescent-competitive systems d. Strictly coalescent systems 3. Blondel (1968): number of parties & their relative size (later considers their ideological placement also) a. 2-party systems b. 2.5-party systems c. Multiparty systems with a dominant party d. Multiparty systems without a dominant party 4. Rokkan (1968): more disaggregation of the multiparty-system category a. 2 + a small third-party systems (UK, Germany) b. 1 large party v. a coalition (Sweden, Norway, maybe Ireland) c. 3 or more relatively even competitors (Netherlands) 5. Sartori (1976): fractionalization & polarization, key role of anti-system parties a. Types: (1) 2-party, ideologically polarized (Sartori considered this category empty) (2) 2-party, moderate ideological distance (US, UK) (3) Moderate Pluralism: multiparty, moderate ideological distance (Denmark) (4) Polarized Pluralism: multiparty, ideologically polarized (Italy) (5) Dominant-Party systems: Japan (until recently perhaps, India, Mexico) b. Virtues of Sartoris scheme (1) most comprehensive, in scope & depth, of available topologies (2) has subsequently been employed fruitfully in a wide variety of theoretical & empirical comparative studiesproven utility (3) explicitly concerned with patterns of competition & with interactions between parties, thus more directly concerned with party system (4) underlines impacts of systemic characteristics on party strategy, & electoral behavior & outcomes c. Changes since Sartoris writings which may be somewhat problematic for the theoretical approach (1) some see a trend toward & thus increasing crowding of the moderate pluralism category (2) some see a diminishing of polarized pluralism, since, by Sartoris definition, that requires anti-system parties on both sides (right & left) & most such are disappearing, esp. on left Mairs Three Distinguishing Characteristics of Different Types of Party Competition for Government 1. Degree of Alternation in Government a. Wholesale alternation (1) single-party v. single-party (2) single-party v. coalition (3) coalition v. coalition b. Partial Alternation c. Non-alternation d. [Think about Powells three aspects of democratic performance, & also about accountability; how might variation in the degree of alternation in government affect those things?] 2. Stability & Consistency of Government Alternatives a. Define: degree to which the alternative government formulas (i.e., combinations of parties in government) are known or predictable before-hand (i.e., stable & consistent) b. [Think about Powells three aspects of democratic performance, & also about accountability; how might variation in the stability/consistency of government alternatives affect those things?] 3. Range of Parties Serving as Potential Governmental Actors a. Define: degree to which access to office is widely or narrowly dispersed b. N.b., what matters is whether each party is viewed by others as a potential government participant, not whether it actually is a legitimate potential participant in any abstract sense c. [Think about Powells three aspects of democratic performance, & also about accountability; how might variation in the range of potential government actors affect those things?] Mair then combines these three to a single dimension: the openness or closedness of the party system 1. Definitions: a. Closed: highly predictable, little or no change over time in the range of governing alternatives or pattern of alternation, & with new and/or outsider parties finding entry difficult Page 50 of 91

a.

E.

LVIII.

Open: highly unpredictable, with different patterns of alternation, frequent and/or large shifts in the make-up of alternatives, & relatively easy access to new parties & few or no real outsiders 2. Closedness or Openness depends on... a. The ideologies & government-formation strategies of parties (e.g., DC in Italy) b. The electoral strategies of parties (e.g., Fianna Fail in Ireland) c. [electoral & government-formation strategies are not so separable really] d. Closure thus depends in large part on the norms of competition among the parties & so requires time to establish ==> new democracies must be open by definition, takes time to close them: a process of structural consolidation (e.g., Greece, Portugal, & Spain) Party Systems & Electoral Outcomes 1. We can now see that party system change can occur without changes in the structure of electoral support for the different parties & vice versa. Mair gives four examples: a. Denmarks Earthquake Election in 1973 (1) prior to 1973, 5 parties had 93% of the vote, 1973 saw 5 new parties, & old 5 dropped to 65% of vote, plus new entrants included Communists & right-wing Progress Party so polarization rose too (2) some time before any government was able to amass majority parliamentary support; typical government duration declined; & elections were called more frequently; (3) but all this was somewhat true before, & on Mairs three dimensions (degree & stability of alternation & range of parties allowed government access), Denmark had been & remained quite open b. Italys Electoral-System Change & Party Electoral-Support Volatility in 1994 (1) total electoral volatility of 37.2%, an Italian record & higher than almost any West European election 1885-1989 (2) many new parties & most old ones reconstituted (and renamed): virtually no party in 1994 parliament had same name or form as in 1987 (3) polarization changed radically also as both communist PCI becomes more acceptable left-wing PDS & far-right MSI becomes a (purportedly) reformed AN (4) But, by Mairs definition, party-system change has occurred only if the pattern & type of government alternation changes accordingly. In this case, seems that it is: (5) New systems major players seem to be coalescing into coalitional left-right camps as opposed to the old DC-dominated centrist multiparty coalitions: [ANLegaForza]Centro[PSIGreensPDS]Rifondazione, + Rete c. Canadian Electoral Watershed in 1993 (1) total electoral volatility was 42%; the Conservatives got only 16% of vote (lowest since 1949) & 2 seats (down from 169 the parliament before), Reform & Bloc Qubcois, two new parties, gained many seats (2) The new government, though, was the Liberals, thus, pending the next change of government, the alternation pattern has been preserved: Conservative <> Liberal d. Ireland: government-alternation pattern changes in 1989 & 1993 (1) Prior to that, alternation had the pattern: (Fianna Fil) <> (Fine Gael + Labour): FFs refusal to enter coalition was an electoral strategy aimed to keep itself the only party able to offer single-party government & which served to keep Labours government options to one: allying with FG (2) in 1989, FF abandoned the strategy, choosing to ally with the new rightish liberal party, Progressive Democrats, rather than go into opposition as it had before when it lost a majority (3) in 1993, they further erased any claim to credibility of the old strategy by allying with Labour (formerly always an opposing party) to form a new coalition government (4) Clearly, party system has changed, but electoral volatility was only 7.8% in the 1989 election 2. Mair concludes that electoral systems can be frozen by three factors: a. Social structure (the Lipset & Rokkan model) b. Institutions: the electoral system & the organizational efforts of established parties c. [the new factor:] the structure of party competition & government formation (1) closure of party systems helps freeze electoral alignments (2) which implies that changes in the party system, as defined by Mair, can destabilize electoral alignments (3) Examples: (a) Italy: PCI becomes PDS, MSI becomes AN ==> new electoral alignments (b) Ireland: FF chooses coalition with PD in 1989 & then Labour in 1993 ==> new electoral alignments 3. Mairs Summary of the Argument: the structure of competition, & the structure of competition for government in particular, may impose a major constraint on voter choice & hence may act to stabilize electoral alignments. In this sense, voters are not simply expressing preferences for individual parties [or candidates]; rather, albeit not always to the same degree in different party systems, & this in itself is an important source of cross-national (and cross-institutional) variance, they are also expressing preferences for potential governments. & in much the same way that a shift in the range of parties on offer can act to undermine established preferences, so too can a shift in the range of governing options, & hence a shift in the structure of competition, act to undermine established preferences & promote instability (pp. 103-4). 4. Finally, Mair perceives some generally shared trends suggesting an increasing openness of party systems a. Many places have seen an increasing range of acceptable governing parties in the past two decades b. Many have also seen an increasing set of coalitional permutations govern in recent times c. Increasing party-system openness <==> increasing electoral instability ==> increasing uncertainty at multiple levels Size & Growth of Welfare State (specifically transfers): some pictures

b.

PARTY SYSTEMS (Part 2) (Mair in LNN, GLM ch. 7-10, Powel ch. 5, L&E ch. ch. 3-5)

Page 51 of 91

LIX.

Gallagher, Laver, & Mair, Inside Political Parties, ch. 10 in Representative Government in Modern Europe. The Standard Basic Party Organization 1. Members of parties belong in local, geographically based units called branches 2. Branches usually have some role in selecting candidates & send delegates to the partys annual conference, which is in principle the final decision-making body 3. The annual conference usually elect most or all members of the partys national executive which runs the party between conferences & adjudicates internal disputes 4. Executive usually selects, & party employs, a permanent party bureaucracy 5. The parliamentary group are the partys MPs 6. [GLM leave out, but I would add: the partys actual & potential electoral supporters] B. Partial Exceptions to the Standard BPO 1. Some parties are highly factionalized: a. i.e., divided into tendencies or groups under often highly personalized leadership b. e.g., USs, Frances, Italys, & Japans parties are or have been highly fractionalized c. Often these factions will have their own organizational structure and, in the extreme, operate nearly as parties themselves 2. In federal systemsSwitzerland, US, Canada, Germany, Australiaprovincial (state) branches often have considerable decision-making autonomy 3. Communist parties are/were usually organized on democratic centralism principles a. More centralized than democratic b. Same written structure, but the party executive has final decision-making power c. Some have argued that this may be partly responsible for the decline in membership & electoral support for most of these parties C. The usual battle lines: 1. The party activists who place heavier emphasis on the partys adherence to its ideals which prompted them to join it in the first place 2. The party legislators who, as parliamentarians, often must compromise ideologically to get elected, to enter government, & to pass legislation 3. The partys actual & potential electoral supporters, & perhaps the members if it is a mass-membership party so many are not activists in this sense, judge this battle. LX. Membership A. Most voters for a party, even those who regularly vote for one party, are not members B. Complications in calculating membership & comparing it across countries & parties 1. Some parties simply do not know themselves how many members they have 2. Even if they know, some are reluctant to reveal their true membership a. Parties have obvious reasons to inflate their count b. Branches have similar reasons to inflate their report to the conference or executive 3. Some parties have affiliated organizations whose members are automatically party members (e.g., UK Labour & the trade unions) 4. The rules in some political systems encourage party membership (e.g., US primaries) C. Comparison of membership levels & trends (see Table 10-1, p. 245) D. Socio-demographics of party members 1. Working class a. Less represented among party members than among voters (esp. if exclude automatic membership via unions) b. This is increasingly true as one goes up the party hierarchy. c. This is becoming increasingly true over time also (theyre becoming still less well-represented) d. All this is true even in left parties (where white-collar public-sector workers becoming more important in many places) 2. Women & minorities under-represented also, but trends are in opposite direction. E. What members do: 1. Elect delegates to conference & thus indirectly choose the party leadership 2. Mobilize the electorate [like what?], esp. at election times 3. Used to be the case that being a party member was like belonging to a separate subculture, with its own newspaper, clubs, social events, & services available to members only. This has & continues to fade. [Why?] LXI. Important Decisions within the Party A. Types of party: Cadre v. Mass Parties (Duverger) B. Resolution of disputes & conflict within the party: over policy & over spoils (e.g., lotizzazione). Specifically differ over... 1. Party Manifesto [define] & Party Program [define] 2. Selection of the party leadership: often, but not always, synonymous with the... 3. Selection of the partys candidates & candidate order [this is the most important thing] LXII. Candidate Selection A. Except in Finland, Germany, & Norway, where its national law, how candidates are selected is decided by parties themselves B. In all European countries, candidates are chosen by party members, not by ordinary party voters (as in US primary system) C. Two basic methods of candidate selection 1. The convention system [define] (p. 255, first paragraph): a. Party members in each constituency choose delegates to a local nominating convention, which picks the candidates b. From as few as 1-2% of party members to (rarely) more than 33% participate in this selection process 2. The national-executive system [define] (p. 255, second paragraph): a. In most places, national executive has right to veto local candidate-selections, but rarely exercises it b. In a few (France, Italy, Malta, Portugal), the executive actually chooses the candidates directly, sometimes with local right to challenge that selection 3. In Greece, the leader of the party (esp. PASOKs Papandreou), chooses the candidates personally; Spain & perhaps Portugal have strong A. Page 52 of 91

but lesser roles for the very central leadership also Consequences of the selection processes 1. Incumbents have an extremely high probability of being re-selected in all systems [why?] 2. In all systems, except maybe UK, candidates always have some local ties 3. In multi-member district systems, the candidate list is almost always balanced [Define. Why?] 4. Party activist, & esp. party-leadership, control of selection increases party discipline. E. Sets up another trade-off: strong parties (disciplined, responsible) v. democracy within the party LXIII. Parties & Money: unlike in US, in most democracies almost all of candidates money comes from the party [==> discipline] A. Sources of party income 1. Internal Sources a. Membership fees (about 1/4 the total) b. MPs salary (up to 10% of salary is requested to be re-donated) c. Fund-raising events d. Party shops, banks, & newspapers (though many of these lose money now) 2. External Sources a. Interest-group backing (in many cases there are no, or only ineffective, laws requiring party disclosure funding sources) b. Public assistance (money & benefits in kind, money usually in approximate proportion to electoral strength) B. Arguments for or against state-funding 1. For (hard to gather evidence, but available evidence generally supportive) a. Reduces reliance on interest-group funding b. Evens playing field as not all supporters are equally able to support parties financially c. Parties are essential to democracy & so perform a public service & so should be publicly funded 2. Against (evidence weak to contrary) a. Public funding supports the status quo by funding existing parties only b. Public funding will reduce party incentive to raise their own money & thus to recruit new activist members c. Campaigns are becoming increasingly costly, so public funding will become increasingly costly LXIV. Trends & the future: A. Duverger expected a trend toward Mass Parties & contagion from the left (in terms of party organizational form) B. Epstein suggested that the rise of mass media & the increasing cost of campaigning would make Mass Parties less the ideal than parties looking for a few big donors (corporations, unions, etc.): contagion from the right C. Kirchheimer suggested that with the rise of the middle class, the success of the welfare state, and, Epstein added, with the rise of individualism, the type of party best suited to competition would be the catch-all party D. Sjblom & Rantala go further to claiming that, increasingly, party members were a liability, not an asset, & that parties would increasingly rely on strategists & TV-time, not members LXV. Final thoughts: A. One might join a party for three reasons (Clark & Wilson) 1. Material: desire to gain some tangible reward (patronage, a position) 2. Solidarity: desire for social contact & sense of comradeship (party as a club of like-minded folk) 3. Purposive: desire to advance certain policy goals B. What we think about party memberships depends in large part on what we think trends are in the typical saliency of these three goals & what we think the impacts are of having members predominantly of one or the other ambitions Party Systems & Election Outcomes chapter 5 of Powell, Contemporary Democracies LXVI. Strong (weak) party systems contribute to good (poor) democratic performance A. Disagreement on how to recognize such strength 1. Partly reflects differing emphasis on participation, stability, & order in defining good democratic performance 2. Partly reflects differing theories about what leads to such performance B. One line of thought emphasizes the value of a de-fractionalized, two-party, centrist party system that will aggregate citizens resources behind governmental majorities responsive to citizen pressures. C. Another approach emphasizes an expressive, mobilizing system of parties that will pull all major factions in the society into its representative, democratic decision-making institutions, co-opting dissent & accommodating demands that might otherwise turn to violence (p. 74) LXVII. Three broad types of systems advocated A. Aggregative party systems 1. Characteristics a. Two or a small number of parties, each capable of achieving legislative majority b. Centrist tendencies, not too stark alternatives c. Parties support cuts across social groups 2. [Examples] 3. [Pros & Cons] B. Representative party systems 1. Characteristics a. More parties representing a broader range of groups b. Clearer representation by these parties of those specific groups c. Accommodating leadership, willing to compromise, form coalitions d. Majorities avoided e. Participatory 2. [Examples] 3. [Pros & Cons] D. Page 53 of 91

Responsible party systems 1. Characteristics a. Two parties, alternation in (majority) governments b. Parties with closer ties to (large) social groups than aggregative systems c. Centralized, disciplined parties representing clearly distinct, somewhat ideologically distant alternatives 2. [Examples] 3. [Pros & Cons] D. Notes: 1. Almost no one advocates fractionalized system with weak ties to social groups 2. All agree that degree of fractionalization & type of alignments with groups the keys; disagree & whats desirable therein 3. All agree that extreme, anti-system parties detrimental, disagree on what to do about them or, more accurately, about the forces that might support them: a. Exclude them b. Co-opt them by allowing them a party role like any other 4. Most agree that electoral volatility a sign of a weak system a. Symptom of citizen dissatisfaction b. Make stable government & policy difficult LXVIII. Summary of hypotheses in Table 5.1 (p. 79) LXIX. Summary of Empirical Findings in Tables 5.6 & 5.7 Political Parties; Political Parties & Social Structure; Party Systems: ch. 3-5 in Lane & Ersson, Politics & Society in Western Europe LXX. Political Parties (ch. 3 in L&E) A. Parties intermediate between social cleavages & government structure 1. Express & organize political demand & support in relation to social cleavages 2. Recruit & select government decision-makers 3. They are the organized collectivities capable of intentional action in promoting their constituencies interests B. Two core themes in the study of political parties & party systems: 1. Degree to which there has been change or not in the patterns of electoral support, government formation, & policy programs 2. Degree to which parties have been, are now, or will in the future perform their intermediary roles C. Lane & Ersson begin their study of parties by distinguishing 10 party families, 5 structural & 5 non-structural [Can you list most of these, & give an example of each type?] 1. Structural parties: strongly socio-economically determined groups of supporters (e.g., by class, religion, ethnicity, sector) a. Religious Parties: L&E include CDs, but most would not (counting them as Center-Right); this because L&E distinguish by nature of supporters rather than party ideology as others do b. Ethnic Parties c. Rural/Agrarian Parties d. Socialist (and Social Democratic) Parties e. Communist & Left-Socialist Parties 2. Non-Structural Parties: weakly to non-socio-economically determined supporters a. Liberal Parties b. (Secular) Conservative Parties c. Discontent or Protest Parties d. Far-Right Parties e. Environment or Ecology Parties D. The Party-System Stability Hypothesis 1. Threats to parties existence: a. Electoral volatility b. Internal dissension c. These often imply trade-off: appeal to new voters vs. maintain loyal support of core constituency 2. Party goals/types are (at least) two-fold a. Governing parties: seek (1) Electoral success & (2) Policy impact b. Protest- or dissent-expressing / ideology- or belief-articulating parties c. Again, a trade-off: all parties are or hope to be at least some of each, but often contradictory 3. Party survival in the face of social-structural change: a. Successful parties are those successfully adapt to soc-structural change in making these trade-offs b. Adaptability is a function of (1) Organizational structure of the party & (2) Governmental role of the party 4. Summary of party & party system evolution in W. Europe a. The key parties in most countries usually date from 1900 or before, which is prima facie evidence of stability, however, party splits & organizational transformation not rare b. Rural, liberal, conservative, & socialist parties reach parliamentary representation by WWI in most places c. Religious, ethnic, fascist, & communist parties come on the scene mostly in the interwar era d. Fascist-party support all but disappears with WWII, neo-fascists re-emerging perhaps e. Socialists gradually come to be considered increasingly legitimate players in government f. Protest & ecology parties begin to emerge in the 70s for the most part 5. Long-run key to party relevance is, of course, electoral support. Fluctuations may be ridden out, steady secular decline in support demands action or precipitates demise. E. Party Organization 1. Membership & membership ratio (members/voters) Page 54 of 91

C.

LXXI. A. B.

C.

LXXII.

Centralization & unity (integration) External organizational connections (segmentation) Parties range from a. Greenssmall membership, decentralized, & weak external organization, to b. Socialists/Communistslarge membership, centralized, strong external organization c. [Flesh these out & consider where others might lie; So what? How might party-organization matter?] Match of the party system to social structure (L&E chapter 4) Focus here is on how well socioeconomic variables predict voters choice of party; what else might we be concerned about in the match of party system to social structure The research design: 1. Divide countries into sub-national units, then ask if social structure of those subunits can explain the percentage share of different parties votes in those units 2. Social-structural explanators a. Industry: proportion of population employed in industry b. Agriculture: size & type of farms in the region c. Affluence: average income of the region d. Religious structure: percent catholic or percent church-attending e. Ethnic structure: share of population belonging to particular ethnic/linguistic group The results 1. First step is to compare inter-regional variation across countries & by party family a. Define: Variance, CV, SCV, & >2 b. Tables 4.2 & 4.3 2. Second step is to see how much of this variation can be explained by structural factors a. Table 4.4 b. Appendix 4.2 3. Comparing the relative importance of different social-structural factors in different countries a. Table 4.5 b. Specifically, the materialist/post-materialist dimension: Tables 4.5 & 4.6 Party Systems (L&E chapter 5): descriptive data

2. 3. 4.

VOTING and CHOOSING REPRESENTATIVES (Part 1) (P ch. 6; Franklin in LNN)

LXXIII. Citizen participation in politics A. Much wider array of participatory options than simply voting: 1. Party organization & mobilization 2. Campaigning, lobbying, protesting B. Voting is distinctive, & especially worthy of study because... 1. Most common form of participation 2. A central feature of democratic society 3. Selects leadership, policy-makers 4. Fundamental source of citizen influence on those policy-makers C. The central questions are: 1. What explains who participates, &, more broadly, the cross-country, cross-time variation in the amount & type of participation? [Show figure plotting voter participation across countries & over time.] a. Relatively consistent variation in the amount of voter participation, though with some secular decline, within nations over time b. This suggests strongly that there are some common features explaining a general decline & some set of broadly stable differences in national characteristics which explain cross-national variation 2. How well do various democratic institutions & party & political systems achieve the democratic ideals of representation, responsiveness, & accountability? LXXIV. Explaining cross-national variation in voter participation (Powell ch. 6, part 1) A. The basic analytic framework is citizens cost-benefit analysis of the decision to vote 1. Presumably people vote if they perceive the net benefits to exceed the net costs of voting 2. Benefits: a. Policy influence: probability ones vote changes the electoral outcome times the policy differences expected to be produced by such an outcome change: probability(PolicyLeft-PolicyRight) b. Social, civic, & psychological benefits: social gathering, participating in a competition, feeling of being a good citizen, etc. 3. Costs: a. Polling fees or fines for not voting (compulsory voting) b. Time required to vote (opportunity costs) B. Powell emphasizes two core elements: 1. Legal (constitutional) & institutional differences across countries 2. Party system differences across countries C. Legal (constitutional) differences 1. Compulsory voting 2. Registration laws 3. Powell finds strong evidence in favor (Table 6.1) 4. Other legal/constitutional differences? Page 55 of 91

Party system differences 1. Strong party-group linkages facilitates turnout a. Strong, continuing expectations about parties policies & the interests of social groups creates easily identifiable choices for voters b. The same features facilitate parties identification of probable supporters for electoral-mobilization efforts c. The external (non-party) organizations themselves can & do aid in this mobilization d. The hypothesis: the strength of the alignment between parties & groups [measured by the] degree to which individuals partisan preferences can be predicted from knowledge of their demographic characteristics... should be positively associated with voting participation (Powell p. 116). e. Powell finds strong evidence in favor (Table 6.2)Switzerland is a major exception (we return to that) f. An implication: if this type of structural voting & these types of party-group ties are on the decline, this may (partially) explain the secular decline in voting participation. 2. The degree of competitiveness of the party system should also contribute to voter participation: Elections expected to be close & meaningful (in the sense of possibly producing important policy changes) should produce the most turnout a. Powell supplies some anecdotal evidence: (1) India 1967 & 1977: Congress Party first challenged & then defeated by a more united opposition (2) Norway & Denmark: late 60s & 70s challenge to Soc.-Dem. dominance produced rises in voter participation (3) Turnout fell in Canada in periods when Liberals gained apparent electoral dominance b. He also notes that three democracies without change in executive control had lower than average turnout among nations with their institutional & other party-system characteristics: Switzerland, Japan, & India; but this difference seems to show only when a very marked lack of alternation arises E. Powells Path Model of Voting Participation (Figure 6.1) LXXV. Relation of Voting Participation & other Systemic Factors with Political Turmoil & Social Strife (Powell ch. 6, part 2) A. Types of political violence 1. Major distinction usually drawn between a. Collective protest: large numbers of citizens expressing discontent, loosely organized or disorganized b. Internal war: smaller groups of well-organized & trained individuals systematically attacking adversaries c. Population size positively related to first, relatively unrelated to the second 2. Riots: illegal, largely unorganized actions by a substantial number of individuals which result in property destruction or assaults a. Goals of riots vary: (1) riots stemming from clashes with authorities, usually based on anti-government grievances (remonstrance) (2) riots stemming from intergroup antagonisms (confrontation) (3) riots lashing out against impersonal, frustrating circumstances such as train delays or food shortages (frustration) b. riots are clearly episodic: come in burstsfeed on each other then absent for some time, etc. (1) idea here is to identify consistent properties of countries that affect their propensities to have riots (as opposed to explaining particular riots or proximate events that cause particular riots) (2) poverty, inequality, & even ethnic/religious divisions not terribly systematically related to riots in themselves, suggesting that something systematic intervenes between sources of tension & expression thereof in riots c. Powell argues that... ...are key political variables intervening to affect propensity to riot (1) support for extremist parties, and (2) degree of fractionalization of the party system d. Citizen turmoil appears best contained by providing all groups direct representation in legitimate political structures & avoiding simple, broadly encompassing majorities directly chosen by the electorate (paraphrasing Powell, p. 125) (1) Caveat: fractionalization can possibly facilitate extremist parties which have opposite effect (2) Even net of this effect, though, fractionalization seems to reduce propensity to riot (3) [See Table 6.3 & Figure 6.2] 3. Peaceful Protests: organized events in which substantial numbers of citizens participate in an endeavor to win the support of others or of the authorities for a political cause a. Protests strongly associated with population size b. Theyre also more common in nations that also experience considerable rioting (or vice versa) c. Require individual & organizational skill & resources ==> more likely in developed than developing nations d. Otherwise similar to riots e. [See Figure 6.3] B. Voting participation negatively related to riots & violence: substitutes? LXXVI. Electoral Participation (Franklin, ch. 8 of LeDuc, Niemi, & Norris) A. Three approaches to explaining voter participation 1. Individual-Resources Approach: what individuals bring to political processknowledge, wealth, timedetermines voters 2. Political-Mobilization Approach: heightened awareness of duty or civicness is inculcated via media, parties, groups 3. Instrumental-Motivation Approach: people respond to the perceived costs & benefits of voting; they aim at political efficacy in terms of influencing policy (at least in concert with like-minded individuals); [see framework above] B. Franklins broad argument: the third of these has been unduly neglected as an explanation of, esp., cross-national variation in voter participation, specifically he focuses upon a small set of legal & political-system variables as explanators: 1. salience of elections [define] 2. compulsory voting 3. proportionality of electoral system 4. postal voting 5. weekend/workday polling C. The first point Franklin stresses is that across-country variation in voter participation is much greater than that between individuals with different Page 56 of 91

D.

D.

characteristics [See Table 8.2] 1. US has greatest differences between voting participation rates of types of individuals, but otherwise individual-level differences account for much less of the cross-country variation than country-level or system-level variables 2. This suggests two things: a. a large part of the explanation for variation in voter participation must come from differences in political system not from differences in demographics or average individual characteristics (e.g., education level) b. it follows, moreover, that if one aims to increase (or decrease) voter participation, changes in political institutions, environments, & rules are the way to go, esp. since most individual-level characteristics are less manipulable Comparing the approaches with these considerations in mind: 1. Individual-resources approach a. argument: people participate who have the time, money, & intellectual where-with-all to do so b. evidence: (1) within country, this is broadly true, though perhaps not as much as youd think (2) across countries: theres not that much variation on average levels of these, & they dont explain much 2. Political-mobilization approach: a. argument: groups, organizations, & media can be very effective in exhorting, inculcating participatory ethic, etc. b. evidence: (1) no doubt that group efforts are effective (esp. driving people to polls, register, etc.), media effect more debatable (2) even less cross-national variation in mobilizational efforts than in individual characteristics 3. Instrumental-motivation approach: a. argument: people vote based on perceived costs & benefits (largely policy-influencing efficacy of voting) of voting (1) both other approaches indirectly address the instrumental motivations of citizens (a) education, experience, time, & money might well increase perception of efficacy & lower costs of voting (b) mobilization is largely an effort to increase those perceived benefits & reduce those perceived costs (2) the I-M approach subsumes these two & also allows the election contest itself be a factor driving turnout b. Specific hypotheses Franklin derives from this approach (1) policy relevance of election & voters role therein increases voter turnout [salience] (a) is (executive) policy-making power clearly, unmitigatedly at stake? [federalism, division of power, etc. reduce turnout] (b) is the election expected to be close? (particularly in my district?) (c) is the outcome likely to affect public policy? (d) are there large perceived differences between the perceived alternatives? (2) an electoral system which reduces the proportion of wasted votes increases turnout (3) anything that makes it more difficult (costly in time, money, opportunity costs) to vote, reduces turnout (a) compulsory voting (b) registration laws (c) weekend/workday polling (d) advance and/or postal voting (e) number of days polls are open (4) voter fatigue: increasing the number of distinct elections reduces turnout [correlates very highly with no on 1a] c. Evidence [review regression analysis]: (1) The individual-level results are given in Table 8.3 (2) Country-level results given in Table 8.4 4. [Discussion about other types of participation in this context]

VOTING and CHOOSING REPRESENTATIVES (Part 2) (Norpoth, Bashevkin in LNN)

LXXVII. Helmut Norpoth, The Economy, in LeDuc, Niemi, & Norris Comparing Democracies A. Economy is a concern that almost everywhere bonds electorates & governments as tightly as Siamese Twins joined at the hip (p. 300) 1. Incumbents presiding over good economic times are more likely to win than those presiding over bad 2. [Therefore] Governments try to bend the economy to their will 3. Economic adversity can kill an infant democracy & vice versa good times strengthens its foundation B. The Vote FunctionReward-and-Punishment Calculus 1. Common or Conflicting Interests? 2. Symmetric or Asymmetric Reward & Punishment? 3. Personal or Sociotropic Concerns? 4. Retro- or Prospective Evaluation? C. The vote function, whatever form, varies across country-times. Why? 1. Clarity of responsibility (Powell & Whitten) 2. Alternatives for Discontentideal is 2-party system [US version or UK version?] 3. [any other ideas?] D. Norpoth demonstrates simply that economic voting exists [show picture] 1. No claim to explain it all [a bit on ceteris paribus, stochastic/systematic world, controlling for...] 2. Still, what exactly does vote function look like, when & where E. Butler & Stokes (1974)Valence vs Position Issues 1. Economic Prosperity is a valence issue 2. How to Pursue It is often a position issue 3. To the degree economy is a valence issue, parties & candidates compete for votes not by offering different promises so much as claiming Page 57 of 91

to have a better plan to achieve those promisesthis suggests perception of economy is as or more important than actual economy (though surely former based on latter) 4. Trade-offs? Sometimes growth, employment, & low inflation cannot be pursued simultaneously a. Valence model says folks unaware of this, so Government constantly in trouble with electorate for failing, G must dissemble, misdirect, etc. as much as possible on this score b. Position Model, contrarily, leads to left-right partisan politics a la Hibbs (1977) F. Uninformed voters? 1. Perfectly rational 2. (Partisan) ideological filters? Reputation 3. Is the electorate still right-on-average (unbiased)? If so, may be irrelevant that uninformed... 4. Why doesnt it pay anyone to inform? ==> incentives to distort, voter must decide whom to believe, etc. G. Simple Reward & Punishment Model (Key 1964; Kramer 1971; Kiewiet & Rivers 1985): If incumbents performance satisfactory, relect; if unsatisfactory, vote for challenger 1. If true, notice how this drastically reduces the informational requirement of voters a. Reduces political universe to one actor b. Retrospectiveno need to evaluate promises of novel & unknown c. Need only decide if you like outcomes, not evaluate policies 2. [But do voters act this way? (And if some do, some dont, who would? Why?)] a. good reason to believe public better able to evaluate incumbent [note on risk aversion] b. but what about Party Reputationscant ignore opposition H. Assymetric Reward & Punishment? 1. Perhaps...certainly economy tends to be bigger news & therefore perhaps bigger issue when bad 2. The negativity effect at first found some evidence, but more systematic, broader study laid it to rest I. Partisan Reward & Punishment? 1. Facts perceived through partisan-tinted glasses 2. Partisan reputations tend to attach certain economic issues to certain partiese.g. Republicans anti-inflation & Democrats antiunemployment a. Votes do seem to respond somewhat in this waywhen inflation a big issue, more R vote b. Notice the really strange incentives this gives a partisan looking for re-election [no evidence they respond to that incentive, perhaps maintaining the reputation is too important] J. Personal or Sociotropic Evaluation? 1. Pocket-book voting has little or no empirical support in US (Kinder & Kiewiet 1979, Kiewiet 1983) a. Ethic of self-reliance in the US? b. [Other reasons?] 2. Howevere, it doesnt show up in UK, FR, GE, IT, or SP either! (Lewis-Beck 1988)==> US self-reliance ethic not a very good explanation a. Empirically, sociotropic concerns seem to dominate personal concerns in estimated voting functions. b. [So, are people really that altruistic? If believe not, you must explain why the evidence seems to go the other way...] (1) [specification errorpeople are selfish after all, & evaluate based on personal economic conditions, but these are poorly measured] (2) [Possibility voters attribute macro performance to politicians, not personal; therefore evaluate accordingly... Though they care about personal, they believe the politicians only effect on the personal is embodied in their effect on the aggregate. or maybe the community?] K. Retro- or Pro-spectiveare politicians rewarded for good past performance or for expected productivity if re-elected? 1. Rationally it should be the latter of course a. In popularity functions, that appears to be the case b. In voting functions, it is much less clear, if anything voters are retrospective 2. Implications for political manipulation of the economy Bashevkin, Sylvia, Interest Groups & Social Movements, in LeDuc, Niemi, & Norris, Comparing Democracies, Sage (1996). LXXVIII. Organized groups: A. At core of democratic debate: 1. Attacked by critics as selfish & destructive of the democratic process 2. Celebrated by supporters as crucial, constructive actors in that process B. Synonyms: 1. Pressure groups, interest group, lobby, & organized interest 2. Social movement organization, social movement C. Definitions: 1. Interest Groups (IGs): a. Bashevkin: what IGs & social movements share is a set of norms, beliefs, or values that keep the interest intact. These shared orientations are the glue binding together the group constituency, leading it to act as a coherent entity...collectivities that have as their basis a shared outlook, identity, or framework of reference (135) b. Truman (1951): IG refers to any group that, on the basis of one or more shared attitudes, makes certain claims upon other groups in the society for the establishment, maintenance, or enhancement of forms of behavior that are implied by the shared attitudes... (pp. 135-6) c. Wilson (1990): organizations, separate from government though often in close partnership with government, which attempt to influence public policy. 2. Social Movements (SMs): a. Like IGs except more pro-change, more challenging, vis--vis the institutional status quo (138) Page 58 of 91

Protest groups that mobilize their adherents in more system-challenging ways than do IGs or parties Kitschelt: movements tend to follow the experience of marginalization Blumer (1951): a collective enterprise to establish a new order of life, meaning a fundamental cultural shift that will either be pursued or, conversely, opposed by SM adherents e. Less formal than interest groups, more heterogeneity of views typically within SMs than IGs (largely because former are typically broader) 3. The substantive distinctions between party-IG-SM are not always terribly clearly drawn LXXIX. Interest Group Functions: A. Social roles of IGs 1. Truman: these [shared attitudes] afford participants frames of reference for interpreting & evaluating events & behavior 2. Like parties, function as social networks as well as political organizations B. Electoral roles of IGs: 1. Evaluate competing candidates 2. Help fund candidates 3. Mobilize voters C. Policy-making roles of IGs 1. Influence or pressure government decision-makers, but not seek office (distinguishes them from parties) 2. Provide information to policy makers a. Where voters/citizens stand on issues b. Information (possibly biased) about the issue itself D. Wilson: IGs provide institutional linkages between the government (state) & major sectors of society LXXX. Substantive questions regarding interest groups & movements A. Why do some groups find it more difficult to organize than others? 1. Heterogeneity/homogeneity of membership: both socio-economic homogeneity & Unity of views 2. Selective benefits (Olson) 3. Size of the groups (Olson) 4. Coercion/suasion B. What explains the difference in means employed by different groups with shared interests? 1. Means: Some form parties [e.g., Green parties]; Others IGs & lobbies [e.g., Sierra Club]; Others adopt politics by other means [Greenpeace] 2. Hypotheses: a. Bashevkin: groups form strategies based on their institutional & ideological settings [Such as?] b. History? diffusion of innovation c. Intensity of views? d. Power configuration of alternative & orthogonal views? C. What explains the general rising number & importance of interest groups? 1. Decline of parties [Define] 2. Education & affluence ==> greater resources & abilities ==> lower relative cost to mobilize groups 3. Advances in communications & transportation technology ==> increased efficiency in organizing D. What explains variation across space & time in the activities of interest groups? (differing uses of Voice?) 1. Change over time in the US a. Traditionally interest groups in US politics lobbied elected politicians in national & state capitals, increasingly they... (1) Work in coalitions to lobby bureaucrats (2) Present their arguments in court (3) Make their cases directly to public via media interviews & advertising (4) Try to influence elections as well as the traditional attempt to influence policy at legislative stage b. What might explain these changing activities? (1) Increasing professionalization of interest groups (2) Societal changes? [Such as...] (3) Voting-behavior changes? (4) [Other ideas?] 2. Why is lobbying of bureaucrats relatively more prevalent in French than US system? a. Differing nature of electoral competition? b. Differing impact of bureaucratic implementation on the net effect of policy? c. [Other ideas?] E. Social Movements are more willing to use extra-political, even illegal, means to seek their ends than IGs. Why? [First of all: is this tautological?] 1. Do the mobilization needs of IGs & SMs differ? 2. Does the nature of IGs & SMs goal differ? 3. Differing opportunities to achieve their ends? [consider Kitschelts point above about SMs resulting from marginalization] F. What explains differing success of various groups of interests in achieving their ends? 1. Tarrow (1983): breadth & flexibility of the groups tactical repertory 2. History, e.g., the novelty of the movement or groups aims or tactics? 3. Institutional differences in the political process? [Such as...] LXXXI. Various Models of Group Activity A. The Power Elite Model: Mills (1956), Hunter (1953) 1. Small elite, sometimes covert elite, controls government, limiting opportunities for average citizens political participation Page 59 of 91

b. c. d.

B.

C.

D.

E.

F.

How does such a minority dominate the majority in a democracy? a. Size, homogeneity, similar interests of these groups serve its coherence b. Relative social, political, & economic power of the elites enables them to dominate the uninformed, excluded masses The Pluralist Model: Truman (1951), Dahl (1961) 1. Struggles among competing groups are the essence of democratic politics; the major role of government is to establish & maintain orderly relations among them 2. The vectors of political inputs image: multiple (vector) forces push on government, the policy produced is the resultant (vector) of those forces 3. Multiple points of access for the diverse interests implies that group participation in public policy-making is virtually assured 4. What, then, explains the extent to which various collective demands are met? a. Emphasis on (relative) internal group characteristics, especially organizational resources. (1) Socio-economic status, media access or control, education, & skill of group leaders (2) Internal organizational cohesion (3) Strategic social position (placement in political issue space relative to other interests) b. Later work has fleshed out four key variables determining the policy success of a group (1) Group Resources: financial & human resources positively related to policy influence (2) Group Cohesion: organizational control, unity, & discipline positively related to influence (3) Leadership & Group Expertise: policy & administrative knowledge always helps (4) Group Representativeness: favorable strategic positioning and/or comprehensive group standing in a policy domain c. Basic assumption of the pluralist model: open, fair competition naturally favors groups with highest net score on these factors [X.B.4.b.(1)-(4)]. Critics of Pluralism: Bachrach & Baratz (1963) & Schattschneider (1960) 1. Non-decisions & agenda setting 2. Government & bureaucracy are interested not neutral actors in the process 3. Insiders v. Outsiders: some groups excluded 4. Basically reject image of government as neutral arena in which group interests fairly compete & policies neutrally emerge. State- & Institution-Centric Model: Katzenstein (1986), Krasner (1978), Evans, Skocpol 1. Three key variables: state preferences, capacity, & autonomy 2. Policy-makers can & do develop preferences which differ from those of constituent groups 3. Relative degree of autonomy of state from societal groups determines degree to which they may differ 4. Capacity is the ability of the bureaucratic state to execute its actions & thus logically depends on its ability to concentrate, coordinate, & exploit governmental resources The Corporatist Model: (Schmitter 1979, Lijphart) 1. system of interest representation in which the constituent units are organized into a limited number of singular, compulsory, noncompetitive, hierarchically ordered & functionally differentiated categories, recognized or licensed (if not created) by the state & granted a deliberate representational monopoly within their respective categories in exchange for observing certain controls on their selection of leaders & articulation of demands & [supporting activities] Schmitter (1979) (149). 2. Interest groups operate as integral parts of, rather than external influences on, government activity 3. State is at the apex of the corporatist system, organizes a hierarchy within which interests operate regarding the government & other interests [keep the example of Governmentunified labor movementunified employer organization in mind] 4. Note the incorporation of insider groups (with perhaps some exclusion implied) so that favored groups play a structured & significant role in policy-making 5. Competition among groups is much more limited than in pluralist systems while states role is more assertive & sustained 6. Contrast liberal or societal corporatism from state corporatism [from below v. from above] All actual country-times could be characterized to some degree by each of these models of politics, but... 1. US among the most nearly pluralist 4. Japan a hybrid of statist & corporatist 2. Britain a bit more statist 5. Germany a bit more purely corporatist 3. France a lot more statist 6. Sweden & Austria most nearly corporatist

2.

LXXXII. Empirical Findings A. Bashevkin claims that business groups are generally the most influential in most systems 1. Note the difficulty of establishing any such claim 2. [If business groups most influential, why are they among the most critical of government policies?] B. Most of the work confirms the importance of internal characteristics of groups to their success C. Most also find that the political opportunity structurethe attitudinal or institutional environment within which the groups operateare key to strategies, success, etc. Thus, influence rests on the combination of internal assets & external situation [a pretty banal conclusion at this level of generality] D. Groups adapt rapidly to their institutional circumstances: they seem to go where it matters in their system (e.g., Congress then courts in the US), subject to their abilities to go there. E. Evidence from transitional democracies indicates that democratic group competition does not emerge simply from the establishment of rules allowing it 1. Pre-transition traditions are key 2. Pre-transition group organization also central LXXXIII. Discussion: How could we evaluate the common claim that special interests as opposed to average citizens & their elected representatives hold power?

VOTING and CHOOSING REPRESENTATIVES (Part 3) (Dalton ch. 9-11; LeDuc in LNN)
Page 60 of 91

LXXXIV. Partisanship & Electoral Behavior [Dalton, ch. 9] A. The funnel of causality predicting vote choice 1. A heuristic model for organizing theory about electoral choices [Figure 9.1, p. 197] 2. From structural factors which are temporally & psychologically removed from voting through explicitly political factors & beliefs more proximate to the vote 3. Model is predictive success: in fact, modern version of it predict voter choices better than individual voters themselves B. Party Identification 1. Definition: a long-term, affective, psychological identification with ones preferred political party a. To be analytically relevant, it must be distinct from immediate voting preferences election by election (1) Creates a difficulty as a comparative concept where the term partisanship means different things in different contexts & the distinction between party affinity & vote choice is variably clear b. Still, voters do seem to hold some partisan allegiances which endure & which influence other perceptions & opinions c. The psychological basis of stable party alignments; it influences other attitudes & behaviors C. Acquiring partisan identification (Party ID) 1. Acquired largely by an early ageprimary school ==> parents & family environment plays a central role a. Correlation of parent & child party ID is very high b. Via explicit reinforcement or subconscious internalization 2. Accumulated experience of party support itself reinforces party ID a. Party ID tends to strengthen with age, & specifically with continuous electoral support of party b. It follows that unstable party systems tend to weaken party ID 3. [Figure 9.2, note E. Germany, French 60+ group] 4. Summary: Partisanship formed early; Reinforced through life; May change in response to experience, but not very easily; Party ID is more stable than opinions on almost any set of issues except perhaps certain moral issues D. Partisanship is more stable than voting preferences in Europe as in the US, but less so, Why? 1. Type of voting: candidate or party 2. Frequency of voting: more frequent voting facilitates distinguishing between partisan loyalty & vote E. The impact of partisanship 1. Provides cues on how to evaluate events, policies, proposals, candidates a. More broadly relevant than class or religion in this respect b. Reliance on partisanship in this way could be thought of as satisficing: a mental short-cut c. Some interesting findings: (1) Washington Post fictitious act experiment (2) Partisanship also colors expectations for national economyis it rational for partisanship to affect economic expectations in this way? d. Has strongest impact on things like candidate image & government-performance evaluations 2. Party ID also mobilizes: increases propensity to vote & to perform other political activities 3. Converses notion of Party ID Vote producing normal vote share for parties: vote-share when all other factors balance 4. Party ID more easily counteracted in more prominent elections (e.g., presidential elections) than less (off-year) 5. Partisanship offers clear & low-cost information voting cue: notice that it can only serve in this capacity if the party label provides information on likely policiesfailure to maintain partisan reputation will eventually weaken party ID among its supporters F. Partisan Dealignment 1. Partisan alignment, like (social) structural voting, is declining, esp. in US (where was largest) but universal a. Some contention that many non-partisans are closet partisans consistently voting for one party, even so, number of partisans is definitely declining b. Decreased turnout might be linked to this c. More delayed voting decisions (made more closely to election day) d. Number of floating voters increasing 2. Explanations: a. Crises & events: Vietnam, Watergate, scandalsbut if this is universal, these explanations weakened b. Many traditionally party functions now done by other org.s & groups, including even candidate selection somewhat c. Mass media changespersonalization of campaigns d. Leading parties inability to deal with modern issuesKeynesian Welfare State crises, new narrower issues, etc. e. Changes in the characteristics of the contemporary voter (Daltons preferred explanation) G. Cognitive Mobilization & Apartisans: two types of independents & two types of partisans 1. Define cognitive mobilization 2. Four types of citizens according to degree of cognitive mobilization & of partisan identification [Figure 9.5] a. Apoliticals (trend 0) b. Ritual Partisans (trend ) c. Cognitive Partisans (trend + ) d. Apartisans (trend ++ ) 3. Implications of rising Apartisans a. Less consistent voting patterns b. More issue-voting c. Greater demand for responsiveness to public opinion d. Press for expansion of citizen-input opportunities 4. Socioeconomic trends producing this voting-behavior trend: parties actions can bolster or resist this trend [discuss] a. Young, educated, post-materialists Page 61 of 91

b. Information revolution [...] Consequences of these trends [discussion] 1. Electoral volatility 2. Electoral unpredictability 3. These may produce increasing responsiveness of elected officials to public opinion 4. Increased divided government / fractionalization 5. Rise in independent candidates & flash parties 6. Factors further along the funnel should be increasing in importance relative to earlier part [...] LXXXV. Issue & Image Voting (Dalton, ch. 10) A. As structural & partisan voting decline, relative importance of issue voting is rising B. Issue voting long seen as key to sophisticated, rational voting. Logical requirements for issue voting: 1. Citizens are interested in the issue 2. Citizens hold an opinion on the issue 3. Citizens know candidate or parties stands on the issue 4. On most issues, most citizens do not meet these requirements ==> early conclusion that voter incapable of issue voting C. Overlapping issue publics: 1. Define 2. Resolves the matter: most voters meet all three requirements on at least one issue D. Classification of issue types [discussion] H. Content of Issue Time Frame Retrospective Position Issues Policy appraisal Performance Issues Performance evaluation Candidate/Party Attributes

Attribute voting Prospective Policy mandate Anticipatory judgement E. Left/right attitudes as a super-issue F. Image as an issue: particularly as it pertains to voter evaluation of how candidates will behave in certain circumstances 1. Interesting that actually the more-educated are more likely to evaluate on personal characteristics 2. More moderate but growing impact outside the US 3. Coming at the end of the causal funnel, as image-evaluation does, creates some analytical problems a. Somewhat predetermined by earlier factors b. Hard to distinguish image evaluation from the vote choice itself G. Opinions on ... issues ... influence vote choice 1. Old Politics issues (economy, moral, etc.) strongly influence vote choice a. Economic issues still at center of political debate; as a group, the category of issues on which the largest share of citizens meet the issue-voting requirements b. Recent revival of economic-policy controversy (Thatcher, Reagan, Kohl), but debate is less structured by socio-economy & partisan predispositiondebate between opposing cognitive partisans & opposing apartisans 2. New Politics issues only modest influence a. Is this because these issues secondary, & could that explain party reluctance to take clear, strong stands on them? b. Or does the causality run the other way: parties vague stands on these issues leave these issues offering voters little reason to choose parties based on them? 3. Foreign-policy issues also typically weak a. Except in crises, attracts the primary attention of relatively few voters b. Party differences on foreign policy also tend to be lesser H. Each issue typically has small aggregate impact because not all salient to all voters, but together issues key to aggregate outcome & individual or small numbers of issues usually key to individual voters I. Performance-based voting: personal v. sociotropic, retrospective v. prospective [see Norpoth for fuller treatment] J. Dalton refers to all of this trend toward personal, issue-centered politics as citizen politics 1. [Define]: a. Decline in long-term structural determinants of vote & partisan attachment/vote b. Growth of short-term determinants of vote: issues & image c. Among a voting population increasingly cognitively mobilized 2. Change in the nature of economic voting illustrates what hes talking about [bottom of page 235, top of 236] 3. Possible pros of citizen politics [discuss] a. Increasing policy implications of electoral results b. May make candidates & parties more responsive 4. Possible cons of citizen politics a. Over-burdening, inconsistent, non-aggregated demands b. Citizens w/o skills for sophisticated, individualistic voting, voting as atomized groups without party or social cues , may be easy prey for demagogues 5. The possible roles of television in campaigns illustrates this two-edged sword LXXXVI. Political Representation (ch. 11): how well do elected representatives correspond to their constituencies? A. Collective correspondence: dispersion between elite & public distributions of attitudes B. Dyadic correspondence: dispersion between constituency & representative attitudes 1. Edmund Burkes Two Models of Constituency Representation Page 62 of 91

a. The trustee model of constituency representation b. The delegate model of constituency representation c. [Figure 11.1 on these two paths of constituency-representative correspondence] 2. The party-government model: dispersion between party & supporter distribution of attitudes a. Necessary characteristics for responsible party government (1) competition between 2+ parties (2) parties have distinct, recognizable policy-orientation differences (3) voters recognize these differences & care about them (4) parties act to important degree as units b. Constituency-representative agreement low-to-non-existent outside the US. Why? c. Bidirectional causality: voters influence party positions, & parties attempt to sway their voters d. [Figures 11.2-11.5] 3. General patterns of elite-citizen correspondence a. Elites tend to be more extreme than their supporters b. Strong constituency-representative/party-voter agreement on economic & social policy, less on foreign policy c. Clarity of party positions on issues tends to increase voter-party correspondence d. Fractionalized party systems tend to increase voter-party correspondence C. Impact of citizen politics on political representation 1. On one hand, issue-oriented, sophisticated electorate may encourage parties & candidates responsivity to voter interests 2. On other hand, partisan dealignment & candidate-centered politics may weaken representation built on party government D. Two distinct patterns of democratic representation have emerged [discussion of pros & cons] 1. The representative-constituency form a. Pro: Greater responsiveness to the different interests of particular districts b. Pro: Potentially more open to interests & representation of minority groups c. Con: Increases difficulty of monitoring & controlling government (as opposed to individual representatives) actions d. Con: Encourages campaigns to focus on personalities & district service, rather than policy & ideological orientations 2. The parliamentary, party-supporter model a. Pro: party votes as a unified bloc so political responsibility for government is more clearly established b. Con: may produce rigidity & resistance to change for this reason; parties may be very responsive to established clientele, but new social groups & minorities may have difficulty gaining such representation E. Empirical research (continues) to show that policy does in fact respond to voter preferences; the overwhelming broad conclusion from what empirically has happened is that democracy works (reasonably well) at least in terms of representativeness & policy responsiveness. So why all the public outcry to the contrary? LXXXVII. Lawrence LeDuc, Elections & Democratic Governance, ch. 14 in LeDuc, Niemi, & Norris, Comparing Democracies A. Focus on linkage between elections & actual policy-making process B. Elections clearly decide many important matter, such as: 1. Establish norms for the legitimate exercise of power 2. Determine/limit how the country may be governed 3. Determines, sometimes indirectly & often in a limited way, who will govern 4. Provide at least a rough guide to what policies will follow C. However, in less than of democracies is the effective executive directly elected, & also there must share governing power with legislatures & with various non-elected bureaucracies & agencies D. Again, an important continuum can be constructed... 1. From those countries where the link between the electoral outcome & the determination of those who will hold power is very direct & simplemajoritarian systems 2. To those where the link is indirect & convolutedconsensus systems. 3. Presidential systems are typically intermediate. E. Do Elections Determine Policy? 1. Agendas (election manifestos, platforms, & programs) v. Policy Mandates (the message sent by the election) 2. Do parties do what they say? [i.e., do they generally implement their election promises?] a. It varies, but, broadly, YES. b. Degree to which they can, however, depends on a number of things (1) necessity of compromise once in office [varies by the political system how much compromise is necessary] (2) circumstances can change once in office (3) some promises may be contradictory or prove practically impossible 3. Do parties do what voters want them to do? [i.e., are election mandates fulfilled?] Much harder to say because a vote is one signal, what millions of voters intend by it is impossible to decipher, parties make their own interpretation & usually fulfill that. 4. Do parties matter? I.e., does policy depend on the partisanship of electoral winners & of governments? YES a. First, are parties recognizably (by voters & experts) distinct in ideologies & policies advocated & enacted? YES & YES b. Matter in simple left/right ways: left governments consistently enact different policies than right ones c. Matter in complicated ways: the number of partisan actors, polarization of governments, who holds which office, etc. F. Again, elected representatives & governments generally do what they say they will; parties have predictably different tendencies in office which voters recognize & vote based upon; so why is everyone so convinced government is unresponsive?

GOVERNMENT FORMATION and DISSOLUTION (Part 1) (P ch. 7; L&E pp. 240-6; GLM ch. 12; L&S pt. I)

Page 63 of 91

LXXXVIII. The Centrality of the Politics of Coalition A. Single party majority government is relatively rare in democracy, esp. in PR systems B. Within parliamentary systems, even legislative majority for 1 party does not really abrogate politics of coalition 1. Single party government more often minority than majority 2. Potential alternative governments (perhaps coalitions) remain in background for next time even when majority exists now 3. Even in single-party government systems, more often than not politics of coalition plays out within party, amongst its factions (Japans LDP, e.g.) C. Even in presidential systems, one can think of day-to-day politics of assembling a majority behind individual pieces of legislation as repeated plays of coalition politics D. Once formed, a government is continually subject to parliaments ability to unseat itmore politics of coalition E. The conduct of policy-making in democracy is virtually all about the forming & maintaining of coalitions LXXXIX. Two Central Questions Regarding the Politics of Coalition: A. The partisan composition of government & the allocation of cabinet ministries therein B. Durability / instability of governments XC. Executive Stability (Powell, ch. 7) A. Powell has already established two important facts about executive performance for us 1. Constitutional factors dominate in the explanation of executive performance (ch. 2-4) a. Presidential systems ==> durable, but often minority, governments b. Majoritarian Parliamentary systems ==> durable, usually majority, governments c. Representative Parliamentary systems ==> unstable, occasionally minority, governments 2. Presence of strong extremist parties is the key feature of representative parliamentary systems that produces these negative executiveperformance outcomes (ch. 5) B. Summary so far: That social-structural factors work through these constitutional & party-systemic factors has been demonstrated & accounts for the relative weakness of social-structural factors per se in explaining executive performance, now fill in remaining link which is type of governments formed by these party systems operating in these constitutional & socio-structural environments. C. The Arguments: 1. Parties Motivations are Two-fold (notice already that parties are being treated as unitary actors) a. Desire to participate in & control (current) policy-making (1) This is both for intrinsic (policy-seeking) reasonsparties care about what policies are madeand (2) for instrumental (office-seeking) reasonsparties & party members wish to remain in government (this requires voter support which presumably depends on policies) b. Parties are also motivated by considerations of the ramifications of current actions for future ability to participate (esp. the future electoral ramifications) 2. What sorts of governments will form then? a. This implies that parties will tend to form governments which... (1) ...are as small as possible to maximize seats/party for spoils & influence reasons (2) ...contain as little ideological dissension possible so as to make policies as close to their ideal as possible (3) ...have sufficient parliamentary support to pass necessary initiatives (4) These goals will often conflict, but knowing them will also often suffice for us to make sound predictions of what sorts of governments will form & endure 3. Evidence on Government Formation a. Rule #1: When a single-party majority can form it will (1) 23/25 such cases (1965-75) follow the rule (2) The two that didnt were oversized governments formed after a strong pre-electoral alliance produced a surprise majority for one of the allies b. Neither Oversize & Minority governments are terribly rare, which is partially explained by... c. Rule #2: Coalitions are usually formed of ideologically connected parties (1) addition of a connecting party to a potentially separated coalition lowers the average ideological distance between parties in government & thereby perhaps facilitates compromise (2) lessens the degree to which the separated parties are seen (by its members & electoral supporters) as moving from their ideals in joining the coalition (3) [parties between can add legis. support to coalition w/o requiring much/any further policy-compromise] (4) however, adding any party, intermediate or not, does require a further division of the spoils] d. Rule #3: Where Minimum-Winning-Connected-Coalition Governments possible, they are usually formed e. Rule #4: Non-majority governments & non-minimum-winning-connected governments are formed most frequently where extremist party support is high (>15%) 4. Legislative Circumstances behind certain types of Minority Government a. Crisis Caretakers: Intense conflict among polarized major players ==> passive minority government as temporary measure (Powell calls these caretakers but the term can be misleading; it has narrower meaning in some contexts.) b. Pre-election minorities: often formed after an election in which a system that usually produces single-party majorities fails to produce one. The larger of the two holds office (assumed temporarily) while new elections are prepared which is assumed will fix the situation. If not, the minority government typically will try to serve a term with outside support c. Active Minority Government: minority party or coalition secures outside support from other legislative groups: these vary in formality (tacit agreement to written, detailed agreements) & in fixity (same outside support all the time to potentially different set of supporting parties each time a law is passed) 5. Typical Government Duration by Government Type: a. Expectations: Page 64 of 91

D.

XCI. A.

B.

C.

(1) Single-party-majority most durable, the minimum-winning-connected (2) Over-sized & unconnected governments should be less durable (3) Minority governments should not be terribly durable; among these caretakers least durable, pre-election a bit more durable, & externally-supported most durable b. Evidence: (1) SPMajG most durable, MWC not much less durable, externally-supported Min can be reasonably durable, all others less durable, esp. crisis caretakers (2) Surprisingly, though, over-sized coalitions were least stable of the bunch c. Figure 7.2 summarizes Powells schematic view; strong extremist presence again key factor Comparing Presidential & Parliamentary Governments 1. Presidential systems all have durable executives relatively independent of party system & political & social environment, but executive frequently lacks legislative majority (divided government common to pres. systems, not a uniquely US phenom.) 2. Presidential systems associated with weaker parties (less unitary actors) perhaps precisely because the party can vote against the executive without risking its dissolution 3. Weak party discipline, however, is what makes executive remaining in power w/o a legislative majority feasible 4. At the same time, it means presidents do not enjoy anything like the degree of legislative control that PMs do 5. If anything, separated powers work to require more complex negotiations between executives & legislatures, esp. since these will usually have different electoral constituencies (President national, Legislators sub-national) Building & Maintaining (Parliamentary) Government (GLM ch. 12) The Party Composition of Governments 1. [Always keep in mind from here on out that...] elections are still paramount in that they determine the legislative weight of the party actors in the government-formation drama 2. Politicians motivated by some combination of (a) fame & power & (b) desire to influence public policy, & these lead to different implications in government formation a. Pure office-seeking motivation leads to minimum-winning logic: dont share seats any more ways than you have to ==> oversized governments are a puzzle from this perspective b. Pure policy-seeking motivation leads to ideological-compatibility consideration ==> median-party govt c. Combination of these motivations suggests: (1) Minimum-winning-connected (Axelrod) (2) Minimum-ideological-range-winning (de Swaan) Minority (and oversize) governments are not at all uncommon in parliamentary systems so any reasonable coalition theory must explain them, all reasonable such theories seem to rely on policy motivation of parties 1. Minority Governments, Hypotheses: a. Strom: as ability of parties to influence policy from opposition increases, frequency of minority government should increase [Figure (mine): data strongly supportsn.b. the Opposition Influence measure used here comes from Laver & Hunt, not from Strom, so helps his case] b. Luebbert: increased role of interest groups & organizations outside parliament (esp. corporatist-type policy-making) decreases necessity of being in government to influence policy so should increase minority governmentsmight explain Scandinavia, but what of Germany, Austria, & Italy? Need multivariate analysis to consider this c. Laver & Shepsle: increased policy divisions among opponents leads to a decreased ability to form an alternative government which should enable minority-government formationespecially centrist governments can do thiswell see this argument up close later as we go through Laver & Shepsle 2. Oversized Governments, Hypotheses: a. Governments of national unity: several of the observed over-sized governments appear immediately postwar, only occasionally after that & then usually short-lived & appear in crises b. Policy agenda in some situations may require super-majorities (Belgium notable in this regard) not all apparently oversized governments are really surplus then c. Laver & Shepsle: extra parties may be included for signals they send [to whom?] about govts policy stance d. Luebbert: dominant party(s) in coalition may want surplus minor parties so no one smaller-party ally is a veto actor e. [surplus governments may be especially likely when a minimum-winning coalition bridges a smaller intermediate party because little further policy-compromise would be necessary to increase the legislative strength of government. May add to Luebberts argument in particular] f. [party discipline? Surpluses in Italy, e.g., may have stemmed from need of extra insurance support] 3. Some examples of government formation a. Figure 12-1: Ireland after the January 1993 election (minimum-winning) b. Figure 12-2: Sweden after the October 1991 election (minority coalition) c. Figure 12-3: Italy after the July 1987 election (surplus majority) The Allocation of Cabinet Portfolios 1. Cabinet government: a. Cabinet serves as central committee for decision-making: by time bills referred from cabinet to parliament for full-parliamentary votes, its a done deal. In fact, most decisions do not require direct legislative assent. b. Vote of confidence theoretically gives parliament control of cabinet, & ultimately it does, but only if parliament willing to risk govt collapse over the issue in question ==> much latitude to cabinet members, esp. within their department competence (more on this in L&S) c. Two questions: how many seats to each party? & which portfolios to which parties? 2. How many seats allocated to each party? a. Proportional to share of legislative majority: 90% of the variation in party share of cabinet seats is explained by party share of the Page 65 of 91

D.

governments legislative majority (e.g., Govt has 40 seats in parl, party Y has 10 ==> party Y gets 1/4 of cab. seats). b. One of strongest empirical relationships in polysci even though its not constitutionally required anywhere 3. Which seats to which parties? a. Tendency for median party on each policy dimension represented by a department to get that portfolio b. Tendency for parties most centrally interested in that policy dimension to get that portfolio c. Cabinet ministers are agenda-setters in their policy area & have considerable latitude in policy-making (1) thus who gets what is central to the policy position of the government & its likely policy outputs (2) thus cabinet reshuffles more important than often realized (3) thus ideological positions of cabinet ministers are perhaps only credible signals of governments policy stance in that policy area (4) thus shifts of power within parties can fundamentally impact the relations between parties The Stability of Parliamentary Governments 1. What exactly counts as a change of government is not universally agreed. Issues: a. Agreed that change in partisan composition of cabinet is a change in govt & that change of PM is change of govt. b. Not quite all agree that every govt forming after an election is a change in govt even if the same make-up as before c. Nor do all agree that if govt resigns & then is reformed with same PM & party make-up, that this is a new govt d. Still, such ambiguity is not so common that it makes tremendous difference. It makes some difference, but not so much that we cannot proceed from here 2. Explaining Duration: Three sets of factors a. Features of Government Itself (1) Number of parties in the government (fractionalization) (2) Ideological distance between parties in government (polarization) (3) Majority/Minority status of government b. Features of the Political Environment (1) Party System: Larger, more diverse party systems reduce government duration because smaller shocks can change what coalition is an equilibrium (2) Other factors? [Powell: extremist party support] c. Shocks, events, & external circumstances [journalistic accounts invariably focus on these] (Warwicks work) (1) Worsening economic conditions, esp. unemployment (2) Unemployment especially bad for ctr-right; inflation especially bad for socialist-led d. [N.b. the difference between journalism/politics & political science: political science is not interested in specific circumstances & events that bring down specific governments, but systematic features which make governments more or less stable in general. Analogy: not interest in what particular spark caused some specific fire but what conditions make fires more likely.]

GOVERNMENT FORMATION and DISSOLUTION (Part 1) (P ch. 7; L&E pp. 240-6; GLM ch. 12; L&S pt. I)
XCII. XCIII. A. B. XCIV. A. B. The Politics of Coalition is central to parliamentary democracy especially but also, more abstractly, to all democracy Two Central Questions Regarding the Politics of Coalition: The partisan composition of government & the allocation of cabinet ministries therein Durability / instability of governments Executive Stability (Powell, ch. 7) Powell has already established that constitutional factors, esp. presidential v. majoritarian parliamentary v. representative parliamentary systems, dominate in the explanation of executive performance (ch. 2-4). Presence of strong extremist parties also a key feature (ch 5). Summary so far: That social-structural factors work through these constitutional & party-systemic factors has been demonstrated & accounts for the relative weakness of social-structural factors per se in explaining executive performance, now fill in remaining link which is type of governments formed by these party systems operating in these constitutional & socio-structural environments. Parties Motivations are to participate in & control (current) policy-making 1. This is both for intrinsic (policy-seeking) reasonsparties care about what policies are madeand 2. for instrumental (office-seeking) reasonsparties & party members wish to remain in government 3. Parties are also motivated by considerations of the ramifications of current actions for future ability to participate (esp. the future electoral ramifications), so the two are not necessarily so much in conflict 4. This implies that parties will tend to form governments which... a. ...are as small as possible to maximize seats/party for spoils & influence reasons b. ...contain as little ideological dissension possible so as to make policies as close to their ideal as possible c. ...have sufficient parliamentary support to pass necessary initiatives Evidence on Government Formation 1. Rule #1: When a single-party majority can form it will 2. Rule #2: Coalitions are usually formed of ideologically connected parties [at least 3 reasons] 3. Rule #3: Where Minimum-Winning & Connected-Coalition Governments possible, they are usually formed 4. Rule #4: Non-majority governments & non-minimum-winning-and-connected governments are formed most frequently where extremist party support is high (>15%) Expectations of Government Duration by Government Type: 1. Single-party-majority most durable, the minimum-winning-connected 2. Over-sized & unconnected governments should be less durable 3. Minority governments should not be terribly durable; among these caretakers least durable, pre-election a bit more durable, & externallysupported most durable 4. Evidence: only surprise is that over-sized coalitions were least stable of the bunch 5. Figure 7.2 summarizes Powells schematic view; strong extremist presence again key factor Page 66 of 91

C.

D.

E.

F.

XCV. A.

B.

C.

D.

XCVI. A.

Comparing Presidential & Parliamentary Governments 1. Presidential systems all have durable executives relatively independent of party system & political & social environment, but executive frequently lacks legislative majority (divided government common to pres. systems, not a uniquely US phenom.) 2. Presidential systems associated with weaker parties (less unitary actors) perhaps precisely because the party can vote against the executive without risking its dissolution 3. Weak party discipline, however, is what makes executive remaining in power w/o a legislative majority feasible 4. At the same time, it means presidents do not enjoy anything like the degree of legislative control that PMs do 5. If anything, separated powers work to require more complex negotiations between executives & legislatures, esp. since these will usually have different electoral constituencies (President national, Legislators sub-national) Building & Maintaining (Parliamentary) Government (GLM ch. 12) The Party Composition of Governments 1. Pure office-seeking motivation leads to minimum-winning logic (Riker) 2. Pure policy-seeking motivation leads to ideological-compatibility considerations 3. Combination of these motivations suggests: a. Minimum-connected coalitions that win (Axelrod) b. Minimum-ideological-range coalitions that win (de Swaan) Minority (and oversize) governments are not at all uncommon, so any reasonable coalition theory must explain them 1. Minority Governments, Hypotheses: a. Strom: as ability of parties to influence policy from opposition increases, frequency of minority government should increase [Figure (mine): data strongly supports] b. Luebbert: increased role of interest groups & organizations outside parliament (esp. corporatist-type policy-making) decreases necessity of being in government to influence policy so should increase minority governments c. Laver & Shepsle: increased policy divisions among opponents leads to a decreased ability to form an alternative government which should enable minority-government formationespecially centrist coalitions can do this 2. Oversized Governments, Hypotheses: a. Governments of national unity: appear in crises (usually short-lived) b. Policy agenda may require super-majorities (e.g., Belgium) not all apparently oversized governments are really surplus c. Laver & Shepsle: extra parties may be included for signals they send [to whom?] about govts policy stance d. Luebbert: dominant party(s) in coalition may want surplus minor parties so no one smaller-party ally is a veto actor e. [surplus governments may be especially likely when a minimum-winning coalitions bridge smaller intermediate parties because little further policy-compromise would be necessary to increase the legislative strength of government.] f. [party discipline? Surpluses in Italy, e.g., may have stemmed from need of extra insurance support] Explicit definition of the different single-dimensional government-formation concepts & some examples 1. From my notes below [revised from the previous notes] 2. From GLM ch. 12: a. Figure 12-1: Ireland after the January 1993 election (minimum-winning) b. Figure 12-2: Sweden after the October 1991 election (minority coalition) c. Figure 12-3: Italy after the July 1987 election (surplus majority) The Allocation of Cabinet Portfolios 1. Cabinet government: a. Cabinet serves as central committee for decision-making: by time bills referred from cabinet to parliament for full-parliamentary votes, its a done deal. In fact, most decisions do not require direct legislative assent. b. Vote of confidence theoretically gives parliament control of cabinet, & ultimately it does, but only if parliament willing to risk govt collapse over the issue in question ==> much latitude to cabinet members, esp. within their department competence c. Two questions: how many seats to each party? & which portfolios to which parties? 2. How many seats allocated to each party? a. Proportional to share of legislative majority: 90% of the variation in party share of cabinet seats is explained by party share of the governments legislative majority (e.g., Govt has 40 seats in parl, party Y has 10 ==> party Y gets 1/4 of cab. seats). b. One of strongest empirical relationships in political science even though its not constitutionally required anywhere 3. Which seats to which parties? a. Tendency for median party on each policy dimension represented by a department to get that portfolio b. Tendency for parties most centrally interested in that policy dimension to get that portfolio c. Cabinet ministers are agenda-setters in their policy area & have considerable latitude in policy-making (1) thus who gets what is central to the policy position of the government & its likely policy outputs (2) thus cabinet reshuffles more important than often realized (3) thus ideological positions / policy-position reputations of cabinet ministers are one of few, perhaps only, credible signal of governments policy stance in that policy area (4) thus shifts of power within parties can fundamentally impact the relations between parties The Stability of Parliamentary Governments (What exactly counts as a change of government is not universally agreed. Still, such ambiguity is not so common that it makes tremendous difference. It makes some difference, but not so much that we cannot proceed from here) Explaining Duration: Three sets of factors 1. Features of Government Itself a. Number of parties in the government (fractionalization) b. Ideological distance between parties in government (polarization) c. Majority/Minority status of government 2. Features of the Political Environment a. Party System: Larger, more diverse party systems reduce government duration because smaller shocks can change what coalition is Page 67 of 91

B.

an equilibrium b. Electoral volatility: for a given party-system structure, greater electoral volatility means larger typical shocks to parties support c. Other factors? [Powell: extremist party support] 3. Shocks, events, & external circumstances [journalistic accounts invariably focus on these] (Warwicks work) a. Worsening economic conditions, esp. unemployment b. Unemployment especially bad for ctr-right; inflation especially bad for socialist-led Evidence: from King, Alt, Burns, & Laver

Single-Dimensional Coalition-Formation Concepts Definitions: Minimum Winning Coalition: a coalition whose member parties control a parliamentary majority & which contains no party which is unnecessary to do so; i.e., a coalition obtaining a majority without any surplus parties. Need not be minimum-connected winning coalition or minimumideological-range winning coalition. Minimum-Connected Winning Coalition: a coalition whose member parties are ideologically adjacent, without ideologically intervening parties that are not also in the coalition, & which contains no strict subset of parties which are sufficient for a majority. Need not be a minimum winning coalition, but will be a minimum-ideological-range winning coalition. Minimum-Ideological-Range Winning Coalition: a coalition whose member parties obtain a majority without containing any other majority within its left-to-right span. I.e., a coalition that spans no more ideological space, left-to-right, than necessary for majority. Need not be a minimum winning coalition or a minimum-connected winning coalition. Single-Dimensional Coalition-Formation Concepts: Examples from the German Parliaments (Bundestagen) Elected in 1994 & 1998

Page 68 of 91

1994 Partisan Composition of the German Bundestag Left to Right PARTY L/R IDEOL. SEATS PERCENT PDS 3.3025 30 4.5% Greens 4.055 49 7.3% SPD 6.605 252 37.5% FDP 11.26 47 7.0% CDU/CSU 13.975 294 43.8% TOTAL 672 100.0% Minimum Winning Coalitions (Need 337): CDU/CSU + FDP 341 50.74% CDU/CSU + SPD 546 81.25% CDU/CSU + Greens 343 51.04% SPD + Greens + FDP 348 51.79% n.b. PDS is a "dummy party" Minimum-Connected Winning (Need 337): CDU/CSU + FDP 341 50.74% SPD + Greens + FDP 348 51.79%

1998 Partisan Composition of the German Bundestag Left to Right PARTY L/R IDEOL. SEATS PERCENT PDS 3.3025 36 5.4% Greens 4.055 47 7.0% SPD 6.605 298 44.5% FDP 11.26 43 6.4% CDU/CSU 13.975 245 36.6% TOTAL 669 100.0% Minimum W inning Coalitions (Need 335): SPD + Greens 345 51.57% SPD + FDP 341 50.97% SPD + CDU/CSU 543 81.17% CDU/CSU+FDP+Greens 335 50.07% n.b. PDS is a "dummy party" Minimum-Connected Winning (Need 335): SPD + Greens 345 51.57% SPD + FDP 341 50.97%

Notes: Both of the minimum-connected winning coalitions are also minimum winning coalitions in this case. This does not have to be so. If, for example, PDS were between the greens and SPD, then: In 1994: Greens + PDS + SPD +FDP would be minimum connected winning but not both MWC and MCW This would leave only CDU/CSU + FDP as both MWC and MCW In 1998: Greens + PDS + SPD would be minimum connected winning but not both MWC and MCW This would leave only SPD + FDP as both MWC and MCW Minimum-Ideological-Range Winning (Need 337): Range 2.715 CDU/CSU + FDP 341 50.74% 7.205 SPD + Greens + FDP 348 51.79% Minimum-Ideological-Range Winning (Need 335): Range 2.55 SPD + Greens 345 51.57% 4.655 SPD + FDP 341 50.97%

The following are not Minimum Ideological Range Winning Coalitions because they span one or the other coalitions above, but it may nonetheless be informative to note that these two minimum winning coalitions span greater ideological range than the above two. 7.37 CDU/CSU + SPD 546 81.25% 7.37 SPD + CDU/CSU 543 81.17% 9.92 CDU/CSU + Greens 343 51.04% 9.92 CDU/CSU+FDP+Greens 335 50.07% Notes: It is possible to construct examples where the minimum ideological-range winning coalitions are not minimum winning coalitions. It's, also possible to construct examples where minimum ideological-range winning coalitions are not minimum connected-winning. The only necessary logical connection is that all minimum connected winning are also minimum ideological range winning. The opposite is not true. E.g., imagine 2 new parties, A&B, between SPD&Greens in 1998. Also imagine PDS and its 5.4% of the seats split among A&B. Then the following coalitions are minimum-ideological-range winning: SPD + Greens SPD + B + Greens SPD + A + Greens SPD + A + B + Greens Of these 4, only SPD + Greens is minimum winning, and only SPD + A + B + Greens is minimum connected winning, but it is still impossible to draw any minimum connected winning coalitions that is not also minimum ideological range winning.

Final Example: A hypothetical Germany with parties arrayed left-to-right thus: PDS Greens SPD A B FDP CDU/CSU R and with seats allocated thus: 1.4% 7.0% 44.5%2.0%1.0% 6.4% 36.62% 1% Minimum Winning Coalitions (MWC): (Riker) Greens + SPD (also MCC & MIRC) SPD + FDP (not MCC but MIRC) SPD + CDU (neither MCC nor MIRC) CDU/CSU + FDP + Greens (neither MCC nor MIRC) Dummy Parties: PDS, A, B, R Minimum Connected Coalitions that Win (MCC): (Axelrod) GreensSPD (also MWC & MIRC) Page 69 of 91

SPDABFDP (not MWC but MIRC) Dummy Parties: PDS, R Minimum Ideological Range Coalitions that Win (MIRC): (de Swaan) GreensSPD (also MWC & MCC) SPDFDP (also MWC but not MCC) SPDAFDP (neither MWC nor MCC) SPDBFDP (neither MWC nor MCC) SPDABFDP (not MWC but MCC) Dummy Parties: PDS, R Whats a game tree?

T h e

P r is o n e r s ' D ile m m a
(1 0 C o n fe s s y e a rs , 1 0 y e a rs )

(fre e , 1 5 S ta y C o n fe s s (1 5 C o n fe s s S ta y S ile n t (5 S ta y P r is o n e r O n e P r is o n e r T w o S ile n t S ile n t

y e a rs )

y e a rs , fre e )

y e a rs , 5

y e a rs )

Page 70 of 91

How do we model actors preferences & decisions?

Indifference Curves for (Simple) Euclidean Preferences

Foreign Policy

Economic Policy
Given Laver & Shepsles assumptions (party discipline, ministerial discretion, separable preferences, etc.) How do we represent the set of possible governments?

Page 71 of 91

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE

CE CA

DE DA

BE BA

EE
EA

Foreign Policy Position

AA

AD

CD

DD

BD

ED

AB

CB

DB

BB
BC

EB

AC Left

CC

DC

EC

Left

Economic Policy Position

Right

Page 72 of 91

GOVERNMENT FORMATION and DISSOLUTION (Part 2) (L&S)

Laver & Shepsle, Making & Breaking Governments XCVII. Background (Here Begins Part I) A. The Centrality of Government Formation & Dissolution 1. The essence of parliamentary democracy is the accountability of the government (syn: cabinet, executive, administration) to the legislature where it must retain a majority. (Legislature, in turn, is accountable to voters.) a. Votes of investiture b. Votes of (no) confidence 2. Nonetheless, cabinet retains wide latitude as long as they remain in office a. Agenda setting b. Control over administrative departments (syn: cabinet ministries, portfolios) 3. Government formation & dissolution lies at the apex of a set of links between voters, legislators, parties, etc. 4. Change of government, thus, implies at least a potential change of policy (evidenced, e.g., by stock & financial markets sensitivity to such changes) B. The Role of Analytic Modeling 1. First fundamental premise is that it is possible to make general statements about the politics of building & maintaining a government, & that such general statements can give us valuable insights into the political processes involved. 2. The advantages of deriving such general statements formally: a. Allows one logically to hold a set of variables constant while manipulating certain key variablesthe model may thus be used as a discovery tool (1) this allows for counterfactuals (within logical, modeled world, if not actual, empirical world) (2) being more precise & explicit about models lets one check (mathematically) her intuitions & tease out further analysis of the problem at handsometimes even discovering counter-intuitive features which may nonetheless prove correct b. Analytical models are also a systematic aid to the construction of empirical modelsthe analytical model can less ambiguously tell us what to look for & how to look for it. 3. Logical abstraction (i.e., simplifying assumptions) are essential to analytic modeling. However, they are no less essential to any other mode of theorizing. Theory is logical abstraction. The main point analytic modelers can (correctly) press is that it is almost always best to be as explicit as possible about these unavoidably necessary assumptions. 4. Points about mathematical (in this case, set theory & geometry) modeling [a brief sermon]: a. Math is simply a language, specifically a symbolic language for the expression of logic. Anything which follows logical laws can be expressed mathematically, & anything which cannot be expressed mathematically does not follow logical laws. Non-mathematical logic is an oxymoron. b. Like other foreign languages, one becomes fluent in mathematical expression by using it & being immersed in it. No one is born understanding it, & no one is born without the ability to learn it. (In my opinion, & from my experience, I suck at math usually means Ive had crummy math teachers in the past, sometimes I havent been immersed in math enough before to follow you there. Ive never met anyone who convinced me that they truly sucked at math though Ive met many who originally thought they did & a few who eventually convinced me they were just too stubborn to make an honest effort.) c. If you have an interest in grad school in the social sciences (not to mention the physical sciences of course), you will be immensely well-served by taking some math: calculus, linear algebra, & probability & statistics in particular. d. I dont advise taking any of the various math escape hatches liberal arts allows. It closes too many roads to leave college without even intro calculus & the like. Trust meI took such an escape-hatch & paid later for it. Im still paying. [OK, enough sermon] C. Rationality & Analytic Modeling in the Social Sciences 1. Only human agency effects outcomes in the social world which we are studying; therefore every social-scientific model must begin with establishing what motivates the actors & how they make decisions a. Thus, we first posit the aims or goals (here: office & the implementation of particular policies) of the key actors (here: politicians, especially MPs) in the phenomenon were studying (here: government formation & dissolution). b. Then we must define the decision rule by which those actors choose between alternative possible actions. Here we assume actors are rational in sense that they calculate (implicitly or explicitly) the costs & benefits of various actions with regard to achieving their aims & select the course of action with the highest net benefit (lowest net cost). 2. Specifically, in this model of government formation, we are assuming: a. Potential government politicians are policy-motivated; doesnt matter whether intrinsically or instrumentally so. b. The policy aims of all the players, all the rules of the government-formation game, etc. are common knowledge to all players. c. Politicians act given this knowledge so as best to achieve their aims; i.e., their votes for or against governments are calculated to bring about their most desired cabinet as best they can. 3. Aside on rationality & mathematical modeling: rationality as a decision-rule lends itself easily to mathematical modeling because maximizing net benefits is something easily expressed as mathematical problem, but any logically immutable decision rule can, in principle, be modeled. Once again: any decision rule which could not be precisely defined in this way is incapable of producing logically consistent theory. D. Legacy: Some Important Theoretical Results which Have Come Before: Hotelling (1929), Black (1958), & Downs (1957) 1. Blacks Median Voter Theorem: If voters have single-peaked preferences defined on a single dimension (e.g., left-right), then the ideal point of the median voter is the only point which is preferred by a majority to all others. 2. Hotelling-Black-Downs Centripetal Tendency of Party Competition: applied to elections between 2 competing parties, the median voter theorem (MVT) implies strong incentives for the parties to converge toward the median voters ideal 3. Multiple Dimension Extensions: a. McKelveys & Schofields Chaos Theorems: when more than one dimension is involved, & choices between various options are Page 73 of 91

not structured in restrictive ways, it is virtually certain that policy proposals will cycle around the policy space, with no proposal majority-defeating all others. (1) This implies either perpetual flux or arbitrary outcomes when game ends (note: Arrows Impossibility Theorem). (2) It also strongly suggests that institutional restrictions on the proposal- & decision-making process are essential to non-arbitrary democracy. (Shepsles Structurally Induced Equilibrium) b. Kadane (1972) has shown that, if an equilibrium exists in multi-dimensional space, & it may not, then it must be the multidimensional median a.k.a. the Dimension-by-Dimension Median (DDM) 4. All of these results refer to policies emerging from majority-rule voting, not to governments forming subject to majority approval the process of government formation & dissolution may add precisely the kind of structure which can bring equilibria to multi-dimensional problems. (A key contribution of this volume, the authors claim, is to show that a particular conception of government formation can produce precisely that.) E. The Institutional Structure of Government Decision-Making 1. Policy decisions are made by the executive. 2. Departments, & especially their cabinet ministers, have agenda power & information, expertise, & resource advantages which give them wide latitude in directing policy within their sphere. 3. Appointment of participating ministers is the governments most credible signal of its policy intent in that area. 4. The overall policy position of the government is given by the partisan position of the politicians parties who are given the various cabinet portfolios. a. This implies that, even in multi-dimensional space, the set of possible policies for government is finite. b. Precisely this specificness, & finiteness of the set of possible government policies, given by the set of possible party combinations in the various cabinet offices, limits the chaos potential of multi-dimensional decision-making. F. The most important lesson of the volume is that... 1. The departmental organization of governmental decision-making structures the environment in which governments are born, live, & die. 2. The problem of finding & establishing equilibrium cabinets therefore differs from that of finding & implementing equilibrium policies in an unstructured environment. 3. Well find that cabinet equilibria are common & usually close to the center of the configuration of party policy-positions. XCVIII. Sketching the Building Blocks of the Cabinet-Formation-and-Dissolution Model A. Motivations: office-seeking (Downs) and/or policy-motivated (de Swaan) 1. The interaction of politicians & voters in elections brings office-seeking & policy-seeking politicians to act alike 2. So Laver & Shepsle take no stand one way or the other here, simply assume they act as if policy-motivated B. Rational Foresight & Common Knowledge 1. Politicians act (perhaps unconsciously) as if conducting (perhaps very sophisticated) cost-benefit calculations regarding their options 2. In doing so, they are capable of looking ahead: they attempt to foresee the consequences of their actions & use that to inform their current cost-benefit analysis (analogy to chess, e.g.) 3. Preferences of the competing players are common knowledge (as are the rules of the game, etc.) 4. An important tool in the analysis of such decision-making is a Game Tree [see the Prisoners Dilemma example at end of these notes] C. Parties are treated as unitary actorscabinet ministers have discretion within their jurisdictions, but they use that discretion on behalf of not against their party. (Practical implication here is that government policy in each portfolio area will be given by partisan identity of its minister.) D. Dimensionality: how many dimensions will we consider 1. Theoretically infinite number may exist, but empirically great correlation in party stances across substantive dimensions. 2. This implies that relatively few core issue areas may suffice for a reasonable description of a partys preferences. 3. Well take the competencies of core ministries (Finance, Foreign Affairs, & maybe Internal Affairs) as defining the dimensions of the policy space. E. Governments make policy, implementing the preferred policy of the party holding each cabinet ministry in that ministrys area. Such ministerial discretion comes from three sources: 1. Ministers may act on matters in their ministerial area which have not been explicitly decided by the cabinet as a whole 2. Ministers may influence which matters will actually come to the cabinet for such decision 3. Ministers may influence the content of the proposals which they or others do bring to cabinet F. Unitary Parties + Ministerial-Policy + Indivisibility of Ministries ==> Limited Number of Possible Governments (and Thus Policies) to Consider, as Given by a Lattice of the perpendicular intersections of the parties ideal points [see example at end of notes]. A pair of key assumptions at this stage: 1. A partys discretion in one portfolio is not affected by who controls other portfolios [can you think of violations?] 2. A partys preferences along 1 dimension are not affected by policies along other dims [can you think of violations?] G. All expected policy decisions are incorporated into the parties calculations in deciding whether to vote for a particular cabinet or not ==> unforeseen matters are the (only) potential sources of cabinet collapse H. The Status Quo: 1. The current government remains in office until defeated by a legislative majority. This can only happen when unforeseen circumstances change the distribution of legislative power that supported the government to begin with, or when some party or parties shift their ideological positions (unforeseen), or when some party or parties split or combine to new ideal points. 2. When a government collapses, it remains in office as a caretaker until a new government receives majority support. I. Any participant in a proposed government can veto that government by refusing to accept its proposed role ==> all governments require unanimous consent of the parties in it & majority support of all parties in legislature J. [The chronological (continuous) play of the government-formation game is given schematically in Figure 3.1, p. 52] XCIX. Working through the Model (Here begins Part II) A. Equilibrium Cabinets: a cabinet for which no actor with the power to bring down the government has an incentive to do so 1. One direct substantive implication of the concept of equilibria here is that equilibrium cabinets are expected to be stable while nonequilibrium cabinets are expected to be unstable Page 74 of 91

C.

There are two types of equilibria in general, & equilibrium cabinets can be of either sort a. Attractive Equilibria [define?] b. Retentive Equilibria [define?] 3. A core notion in the determination of whether there is an equilibrium cabinet & in characterizing it if so is that of a strong party B. The information necessary to analyze cabinet formation in the model 1. Information about parties a. Their identity (how many distinct parties) b. Their legislative weight (seats for each party) c. Their policy positions (ideological positions in the policy-space determined) 2. Information about the Policy Space & the Departmental Structure of Government a. The set of (key) relevant policy dimensions b. The set of (key) cabinet ministries & their competences (i.e., their issue domains) c. The relation between the twoassumed for most of the book to be a simple one-for-one relationship 3. The formation process (assumptions) a. Proposals can come from anyone at any time b. Unanimous internal consent is required c. Majority legislative consent is required C. Important concepts 1. Indifference Curves for Euclidean Preferences [example figure] 2. The winset & the lattice winset: a. The winset of some policy x is the set of all alternative policies preferred to x by a majority b. The lattice winset of some cabinet X is the set of all alternative cabinets preferred to X by a majority c. The only difference is that the lattice winset is concerned only with alternative cabinets which must be on the lattice d. [We can use the lattice from the end of these notes to show how winsets are found; Figure 4.1 is L&Ss example] 3. The Dimension-by-Dimension Median (DDM) in this context a. A cabinet is the DDM iff the policy associated with it is the median on each dimension (has 50%+ on either side of itself, including itself). b. The easy way to find this is to read from left to right & top to bottom (or vice versa) until you find the lattice line that turns a minority into a majority. [example again from lattice below or Figure 4.1] c. The DDM is an equilibrium government if no alternative government is in its winset, but there is no guarantee the DDM will have an empty winset (so it may not be an equilibrium) (1) the DDM is always a potential equilibrium point, so a useful place to start the analysis (2) the DDM is more likely to have an empty winset (and so be an equilibrium) when dimensionality & number of parties low d. [the example I give can have either an equilibrium DDM (empty winset) or non-equilibrium DDM (non-empty winset) depending on the distribution of legislative seats; L&Ss Figure 4.2 shows a non-empty winset DDMone in which cycling can occur] 4. Strong Parties a. Party S is strong if it participates in all cabinets majority preferred to its ideal point (so that it can veto all cabinets majority preferred to its ideal & thus perhaps secure its ideal point) b. A party may be very strong or merely strong (strong refers to a party which is either very strong or merely strong) (1) very strong: there is no cabinet majority preferred to its ideal (in which case the partys ideal must be the DDM) (2) merely strong: there are cabinets majority preferred to its ideal, but it participates in all of them & so can veto them c. [Examples: (1) a relatively even distribution of seats in our example gives a DDM with an empty winset at DD, thus party D is very strong (2) Party D may be merely strong in a less even distribution of seats in our example (3) Party B is merely strong in L&Ss Figure 4.3] d. There can be at most one strong party; L&S claim they are in fact common e. Theorems (syn: predictions, hypotheses) (1) A very strong party produces an attractive & retentive equilibrium cabinet at its ideal point (it gets all seats): (its a special case of a DDM with an empty winset)this is so whether or not the very strong party is itself large (in terms of its seats) (2) When a merely strong party exists, it will be a member of every equilibrium cabinet, & an equilibrium cabinet can be no less desirable from the merely strong partys view than the DDM f. Whether & whos strong will depend on the weights & positions of the parties (1) changes entirely external to a strong party itself can change the identity of the strong party or whether there is one (2) striking discontinuities (a) small changes in weights or positions can have large effect on whos strong & thus on cabinet formation (b) conversely, large changes in these same could have no effect on whos strong (c) it depends on how close the equilibrium is to some alternative government in the lattice which could be an equilibrium if things were just somewhat different [refer back to diagrams] g. Merely strong parties depend on their ability to continue (credibly) to veto majority preferred cabinets in which they participatethus standoffs can occur [refer back to diagrams] h. Notice the strong centripetal tendencies: (1) either the DDM is an equilibrium, (2) or a strong party can move the equilibrium toward its ideal, but strong parties tend to be toward the median Strong Parties Making & Breaking Governments A. Analyzing a case vs. analyzing the process 1. For analyzing a specific case we need to identify which, if any, party is strong 2. For general analysis of government formation & dissolution, we need to discover the conditions (e.g., distribution of legislative power, Page 75 of 91

2.

etc.) which produce them. Simplest Case: 2 Dimensions, 3 Parties, any 2 parties make a majority ==> Triangular system 1. Only the middle party can be strong, & it usually is (but doesnt have to be) 2. Middle as defined here: connect the party ideal points to make a triangle, party whose ideal point is the vertex between the two smaller legs of the triangle is the middle party 3. [L&Ss first example shows this] C. Outside the simplest case: 1. Formal analysis produces few intuitive results (the mathematical conditions for strong parties are not readily understood substantively) 2. Empirical analysis is limited because the universe of coalition-generating party systems is too small to produce inductive generalizations of this breadth 3. This leads to Simulations: a. Program a computer to find equilibrium cabinets, strong parties, etc. from inputted information about the parties, cabinet portfolios, etc. b. Then, holding some set of parameters fixed (e.g., # parties, their relative strength, & # dimensions), generate a large number of random virtual parliaments which vary some other parameter(s) (e.g., their policy positions). c. Try to find patterns in the outcomes (i.e., inductive reasoning from virtual reality) d. Simulations are NOT empirical tests b/c the program assumes the model is true to begin with, but (1) Useful for finding substantive bite of formal conclusions which may be too complicated for intuitionalways with caveat that virtual world assumes random & assumes the model (2) Often useful for various kinds of sensitivity analysis (3) [In other contextsmany statistical procedures have known properties in infinite samples (asymptotic properties), but unknown properties in limited samples ==> simulate large number of small samples to observe properties: called Monte Carlo Experiments] D. L&Ss Simulations 1. Finding frequency of strong parties under various distributions of legislative strength, numbers of parties, & numbers of dimensions (party policy-positions are varied randomly) [Table 5.1] a. Dominated Decisive Structure, Fewer Parties, Fewer Dimensions all foster SPs b. Most usually its the dominating party thats strong if there is an SP, but not always, even a dummy party can be strong (though not very strong)! c. They argue that dominant position in the decisive structure (i.e., size) matters because it makes a party more likely to be strong or even very strong (almost the case that only dominant parties can be very strong). 2. Finding frequency of strong parties as a function of party positions a. Almost impossible for a party at the median on no dimension to be a strong party, this seems true almost regardless of the partys size b. In most cases, being at median on 1 of 2 dimensions seems to give a party about 25% chance of being strong; these odds are only noticeably better for a dominant party in a very dominated sys. c. Parties at the DDM are usually strong regardless of sizethough size helps it be at medians 3. Conclusions a. Size & median-ness central to being a key player in formation, being in a small, low-dimension party system also helps. However, L&S argue, these factors matter because it makes a party strong in their sense. [Parties can be strong without these factors, & such parties would seem therefore to be the key test of L&Ss theory against alternatives.] b. Strong parties are common (and very strong parties surprisingly so). They should always obtain government according to the theory. CI. Empirical Analysis (Here Begins Part III) A. Case-study Applications of the Theory 1. Necessary Info: a. Number of parties & their seats in legislature b. Number of dimensions & party positions on them 2. Germany 1987: Figures 6.1 & 6.2 3. Ireland 1992-3: Figures 6.3 & 6.4 4. I strongly recommend you get comfortable with applications like thesesuch an example may appear on the test for you to analyze in various ways B. Testable Implications of the Theory 1. Status Quo (SQ) government at beginning of period either remains in place or is replaced by a coalition in its winset 2. If SQ has an empty winset, then SQ remains in office 3. If party is very strong, then it gets all seats in the cabinet 4. If a party is merely strong, then it gets at least one seat in the cabinet a. If a party is strong, & every cabinet in the winset of its ideal gives some particular cabinet seat to that party, then the party will get that cabinet seat b. If there is a strong party, then each cabinet seat will be assigned to that party or to one of its partners identified as those parties participating in governments in the winset of the SPs ideal C. Needed Data: 1. # Seats for each party across some set of countries over some period of time 2. Which party has which portfolio before & after each government change 3. The party positions on the salient dimensions: prior issues a. Which are the key portfolios? b. Which are the key policy dimensions? c. What is the correspondence between those two? d. [Salience weighting is possible] B. Page 76 of 91

L&Ss test of these implications against a null of random portfolio allocation [this is a wimpy null to say the least. I think L&S way overstate the absence of alternative theories] 1. Table 8.1 assesses implication B.1 from above 2. Table 8.2 assesses implication B.2 from above 3. Table 8.3 assesses implication B.3 from above 4. Table 8.4 assesses implication B.4 from above a. Table 8.5 assesses implication B.4a from above b. Table 8.6 assesses implication B.4b from above 5. Conclusions [Mine, not L&Ss]: a. Not all of these do terribly well, & theres clear variation across countries in how well they do b. The random alternative hypothesis is a weak & irrelevant alternative, why not use Minimum Winning Coalitions to at least narrow the range of parties from which randoms are drawn, for example. Hard to say, therefore, whether the performance reported is good; how would alternatives have done (and whatever L&S say, they do exist) c. That said, this is one of the only going theories which offers predictions at the level of which parties will get which seats [can you think of any alternatives?] E. L&Ss much better test: the regression on page 189 1. Dependent Variable: Gij = 1 if party i is in government j , Gij = 0 if not 2. Independent Variables: a. MSP = 1 if party is a merely strong party, = 0 if it is not b. VSP = 1 if party is a very strong party, = 0 if it is not c. PSP = 1 is party is a partner of a strong party, = 0 if it is not d. W = the partys percentage of the legislative seats e. MD1 = the partys policy distance from dimension 1 median f. MD2 = the partys policy distance from dimension 2 median 3. The results: G=-0.970 -0.151 MSP +0.630 VSP +0.248 PSP +0.041 W -0.047 MD1 -0.157 MD2 s.e. (0.34) (0.184) (0.182) (0.144) (.004) (.031) (0.041) t: (2.85) (0.82) (3.46) (1.72) (10.5) (1.54) (3.84) We like coefficients near twice their standard error or better. Thats the same as wanting t-stats around 2 or better. CII. The Portfolio-Allocation Model & Cabinet Stability (Here Begins Part IV) A. Equilibrium cabinets are more durable than any non-equilibrium cabinets which may form B. The same factors (information) which we needed to determine equilibrium cabinets are the factors which must change to alter what is the equilibrium cabinet: 1. [The list of parties, their weights, & their policy positions] 2. [The dimensionality of the policy space & its allocation to portfolios] C. Only unforeseen shocks can destabilize an equilibrium cabinet [why?]. Examples: [how do these link with I.B. above?] 1. Party splits or fusions 2. Defections or by-elections 3. Emergence of new issues or fading of old 4. Events may shift party preferences 5. Change in parties perceptions of each other (esp. regarding ability to win standoffs) D. Compare figures 10.1 & 10.2: which cabinet is more stable? Why? What general implications might be drawn from this? E. General Simulations 1. Take some set of fixed situations, described by... a. The # parties, b. Their initial policy positions, and c. Their weight in parliament 2. Calculate initial equilibrium 3. Use a computer to generate 1000 random perturbations of party positions 4. Calculate new equilibria 5. Try to draw inferences from the proportion of the time the equilibrium changes in the various settings all subjected to the same variance random shocks 6. [Tables 10.2 & 10.3 show results] Conclusions: a. Empty winset DDMs are most stable (1) if empty winset DDM is an ideal point (very strong party), then extremely stable b. Least stable appears to be where dominant party is not median F. Note on PMs ability to call elections: 1. Since PM can call an election, it can threaten its coalition partners & parliament with doing so when expects to gain a. It certainly wont do so when expects to lose b. This threat, if successful, will cause a change in government in PMs partys favor 2. Thus, L&S conclude, shifts in potential electoral support favoring PMs party can cause government change [do you think changes in other parties electoral potential will have effects? Governing parties? Non-governing parties? Parts of majority parliamentary support? Opponents in parliament?] CIII. Relaxing some of the assumptions; specifically: Reconsidering aspects of the policy space A. First a reminder: the options are not whether to make assumptions or notall logical argumentation begins with some set of assumptions. The options surround which assumptions are made & how explicitly they are made. So, when re-considering assumptions like this, the question is how much do alternative assumptions alter the conclusions. The next question, of course, is how do the conclusions drawn from one set of Page 77 of 91

D.

B.

C.

D.

E.

CIV. A.

B.

C.

assumptions fare empirically compared to those drawn from another. I would agree with L&S that, ceteris paribus, more explicit assumptions (usually following from formal argumentation) are better than less (usually following from informal argumentation). Dimensionalityallowing for many dimensions? 1. In principle, it may well seem that dimensionality of policy space is infinite or at least large & indeterminate a. even in the lattice subset of all possible policy locations, raising the dimensions decreases the probability of equilibria in random party systems 2. Two approaches to determining policy space: a. take the existing portfolios as simple & given (fixed at start of process) b. use policy issues on which parties take a public stance c. both seem to produce dimensionality which is too large given the apparent stability of governments [whats the logic of the argument here?] 3. Empirically, policy positions of parties are highly correlated across dimensions a. especially across related dimensions, but even across seemingly unrelated dims 4. such correlation reduces the effective dimensionality of politics a. [demonstrate this in two dimension, perfect correlation case: figures 11.1 & 11.2] b. [why might parties adopt correlated policy stances like this?] Differing salience of issues & dependence of policy preferences across dimensions 1. Assuming equal salience gave us circular indifference curves: pure distance-based preferences a. if unequal salience, distance along one dimension is more distasteful than distance along others ==> ellipsoidal indifferences b. Still linear, orthogonal policy reaction curves ==> theoretical propositions continue to hold, but actual equilibria case by case may well be different 2. Assuming independent preferences across dimensions gave us indifference ellipses or circles which are perpendicular to the axes a. if policy preferences non-separable then policy reaction-curves not orthogonal any more b. this can radically change the analysis Complex jurisdictions [define] 1. If we continue to assume that ministers have free reign in their portfolios areas, then allocating multiple portfolios to one party reduces the set of possible cabinets (the lattice space): fewer ways to allocate control over policy areas 2. Some new results a. party ideal points are always in the set of possible cabinets: e.g., pure prime-ministerial government b. if strong (either type) under simple jurisdictions, then strong under any complex jurisdictioning [why] c. if empty winset DDM under simple jurisdiction remains an option under complex then remains eqbm [how might it disappear?] d. reminder that effective jurisdictional complexity is what matters (policy stances are correlated across issues) Implications of these three extensions 1. Correlated preferences between parties across dimensions reduces dimensionality & simplifies analysis, increases prob. eqba 2. Jurisdictional complexity reduces the set of possible cabinets & so likewise simplifies & increases prob. eqba 3. Unequal salience doesnt make that much difference, but non-separability makes a large difference Extensions Factions & intra-party politics 1. Party leaders are potential cabinet ministers a. Their leadership involves having established a set of policy reputations & credibility b. These reputations take time to build & are generally valuable enough to be maintained c. This implies that parties have a set of possible issue stances they can credibly commit to in government given by the established policy reputations of their leaders d. Factions, from this standpoint, arise from party leaders having (probably slightly) different preferences e. By having a diverse set of leaders, a party retains some ability to strategically shift its policy stance on issues by rearranging its leadership assignments 2. Three issues arise from this possibility a. Effect of having more than one possible party ideal point (L&S still maintain that its not any possible point, but those for which some combination of leaders have established a reputation) b. Effect of factionalism on the stability of cabinets c. Incentives for leaders to split from or fuse with parties 3. L&S work from the premises that: a. There is a dominant party leadership which controls the entire partys votes b. Subordinate party leadership: (1) does not control any votes (2) cannot veto any cabinet, cannot even veto its own participation in a cabinet (3) BUT, once in a ministry it can pursue its own ideal point, not forced to pursue that of the dominant leadership c. Under these conditions, the addition of factions adds lattice points to the set of possible cabinets. Its like adding parties with zero votes & no veto power. 4. One interesting implication is that changes within opposition parties can affect the strategic options of governing party(ies) Minority & surplus-majority governments 1. One of the portfolio-allocation models most attractive features is the way it predicts minority governments under certain conditions 2. Surplus-majorities are also understandable from this perspective: parties contribute more than votes, they also contribute credible policy stances (i.e., they add lattice points to the possible set of governmentssuch an added point may be an equilibria even though the party in question is not necessary to a majority) Endogenizing issue assignment to portfolios: since different jurisdictional allocations can produce different equilibrium cabinets (as noted above), parties have preferences over portfolio allocation schemes because they prefer some cabinets to others. This too, then, can be an element Page 78 of 91

in bargaining to form a government. CV. Some very key elements of the theory & what would happen if they were relaxed A. Extreme departmentalism: 1. But compromise within cabinet goes on all the time 2. Finance minister especially has considerable influence on other ministries 3. Prime minister has strong influence across board 4. As departmentalism relaxed, lattice begins to get fuzzy ==> back toward government policy is some compromise among its members ideals on all issues rather than policy is ideal of party holding that portfolio ==> back toward chaos theories B. Party positions (or faction leadership positions) are given, fixed exogenous to model: if these are instead strategic, we may get a very different dynamic CVI. A. Table 9.1: Variables Involved in Evaluating L&Ss Making & Breaking Governments Variable Variable Label Operationalization 1 = party i is in govt j Government Gij 0 = party i is not in govt j 1 = party i is merely strong at time of government j Merely Strong Party MSPij 0 = party i is not merely strong at time of government j 1 = party i is very strong at time of government j Very Strong Party VSPij 0 = party i is not very strong at time of government j 1 = party i is a partner of a merely strong party at time of government j Partner of a Strong Party PSPij 0 = party i is not a partner of a merely strong party at time of government j Party is percentage of legislative seats at time Weight Wij of government j Ideological distance on dimension k of party i Centrality MDijk from the median on that dimension at the time of government j Who Gets Into Government? 1. Gij = -0.72 +1.01(Sij) t-statistics: (2.72) (10.93) 2. Gij = t-statistics: C. -0.51 (1.90) B. Source EJPR Laver & Shepsle: Empirics

WINSET

WINSET

WINSET

MR LH

where Sij = MSPij+VSPij

+0.68(MSPij) +1.66(VSPij) +0.73PSPij (4.62) (11.01) (5.86)

What Makes a Party Strong? 1. MSPij t-statistics: 2. VSPij t-statistics: 3. PSPij t-statistics: = -0.77 (5.23) -1.77 (9.67) -0.86 (6.96) +0.03(Wij) (6.24) -0.49(MDij1) -0.03(MDij2) (8.79) (0.55)

+0.05(Wij) -0.29(MDij1) -0.10(MDij2) (10.19) (6.43) (1.61) +0.01(Wij) (2.89) +0.18(MDij1) -0.45(MDij2) (6.75) (10.51)

D.

Who Gets Into Government? Take 2 -.15MSPij (0.82) +.63VSPij (3.46) +.25PSPij (1.72) +.04Wij -.05MDij1 -.16MDij2 (10.53) (1.54) (3.84)

1. Gij= -0.97 t-statistics: (2.85) CVII. A.

Laver & Shepsle: Examples of Applying the Model: Example 1: Empty inset DDM at a partys ideal point ==> that party is very strong & the government where it gets all the portfolios is the (1) equilibrium:

Page 79 of 91

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE
20 Seats

CE CA

DE DA

BE BA

EE
EA

20 Seats

Foreign Policy Position

AA
AD

CD

DD
20 Seats

BD

ED

AB

CB

DB

BB
20 Seats

EB

AC Left

CC
20 Seats

DC

BC

EC

Left

Economic Policy Position

Right

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE
20 Seats

CE CA

DE DA

BE BA

EE
EA

20 Seats

Foreign Policy Position

AA
AD

CD

DD

BD

ED

AB

CB

DB

BB
20 Seats

EB

AC Left

CC
20 Seats

DC

BC

EC

Left
B.

Economic Policy Position

Right

Example 2: Non-Empty Winset DDM, not at any partys ideal point, but a merely strong party exists ==> equilibrium governments (more than one) are that partys ideal point & all governments majority preferred to it (in all of which it participates, by definition of merely strong)

Page 80 of 91

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE

CE
30 Seats

DE DA
9 Seats

BE BA

EE
EA

27 Seats

Foreign Policy Position

AA

CA

AD

CD

DD

BD

ED

AB

CB

DB

BB
BC

25 Seats

EB

AC Left

9 Seats

CC

DC

EC

Left

Economic Policy Position

Right

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE

CE
30 Seats

DE DA
9 Seats

BE
DDM

EE
EA

27 Seats

Foreign Policy Position

AA

CA

BA

AD

CD

DD

BD

ED

AB

CB

DB

BB
BC

25 Seats

EB

AC Left

9 Seats

CC

DC

EC

Left

Economic Policy Position

Right

Page 81 of 91

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE

CE
30 Seats

DE DA
9 Seats

BE BA

EE
EA

27 Seats

Foreign Policy Position

AA

CA

AD

CD

DD

BD

ED

AB

CB

DB

BB
BC

25 Seats

EB

AC Left

9 Seats

CC

DC

EC

Left
C.

Economic Policy Position

Right

Example 3: Empty Winset DDM, not at any partys ideal point, but a merely strong party exists ==> equilibrium governments (more than one) are that partys ideal point & all governments majority preferred to it (in all of which it participates, by definition of merely strong), but only those governments which are no farther away from the merely strong partys ideal point than the empty-winset DDM.

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE

CE
9 Seats

DE DA
16 Seats

BE BA

EE
EA

15 Seats

Foreign Policy Position

AA

CA

AD

CD

DD

BD

ED

AB

CB

DB

BB
BC

30 Seats

EB

AC Left

30 Seats

CC

DC

EC

Left

Economic Policy Position

Right

Page 82 of 91

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE

CE
9 Seats

DE DA
16 Seats

BE BA

EE
EA

15 Seats

Foreign Policy Position

AA

CA

AD

CD

DD

BD

ED

AB

CB

DB

BB
BC

30 Seats

EB

AC Left

30 Seats

CC

DC

EC

Left

Economic Policy Position

Right

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE

CE
9 Seats

DE DA
16 Seats

BE BA

EE
EA

15 Seats

Foreign Policy Position

AA

CA

AD

CD

DD

BD

ED

AB

CB

DB

BB
BC

30 Seats

EB

AC Left

30 Seats

CC

DC

EC

Left
D.

Economic Policy Position

Right

Example 4: Non-empty winset DDM & no strong party ==> no (0) equilibrium governments

Page 83 of 91

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE
20 Seats

CE CA

DE BE DABA

EE
EA

35 Seats

Foreign Policy Position

AA
AD

CD

DDBD

ED

5 Seats

AB

CB

DBBB DCBC

20 Seats

EB

AC Left

CC
20 Seats

EC

Left

Economic Policy Position

Right

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE
20 Seats

CE CA

DE BE DABA

EE
EA

35 Seats

Foreign Policy Position

AA
AD

CD

DDBD

ED

5 Seats

AB

CB

DBBB DCBC

20 Seats

EB

AC Left

CC
20 Seats

EC

Left

Economic Policy Position

Right

Page 84 of 91

The Lattice of Possible 2-Ministry Coalitions in an Arbitrary 5-Party System

Right

AE
20 Seats

CE CA

DE BE DABA

EE
EA

35 Seats

Foreign Policy Position

AA
AD

CD

DDBD
5 Seats

ED

A is Not Strong: e.g., a coalition of B&E prefer BD to AA. B is Not Strong: e.g., a coalition of A&E prefer CA to BB. C is Not Strong: e.g., a coalition of B&E prefer DB to CC. D is Not Strong: e.g., a coalition of A&E prefer BA to DD. E is Not Strong: e.g., a coalition of A&B&C prefer CD to EE.

AB

CB

DBBB DCBC

20 Seats

EB

AC Left

CC
20 Seats

EC

Left
E.

Economic Policy Position

Right

There is one more possibility which was not drawn: Empty Winset DDM with No Strong Party ==> the (1) equilibrium is the empty-winset DDM.

OUTCOMES (Part 1) (GLM ch. 13, L&E ch. 9-10, P ch. 9-10)
CVIII. A. Lane & Ersson on stability (ch. 9) First of all, we must be clear about the stability of exactly what we are talking [Table 9.1]: some possibilities 1. Stability of Regime: uninterrupted continuation of political democracy 2. Stability of Government (cabinet): cabinet duration, with changes in the composition of government (i.e., a change of PM or of the party holding any ministry) or an election signaling a new government 3. Stability of Party System: continuity in the identity & relative strength of parliamentary parties 4. Stability of Society (what Powell calls / we have been calling Social Order): the relative absence of socio-political turmoil of some specified sorts We are interested in developed democracies; in other words, we assuming (democratic) regime-stability; we already talked about party-system stability extensively, so now we are talking about government & societal stability 1. Measurement: a. Government stability = duration of cabinet (as defined above) as a % of the constitutionally allowed max. b. Social stability: two-fold (1) Protest: (per capitized, logged) count of political strikes, riots, & protest demonstrations (2) Violence: (per capitized, logged) death count from domestic political violence, armed attacks, & assassinations 2. Though somewhat correlated, these are clearly separate things [Table 9.3] 3. [Figures showing stability across countries & over time; generate hypotheses] Explanations for Political Stability 1. Traditional explanations a. Almond, Lipset, et al.: 2-party systems in homogeneous societies are stable b. Lijphart, et al.: multiparty systems in heterogeneous societies can be stable if consociational devices present c. L&E suggest that neither of these fares terribly well as a systematic explanation 2. L&Es suggested framework for thinking about stability (& outcomes more generally) parallels the outline of this course: a. Social structure (particularly, the nature & severity of social cleavages) feed through... b. The electoral, party, & political system via which... c. Governments form which, in turn, must address the issues facing society. d. This process produces the outcomes about which we care [here, government durability & social order] e. [Their figure 9.1, my Figure I.10] They conclude with suggesting a framework for understanding socio-political instability [I modify it slightly here]: 1. Consider first the popular demand for certain things to be done publically: e.g., provision of public services, of social welfare, etc.

B.

C.

D.

Page 85 of 91

2.

E.

Then, to the degree society is divided about how these things ought to be done, means for citizens with different opinions & interests to influence policy need to be provided 3. This can be done either by providing direct mechanisms for the public to influence the central government, or, in certain cases, by devolving authority to more local authorities. 4. Insufficient supply of popular access to government through centralized influence mechanisms [e.g., effective voting, lobbying, corporatism] or decentralized government [e.g., federalism, local autonomy], relative to the differential demand for public provision, produces socio-political instability. Evidence: Tables 9.7-9.15 1. Measurement: a. Socioeconomic Structure (1) Affluence: real GDP per capita (2) Religious Awareness: % of population claiming weekly (monthly) church attendance (3) Ethnic Structure: (a) Fragmentation Index: probability two people chosen at random belong to different ethnic groups (b) Dominant-Group Measure: percent of population belonging to largest ethnic group (4) Class Structure: (a) Quintile Measures: % of national income going to top 20% of the population (b) GINI Index b. Electoral-System Properties (1) # of Referenda (2) Electoral Formula: 1=plurality/majority, 0=proportional representation (3) (average) District Magnitude (4) Legislative Proportionality c. Party-System Properties (1) Fractionalization index (2) Polarization index (3) Electoral-volatility index d. Political-System Properties (1) Index of the Degree of Individual Access to Government (Individual Autonomy) (2) Index of the Degree of Institutional Autonomy (Regional Autonomy & Federalism) (3) Chamber System: 1=unicameral, 0=bicameral (4) Presidentialism Indicator (5) Consociationalism Indicator (6) Corporatism Index (Degree of extra-parliamentary institutional access to policy making) e. Issues: index of the number & intensity of issues in political debate (?) f. Government-Formation Patterns: (1) % of governments that are Minimum Winning (including single-party majorities) (2) % of governments that are oversized coalitions (3) % of governments that are minority coalitions (4) % of governments that are grand coalitions g. Stability: Government Duration, Protest Occurrence, Violence Occurrence [as above] 2. Empirical Analysis: a. Relating Social Structure (Table 9.7) & Electoral-System Properties (Table 9.8) to Party-System Properties (1) Table 9.7 (a) Affluence is associated with party-system fractionalization: may be spurious, wealthy countries, esp. in Europe, tend to have more proportional electoral systems (b) More frequent religious-attendance associated w/ lower party-system polarization: unclear why (c) Ethnic fragmentation associated w/ greater party-system fractionalization (d) Income inequality associated w/ greater party-sys. polarization & possibly greater electoral volatility (2) Table 9.8 (a) Plurality/Majority reduces party-system fractionalization & increases party-system polarization (b) District Magnitude increases party-system fractionalization, perhaps reduces volatility (c) Proportionality reduces volatility & probably also polarization, little effect on frctnlztn [why?] b. Relating Party-System Properties (9.9), Political-System Properties (9.10), Government-Formation Patterns (9.11), Issues (9.12), & Social Structure (9.13) to Stability (1) Table 9.9. (a) Electoral participation probably reduces govt stability, may reduce social disorder [see Powell] (b) Party-system fractionalization probably reduces social disorder & perhaps govt stability (c) Party system polarization reduces government stability & may increase social disorder (d) Electoral volatility positively associated with political protest & possibly political violence, may be negatively related to government stability [see Laver & Shepsle] (2) Table 9.10 [discuss: (a) Autonomy measures: which way does causality go? (b) Chamber system: unicameral systems have more stable governments. Why? (c) Consociationalism & Corporatism: discuss] (3) Table 9.11 [discuss: (a) Main impacts are on government stability: Page 86 of 91

i) MWC & Grand Coal ==> higher govt stability ii) Minority Coalitions ==> lower stability (b) Smaller correlations with social disorder, but notice: i) MWCs may conduce toward political violence: any ideas why? ii) Grand & oversize coalitions may conduce to less social disorder, esp. violence: any ideas why?] (4) Table 9.12 [these more or less obvious] (a) Class-conflict & political-system issues have largest impacts, religious-conflict issues next (b) Regional-conflict issues can have large positive impact on political violence (c) The number of issues, if this has been reasonably accurately measured, also has considerable impact (5) Table 9.13 (a) Affluence highly negatively correlated with social disorder (b) Religious structure has strong impact on government & social stability [n.b., fragmentation index has reversed sign. The table is misleading in that respect.] (c) Ethnic structure, conversely, has remarkably little impact on stability (d) Income inequality increases political protests, probably reduces government stability, & may increase violence. c. Table 9.14 shows the stability measures & creates a summary combination index of socio-political stability d. Table 9.15 relates those to % Catholic, Income Distribution, & the Political-System Autonomy Indices CIX. Lane & Ersson on the Tax & Welfare State (ch. 10) A. Trends & Scope of the Tax & Welfare State [Figures I.1-I.6] B. Approaches to & explanations of the public-sector growth 1. Environmental approach (Dye, Wilensky, et al.): affluence, urbanization, & industrialization: Wagners Law 2. Left (Schmidt et al.) or Right (Castles et al.) Dominance increases or decreases growth 3. Influence of neighboring states (Tarschys) 4. Economic openness (Cameron) 5. Convergence hypothesis (Pryor et al.) 6. Approaches focusing on Public Bureaucracy: a. Bureaucratic budget-maximization (Niskanen) b. Bureaucratic waste (Tullock) c. Negative productivity-growth in public sector (Baumol) 7. Fiscal illusion (Buchanon & Tullock, Oates, et al.) 8. Growing number of special-interests over time (Olson) 9. Demographic factors (age structure, income structure, etc.) 10. Incrementalism (Alt & Chrystal, et al.) 11. Institutional approachpolitical decision-making structures & rules (Austen-Smith, Banks, Cox, McCubbins, et al.) 12. Trade-union strength (Schmidt et al.) 13. Economic conditions, esp. unemployment C. Evidence: 1. L&E Tables 10.13 & 10.14 2. [Some Crude New Evidence: Dependent Variable is Change in Government Expenditure as % GDP: VARIABLE COEFFICIENT STD. ERROR T-STATISTIC PROB. INTERCEPT +5.374483 2.130174 2.523025 0.0119 Expendt-1 -0.021793 0.011128 1.958294 0.0506 Unemp +0.044390 0.026856 1.652899 0.0988 Growth -.2018417 0.025520 7.909260 0.0000 Wealth +0.441837 0.215758 2.047839 0.0410 Trade +0.592740 0.398077 1.489009 0.1370 GInstab +1.217048 0.261696 4.650611 0.0000 GPart -0.103148 0.052538 1.963302 0.0500 EleYr +0.229336 0.212788 1.077764 0.2815 Aged +0.086050 0.042627 2.018676 0.0439 Vpart +0.401144 0.522074 0.768366 0.4426 RelWage +0.374223 0.477088 0.784390 0.4331 UnDen +0.001524 0.006733 0.226277 0.8211 R2 = 0.146367 3. We will look more closely at this sort of thing (transfers & debt to be exact) next week CX. Powell: (Civil) Liberty, (Political) Competition, & (Policy) Responsiveness (ch. 9) A. Questions: 1. Do democracies deliver on these qualities? 2. Do some constitutional & party-system characteristics of some democracies enable them to deliver on these qualities better than others do? B. Civil Liberties: Freedom House indices of Degree of Democracy & Extent of Civil Liberties 1. Correlation between the two extremely high ==> democracies provide high civil liberties 2. Among established capitalist democracies, there is virtually no variation on this dimension: all receive highest rank over most of the postwar period C. Political Competition 1. Two Issues: Frequency of Incumbent Replacement & Inclusion/Exclusion of Parties from Office-Holding 2. Evidence given in Table 9.1 Page 87 of 91

D.

CXI. A.

B.

C.

D.

3. Keys for Competition: presidentialism (+), Fractionalization (-) / Majoritarianism (+) 4. Keys for Inclusion: Fractionalization (+), Extreme-Party Support (-) Policy Responsiveness 1. Setup: If (a) party promises follow the interests of their support bases, & (b) policies follow promises, & (c) changes in the parties governing produces changes in policy, then (d) democracies are responsive 2. Hibbs, Tufte, & many others have found that L/R partisanship of government seems to affect, e.g.,... a. frequency & duration of labor strikes b. the redistributiveness of the tax system c. public-sector growth: (taxes + expenditures)/GDP d. unemployment rate, inflation rate, output-growth rate 3. Powell adds the following considerations: a. We should expect greater policy changes where alternation of governments is wholesale & of majority governments than where they are partial and/or of minority governments b. Where the class divide is more strongly linked to the party system (party-group linkages on class bases), we should see greater policy changes with L/R alternation in government 4. Evidence: Tables 9.3 & 9.4 Powell: Explaining Participation, Stability, & Order (ch. 10) First question was the consistency of these qualities in a democracy 1. Summarizing what has come before a. Arguments: Participation increases legitimacy, perhaps, but also increases demands on government b. Arguments: Stability increases government effectiveness, perhaps, but also increases outsider exclusion c. Ch. 2 & 4 found that voter part. appears to increase government instability but also to increase social order d. In general, these different dimensions of performance were related to different causes 2. Table 10.1 shows Powells results for executive durability, riots, & political deaths a. Deaths explained largely by environmental & constitutional factors, neither voter part. nor exec. stability very related b. Levels of rioting were more connected to these variables (1) Data support the voter-participation-yields-legitimacy argument in this respect (2) Executive stability weakly if at all related to turmoil & violence c. Participation does appear negatively related to government durability among developed democracies 3. Summary of these findings: a. Deadly violence does not appear affected by the other performance dimensions b. Rioting & turmoil are inhibited, to some extent, by participation & government stability c. Vote Part. weakens govt stability, directly & through extremist-party support, even while it enhances social order Political business-cycles: electoral manipulation of the economy 1. Likelihood that incumbents will manipulate the economy depends on a number of factors a. Available fiscal & monetary tools b. Openness of the economy c. Personal philosophy & party ideology of the incumbents d. Stake incumbents see in the coming election e. [more ideas?] f. Powell suggests that intense competition among [ideological] neighbors in multiparty systems should help restrain manipulation by [esp. connected] coalition governments (p. 208), & that this should be even more true for minority governments who are (usually) relying on ideologically-adjacent-party support from outside govt 2. Theres a large & continuing empirical literature on this; Powell shows some of Tuftes evidence [Table 10.2] Consociationalism (concept is associated with Lijphart) 1. Features a. Grand-coalition-type governments b. Provision of a veto somehow to any substantial groups affected by policies c. Proportionate sharing of government expenditures & patronage across these substantial groups d. Substantial autonomy for each group to regulate & control its supporters e. All of this is institutionalized in a set of elite implicit or often even explicit rules of the game; formal organizations of power sharing often established as welle.g. Netherlands Social & Economic Council 2. Purpose a. Elites work to establish this consensus-based decision-making structure where bitter rivalry between clearly defined socio-economic groups has historically been a problem: Neth., Austria, Belg. key exemplars b. If the above features can be established & maintained it appears to work as a peace-keeping device 3. Powell, Table 10.3 shows evidence 4. [comment: these books were written when the concept of consociationalism was all the rage in political science; enthusiasm has waned, I think correctly: consociation isnt an independent means to peace but rather a sign that peace is being institutionalized, & besides it is neither the only route to peace (e.g., regional devolution can work too) nor is it without its problems] Broad properties (strengths & weaknesses) of the three broad types of political systems identified by Powell 1. Presidential systems a. Executive durability b. Separation of powers ==> can withstand legislative opposition to exec, but divided government common c. Can hinder party discipline b/c executive & legislative branches of parties have different constituencies d. Low party-group ties e. Strong centripetal forces Page 88 of 91

CXII. A.

B.

C.

f. Previous two (d+e) ==> relatively high turnover g. Seem to be associated with smaller governments than might otherwise expect, h. But also seem to be associated with more political manipulation of economy i. Susceptible to take-over of democracy by strong executives in some settings 2. Majoritarian parliamentary systems a. Stable governments with legislative majorities b. Single-party disciplined governments c. Previous two (a+b) ==> stable & effective governments d. Even more likely to manipulate economy for electoral advantage e. More susceptible to violence than other system types f. Minorities severely disadvantaged g. Little restraint on executive between elections ==> susceptible to suspension of democracy h. Strong centripetal forces 3. Representational parliamentary systems a. Encourages multipartism b. Strong party-group ties c. Previous two facilitate direct, legitimate entry of minority groups into the system d. High participation e. Proportionality f. Facilitates extremist representation g. Less susceptible to violence h. Unstable governments i. Less likely to manipulate the economy for electoral advantage j. Possible deadlock & immobilism: ineffective governments k. Relative lack of accountability l. If single party or tight coalition emerges to/toward a sustained majority, this can be most dangerous in traditionally representational system; conversely deadlock can invite military intervention in some settings ==> susceptible to suspension of democracy Gallagher, Laver, & Mair, chpt. 13. The central question: does politics matter? 1. Difficult to determine whether a particular government made a difference in particular instance because question is counter-factual: would something different have happened if a different government had been in office? 2. Approaches a. Detailed case study of major policy interventions (1) Logic: pick a dramatic policy initiative, if world changes in some way, then the initiative caused it (2) [Strengths & Weaknesses?] b. Compare election promises with government programs announced at investiture c. Compare election promises and/or government programs with policy actions taken by government d. Specifically consider patterns of government spending or other directly observable indicators of govt. action e. Consider broad-scope macro outcomes like size of government, macroeconomic outcomes (unemp., infl., ineq...) Case Studies: privatization in France & UK 1. Privatization in France 1986-1993-Present a. The French right, RPR (Gaullists), fought the 1986 election heavily on the platform of privatization b. 20 March 1986, Gaullists, with their UDF allies, having won control of parliament, install a government replacing the left-wing PSFled coalition c. The decrees implementing the first stages of privatization were officially recorded on 24 October 1986 d. Plan (Decree): privatize 66 firms, with workforce of 900,000, & valuation of 300 Billion Francs: roughly 1/4 the value of the Paris Bourse (admittedly small). (On the order of 50 Billion Dollars) e. Within a year, 1/3 of the 5-year plan had been implemented. f. Number of small shareholders rose from 1.5 Million in 1985 to 8 Million in 1987 [Significance of this?] g. October 1987 stock-market crash ==> heavy losses on those investors (and others, but perhaps especially on newer) h. The right loses the presidential & the subsequent parliamentary elections in mid 1988 i. PSF-led coalition halts the privatization j. The 1993 election returned the right-wing coalition to control of parliament, government. k. The new Balladur government begins the privatization movement anew 2. Privatization in the UK 1979-Present a. Conservative government that took office in 1979 (PM Thatcher) had not fought previous election on privatization b. Began with selling off the profitable British Aerospace & Britoil for needed cash (unwilling to raise taxes or run larger deficits). c. These early successes fed through & privatization was more central the partys 1983 election campaign d. Privatization Plans now extended to electricity, gas, water, & other previously untouchables e. Virtually all of these privatization completed by 1990: (1) 50 +/- companies, the total state sector +/-, well over 24 Billion Pounds by 1988 (about 40 Billion Dollars) (2) shareholding population tripled to 10 million people (3) 600,000 jobs shifted from public to private sector 3. Some but more limited privatization also in Germany, Italy, the Netherlands, Belgium, even Sweden & Austria Party manifestos & government policy-programs 1. First step from campaign promise to outcomes is govts announced program (its intentions) at start of tenure 2. Need to code both documents consistently: content analysis [describe content analysis] Page 89 of 91

D.

E.

a. GLM focus on % of document expressly promoting free-market economics b. The question is whether alternation in government produces the promised changes in government programs 3. In some cases (e.g., Norway: Table 13-1, 13-2), the correspondence is remarkable; varies some generally (Table 13-3) 4. [Discuss Table 13-3] 5. General conclusions: a. When parties diverge considerably in their electoral manifestos, the divergence is reflected in their programs b. Conversely, where parties manifestos dont reveal much policy debate, the lack of debate is reflected in their respective programs c. Where alternation in government is partial [define?], response in government programs also partial d. Programs tend to be closer to permanent fixtures of government than its partners Party manifestos & actual government policy 1. Methodological issues a. Whats a pledge & whats just a platitude or aspiration? b. Distinguish pledges which person would have the power to fulfill if elected from those not c. Fulfillment is going to be a matter of degree 2. Previous findings: a. Richard Rose: Do Parties Make a Difference? (1980) compared Conservatives 1970-4 with Labour 1974-9: Finds that parties overwhelmingly do what they promise b. Rallings (1987) extends it to UK 45-79 & finds 70% fulfillment rate (similar rates found elsewhere in other studies) (1) Clarity of promise (esp. those regarding increases in benefits etc.) increases its odds of being fulfilled (2) Promises to repeal ideologically unacceptable acts of previous government almost always fulfilled c. Evidence from multiparty coalition governments is currently lacking: (1) [Some rough evidence from Hofferbert et al. 1996 indicates generally, across countries & time, policies react to the voteweighted policy-programs of parties, though with some lag (inertia). See Figures] d. Criticisms: (1) Many focus only on doable promises (2) Lots of promises are straightforward & uncontroversial (3) Parties may choose to make only promises which are easily kept so that it can carry their fulfillment to the electorate [does this seem a criticism to you?] 3. However you slice it, the evidence continues to mount that reasonable promises are remarkably well fulfilled Party government & public spending 1. Approaches a. Compare different countries at the same time [strengths & weaknesses?] b. Compare same country at different times [strengths & weaknesses?] [Could do both of course] 2. Left/Right & the Size of Government a. Broad association b/w L/R government partisanship & government size [see Figure], but considerable exceptions b. Resolving this apparent puzzle (1) Some have narrowed focus to more specific elements of budgets, e.g. welfare spending [see Figure] (2) Many others have noted that partisan effects, especially in the short run, should be more readily observed in changes in budgets than in their levels since the latter tend to reflect long accumulated histories of decisions & change incrementally (3) Many acknowledge a greater need to control for other factorseconomic & structural conditions for example (4) Some have moved beyond simple left-right partisanship as the mode of government impact on policy (e.g. Castles & others have considered the distinctive impact of Christian Democracy)

OUTCOMES (Part 2) (Downs, M&R, Franzese)

CXIII. A. Downs, Anthony, Why the Government Budget is Too Small in a Democracy Voters & the budget 1. Budget is sum of items decided upon individually; all which win net positive votes 2. Voters evaluate budget on basis of particular benefits it brings them individually 3. In 2-party democracy, each election can be considered a contest b/w prospective budgets Rational ignorance [define] leads to mis-evaluation of public v. private spending 1. Whether the budget is too large or too small depends on the specific forms of rational ignorance present 2. Private exchange is quid pro quo, but public benefits divorced from costs by nature of public goods & taxation & because government seeks redistribution ==> coercion necessary & remoteness [define] Individual preferences over distribution of taxes & spending differ; no one person has his/her desired equilibrium amount of the good because lumpy public-good provision, everyone therefore thinks the govt is spending too much. Thus the strong anti-spending bias in public provision 1. Remoteness: benefits more remote than taxes 2. Uncertainty: benefits less certain than taxes However, to the degree revenue costs can be concealed, the contrary is true 1. ==> Net effect is ambiguous, but Downs claims that even indirect taxation is more apparent than benefits, so should be that budgets are too small. 2. ==> Increasing complexity of polity & economy implies that gap between correct size & actual size is increasing. Meltzer, Allan & Scott Richard, Why Government Grows (and Grows) in a Democracy By all accounts, government has grown remarkably & remarkably consistently & swiftly in the US & elsewhere. Given that trend: If a majority believes that efficiency or freedom is sustained only if there is some private output & employment, there will be a growing consensus that government is too big (p. 114) Common explanations Page 90 of 91

B.

C.

D.

CXIV. A.

B.

C.

CXV. A.

B.

CXVI. A.

B.

C.

Cold War / Military Industrial Complex a. Explanation is that military confrontation, with or without a kind of lock-in facilitated by close government ties to military producers, produces the growth of government b. Problem is that government has been growing more or less continuously for two centuries; besides, Vietnam came & went & the state kept on growing; besides, other states with much smaller militaries have seen similar growth of government 2. Schumpeters explanation: a. Capitalisms success in efficient production & in extension of private property & freedom would challenge many of the institutionschurch, bourgeois family, dynastic ambitions of capitalists, & respect for privacy in private & public affairs. Each of these institutions would be challenged & criticized & only weakly defended. b. The weak defense comes from collective-good nature of defending free-market capitalism. While all firms might gain by the extension & security of free-market, for example, no single firm has the incentive to do anything about it on its own. Much less costly to cooperate & to work with regulatory agencies than to combat them. c. Problem is that, if Schumpeter were right, those capitalist countries doing best would be breaking down most & seeing the most growth of government, but that appears to be false. Third explanation focuses on differential distribution of political influence (votes) & income [M&Rs preferred]. 1. Free-market capitalism tends to distribute income skewed left, i.e. such that there are relatively many poorer than average & relatively few wealthier than average. 2. This implies that the median person is poorer than the average person 3. Given universal suffrage, then, the median voter is poorer than average & so desires government redistribution. 4. Universal suffrage & income inequality combine to imply a desire for the growth of government by median voter Franzese, Political Participation, Income Distribution, & Public Transfers in Developed Democracies Start with the standard Meltzer-Richard-type argument: 1. Median person poorer than average, therefore median desires positive net transfers 2. Taxes / Transfers reduce incentives to work / to invest, therefore median desires larger net transfers until this cost outweighs the greater redistribution garnered. The greater the difference between the median & the mean, the larger transfers before this occurs. However, not everyone votes, and, in particular, the relatively wealthy vote more than the relatively poor. 1. Therefore, the positive impact of increases in income skew on transfers is larger the higher voter-participation 2. Conversely, positive impact of increases in voter participation on transfers larger the greater underlying income skew Franzese, The Positive Political Economy of Public Debt: An Empirical Examination of the OECD Postwar Experience Much of the expansion in public debt since the mid-seventies is due to economic conditions; specifically: 1. Rise in unemployment & slow in growth reduced revenues & raised expenditures, given the current taxation & expenditure systems 2. That initial impact was amplified by ensuing rise in real interest rates ==> explosive public-debt path in many places The key political factor is then fractionalization (and less so polarization) of governments 1. Where governments were fractionalized, fiscal-policy adjustments to address the rising debt were delayed as parties in government, while they may have agreed on the need for adjustment, naturally disagreed over whose constituents were going to pay the costs 2. Where governments were more unified (fewer parties, not divided government), adjustment plans were easier to implement: the governing party more easily placed the adjustment costs on those out of power 3. Where such economic conditions hit environments with fractalized governments, debt path was yet more explosive Several other political, structural, & institutional factors appear to have had some, but more minor, effects

1.

Page 91 of 91