ThirtySix Countries

1
Thirty-Six Countries

• Lijphart sample
• Democratic in mid-1996
• Continuously democratic since 1977

2
Dahls Democracy

• Eight criteria
• Right to vote
• Right to be elected
• Right to compete for election
• Free and fair elections
• Freedom of association
• Freedom of expression
• Alternative sources of information
• (Electorally) responsive institutions

3
Democratic Countries

• Lijphart relies on the Freedom House ranking
• Independent organization
• Based on principles similar to Dahl
• Produced since 1972
• If ranked free by FH included in Lijpharts
  sample

4
Borderline cases

• India, Papua New Guinea, Columbia Venezuela.
• Better to err on the side of inclusion?
• Do institutions determine inclusion in sample?
• Is it a serious problem if the above countries
  are all in the majoritarian model?

5
Other exceptions

• Switzerland (1971)
• Australia (1962)
• U.S. (1965)
• U.K., France, the Netherlands, Belgium
  (colonial)
• Allied Powers, Israel
• Focus on years after 1945 minimizes these
  problems

6
Extended Period of Democracy

• Stable democracies interest in outcomes
  generated by democracies
• More observations more accuracy
• Statistical outliers
• Events
• Why 19 years ?
• Why only current democracies ?
• Big country bias ?

7
The Sample

• Variation in time of democratization
• Democratic since 1940 Homogenous
• Economically developed
• Industrialized
• Urbanized
• North Atlantic
• Exception Societal Divisions

8
Societal Divisions

• Easily quantifiable often used in studies
• Problems
• Ethnic divisions not the only important ones
• Within group differences
• Depth of division
• Political significance/Organization
• Basically, problems of defining salient divisions.

9
Societal Divisions

• Lijphart opts for a simple classification
• Plural, semi-plural and non-plural
• Plural primarily linguistic
• Reflects mid-1990s
• Non-plural does not mean homogenous

10
Measure of Socio-Economic Development

• GNP
• PPP
• Human development index
• Length and Health of Life
• Knowledge
• Standard of Living

11
Population, etc.

• Big variance in population
• These factors (pluralism, socio-economic
  development, population) may influence choice of
  institutions and outcomes
• Not statistically significant.
• Why is that important ?

12
Party Systems

• Two-party vs. multiparty systems
• Two-party systems fall in the majoritarian
  category
• Benefits
• Offer voters a clear choice
• Moderating influence compete for the median
  voter
• Contradictory ?

13
Party Systems

• Further benefits of the two-party system
• Requirement (?) for the formation of stable
  single-party government. Coalitions considered
  unstable.
• Lowells hypothesis.
• Accountability

14
The Effective Number of Parties

• How do we measure the number of parties?
• Concentration of power.
• The U.K., New Zealand commonly considered
  two-party systems despite presence of other
  parties.
• Should we count small parties as equal to big
  parties ?

15
The Effective Number of Parties

• Sartori
• Disregard parties without representation.
• Consider relative size of parties
• Eliminate minor parties.
• Relevant parties are those with coalition
  potential or blackmail potential.
• Attempts to find significant parties
• Lijphart not sufficient attention to size

16
The Effective Number of Parties

• Blondels four categories
• Two-party system 55-45
• Two-and-a-half 45-40-15
• Multiparty w/dominant 45-20-15-10-10
• Multiparty w/o dominant 25-25-25-15-10
• An attempt to capture bargaining situation.
• Imprecise ?

17
The Effective Number of Parties

• Laakso-Taagepera
• The Effective Number of Parties
• N1/(?si2), where si is party is proportion of
  seats won.
• If n equal sized parties then Nn.
• Size disparities reduce N.

18
The Effective Number of Parties

• Blondels vs. Laakso-Taagepera
• Two-party system 55-45 2.0
• Two-and-a-half 45-40-15 2.6
• Multi w/dominant 45-20-15-10-10 3.5
• Multi w/o dominant 25-25-25-15-10 4.5

19
Closely Allied Parties

• We now know how to count parties.
• But how do we define a party?
• CDU, CSU in Germany
• Liberal and National parties of Australia
• Christian Democrats split along linguistic lines
  in Belgium.
• German and Australian parties usually treated as
  one.

20
Closely Allied Parties

• When should we treat parties as one?
• Dont compete for votes
• Degree of cooperation in Parliament
• Parliamentary Party Groups
• Behavior in cabinet formation
• Bargain together ?
• Length of cooperation.
• Not only electoral alliances
• Lijpharts solution Calculate both average.

21
Factionalized Parties

• Divisions in with parties can also be
  problematic.
• Degrees of cohesiveness.
• Necessity of intra-party bargaining.
• Lijpharts solution Factionalized party two
  equal sized parties.

22
Thirty-six Party Systems

• Things to note
• Considerable range of size
• Why does the U.S. have 2.40 effective parties?
• Does experience have an influence on the number
  of effective parties ?

23
Issue dimensions

• How do we determine what parties want ?
• Manifestos, statements
• Party in power
• Its voters ?
• With regard to issue dimensions we should focus
  on difference between parties.
• Relevant parties (Sartori)
• Durability of issues

24
Issue Dimensions

• Lijphart identifies seven issue dimensions in
  the 36 countries
• Socioeconomic.
• Most important.
• Declining importance?
• Religious.
• About half of sample.
• Variation in importance across countries and
  time.
• Often related to moral issues

25
Issue Dimensions

• Cultural-Ethnic
• Important in plural societies except Israel.
• Usually high salience, India Switzerland the
  exception.
• Urban-Rural
• Agrarian Parties Nordic countries.
• Increased urbanization often has the effect of
  de-emphasizing the parties programs.

26
Issue Dimensions

• Regime support
• Parties opposing the democratic regime.
• Communist parties.
• Outlawed in some places.
• Foreign Policy
• Pro-Soviet Communist Parties
• Opposition to NATO
• EU membership
• Usually moderate

27
Issue Dimensions

• Materialist
• Participatory Democracy
• Environmental Issues
• Post-materialist parties only appeared in few
  countries Germany, Norway, Sweden, the
  Netherlands.

28
Issue Dimensions the Party System

• Two views on the relationship between the number
  of issue dimension and the party system
• The first hypothesis that the number of issue
  dimensions, or salient issues, determines the
  number of parties.
• That is, parties form around issues that the
  voters care enough about

29
Issue Dimensions the Party System

• The second view posits that the party system
  determines the number of issue dimensions.
• A more common view emphasis the role of the
  electoral system in determining the number of
  parties.
• Electoral system -gt Party System -gt Salient
  Issues

30
Issue Dimensions the Party System

• The first view, taken to its extremes, might
  argue that there is a clear correspondence
• Two issue dimension -gt 4 parties
• E.g., socio-economic and language cleavages might
  require a left-french party, a right
  french-party, a left-dutch party, and a right
  dutch party.
• Three issue dimension -gt 8 parties

31
Issue Dimensions the Party System

• That is clearly not the case
• We dont observe that many parties
• Parties face a trade-off between taking strong
  positions on issues and electoral strength.
• Issue dimensions may be correlated to some
  degree, i.e., their may be a relationship between
  class and religion/ethnicity.

32
Issue Dimensions the Party System

• Intuitively, there is a strong case for some
  correspondence.
• Parties need issues to distinguish themselves
• At the least, the number of issue dimensions may
  provide a ceiling for the number of parties
• The second view will be examined in more detail
  later in the semester

33
Issue Dimensions the Party System

• Lijpharts analysis adds an additional reason
  for why we should expect to find a relationship
  He has defined the number of issue-dimensions as
  differences between parties.
• Is this problematic ?
• Only includes differences that the political
  parties have articulated.
• Other significant cleavages may exist.

34
Issue Dimensions the Party System

• Lijphart finds a fairly strong positive
  relationship between issue dimensions and the
  party system.
• Taagepera and Grofman suggest a relationship
  NI1
• Based on empirical observation
• Holds reasonably well (surprising ?)
• Germany, the U.S., the Bahamas Norway the
  furthest away from the T-G prediction.
• Why ?