3: Interinstitutional Decision Making

1
3 Interinstitutional Decision Making
European Co-operation and Integration
7.05.2009
Universität Konstanz, Dept. Politics and
ManagementSummer 2009Gerald Schneider
2
Outline

• Neo-institutionalist research
• Classification of decisions and decision making
  modes
• Rationality, Expected Utility, Game Theory
• Spatial Theory of voting
• Voting power analysis
• The core of the European Union

3
Neo-institutionalism Aims and Definitions

• "to provide fresh answers to old questions about
  how social choices are shaped, mediated and
  channeled by institutional arrangementsPaul J.
  DiMaggio/Walter W. Powell 1991. The New
  Institutionalism in Organizational Analysis.
  Chicago.
• General conceptualization of institutions
• Institutions are "the formal rules, compliance
  procedures, and standard operating practices that
  structure the relationship between individuals in
  various units of the polity and economy". Peter
  Hall 1986. Governing the Economy The Politics of
  State Intervention in Britain and France.
  Cambridge Polity Press.
• Institutions are "legal arrangements, routines,
  procedures, conventions, norms, and
  organizational forms that shape and inform human
  interaction"" Asbjorn Sonne Norgaard 1996.
  "Rediscovering Reasonable Rationality in
  Institutional Analysis," European Journal of
  Political Research 2931-57.

4
Extreme definitions

• "I treat institutions .... as being a "long-lived
  equilibrium patterns of rational behavior" and
  thus realized outcomes in a strategic game "that
  society plays" "A more successful theory must
  clarify three fundamental phenomena ... the
  establishment and maintenance of endogenous
  incentives for cooperation among selfish
  individuals the achievement of coordination,
  that is, of standards, organization, or
  conventions, in complex settings and the way in
  which communication .... can take a life of its
  own in the definition and functioning of
  institutions." Randall L. Calvert 1995. "The
  Rational Choice Theory of Social Institutions
  Cooperation, Coordination, and Communication". In
  Modern Political Economy, eds. Jeffrey S.
  Banks/Eric A. Hanushek. New York Cambridge
  University Press.
• "The institutionalists' structure is a cultural
  one it is Western rationality and individuality
  that creates states, markets, bureaucratic
  organizations.Martha Finnemore 1996."Norms,
  Culture, and World Politics Insights from
  Sociologys Institutionalism." International
  Organization 50325-347.

5
Sociological and rationialist conceptualizations
of institutions
6
Effects of institutions according to March/Olsen

• Introduction of a framework to assess whether
  actions are appropriate (logic of
  appropriateness9
• Structuring of actions through rules and routines
  
• Accumulation of historical experience
• Adding conflict, contradiction, conformity and
  standardization to the repertoire of human
  interactions
• Creation of confidence in interpersonal relations
  Source James March und Johan Olsen 1989.
  Rediscovering Institutions The Organizational
  Basis of Politics. New York The Free Press.

7
Three institutionalisms
8
Continued
9
Sociological institutionalism

• The material context only becomes relevant
  through the social context Constructivists
  argue that it is not (nuclear) weapons that
  matter. After all, the US worries very little
  about the large quantity of nuclear weapons held
  by the British...
• Against methodological individualism
  Constructivists emphasize a process of
  interaction between agents and structures the
  ontology is one of mutual constitution, where
  neither unit of analysis --agents or structures
  --is reduced to the other and made ontologically
  primitive.
• Jeffrey T. Checkel 1998.Social Construction,
  Institutional Analysis and the Study of European
  Integration. Unpublished paper, ARENA,
  Universitetet i Oslo.

10
Historical institutionalism

• Pierson, Paul 1996. "The Path to European
  Integration A Historical Institutionalist
  Analysis." Comparative Political Studies
  29123-163.
• "In a number of instances, the short-term
  preoccupations of institutional designers have
  led them to make decisions that undermined
  long-term member-state control."
• Further important explanatory variables
• -"ubiquity of unintended consequences"
• -"instability of policy preferences"

11
Decisions in the European Union
12
Decision Making Modes

• Bargaining Power and information asymmetries
• Voting Ideological motivation of actors vs.
  ability to cast the decisive vote (spatial theory
  of voting vs. voting power models)
• Delegation (Ideological vs. monetary incentives
  of agent information asymmetries)

13
Bargaining, voting and delegation
L C
R
Bargaining
Zone of agreement (win set or bargaining set)
Voting (Median voter theorem)
Delegation
Unanimity
Majority V.
C
14
Historical development of the power division in
the European Union
15
Decision Making in the EU and Beyond A
Classification

• Constitutional choice (Decision making about
  rules) Crucial component Veil of ignorance
• Decision making within rules a) Legislative
  decision making b) Administrative decision
  making
• Forms of power 1) Veto power2) Gate keeping 3)
  Agenda setting

16
Players and decision making mode in the EU

• Who are the players and on what decision making
  mode do they rely- in the IGC or in the European
  Council- in the interactions between EU bodies-
  in the Council of Ministers- in the European
  Commission- in the European Parliament- in the
  European Court of Justice- between the member
  states and the institutions of of the EU a)
  before the first decision in the EU is made b)
  after the decisions at the EU level have been
  made

17
The concept of rationality

• Actors are goal-oriented and develop a complete
  preference ordering over the possible outcomes.
• Actors have transitive preferences over outcomes.
  If an actor makes two decisions 1) Option
  1gtOption 2 (Choice Set Option 1, Option 2) 2)
  Option 2gtOption 3 (Choice Set Option 2, Option
  3), then transitivity requires that she
  prefers 3) Option 1gtOption 3 (Choice Set
  Option 1, Option 3)
• Note that a) Preferences are not observed. They
  are deduced from actions. b) Preferences over
  outcomes are fixed. Preferences over strategies
  can change depending on the information an actor
  receives. c) Rationality is not used as a
  normative criterion. The choices of an actor do
  not have to be socially desirable

18
Decision theory and

• Decision theory analyzes choices by single
  actors. Such decisions can be formally described
  by the (i) Strategies A player can choose
  between m alternative strategies si?Si of her
  strategy set Si?Rm.(ii). States of the world
  The set E contains all states in which a problem
  can occur.
• If we move to a social situation and the choices
  by one actor also depend on the choices by at
  least one other actor, we are in a strategic
  situtation, which is analyzed through
  game-theoretic approaches.

19
game theory
Game theoretic decisions can be formally
described through criteria (i), (ii) and (iii)
Players The set N1,...n includes the players
of the game. Every game can thus be formally
described through the game form ?(N, S, E). If
the preferences and utility functions are
specified this transforms to ?(N, S, u).
20
Definitions of game theory I
Game theory analyzes situations in which (i) the
final outcome depends on the decisions of
several players (ii) every player knows about
this strategic interdependence (iii) every
player assumes that all the other players
equally know about this strategic
interdependence (iv) every player considers (i),
(ii) and (iii) in her decisions Adapted from
Manfred J. Holler/Gerhard Illing. Einführung in
die Spieltheorie. Springer Verlag 1991.
21
Definitions of game theory

• Game theory is a mathematical theory of rational
  strategy selection used to analyze optimal
  choices in interdependent decision situations
  the outcome depends on the choices of two or more
  actors or players, and each player has
  preferences for all possible outcomes.Steven J.
  Brams. Negotiation Games. Routledge, 1990. P.
  268.

22
Expected utiliy theory I

• Imagine a situation where the Commission
  President wants to introduce a new EU
  constitution with a EU-wide referendum, but is
  not sure about the nature of the population.
  The constitutents can either be good-natured or
  unforgiving. The uncertain decision against
  nature can be described with a game tree. The
  Commission President can either opt for an
  activist or a rather passive constitution.

23
Expected utility theory II
Decision node
Information set
4
activist
CP
Good- natured
1
passive
N
2
activist
Unforgiving
CP
3
passive
24
Expected utility theory III

• The expected utility of the Commission President
  depends on her assessment of the nature of the
  constituents. She opts for the activist
  constitution if the expected gains of this
  project exceed the expected benefit of the
  alternative project. Formally, EU(A)4p1(1-p)gtEU(
  P)2p3(1-p)
• We can calculate the Commission Presidents
  choices by assuming that she is indifferent
  between the two options (i.e. EU(A)EU(B).
• 4p1-p2p3-3p
• 3p1-p3
• 4p2 ? p½

Activist constitution
Passivist constitution
p
0
1
25
Expected utility theory IV

• Von Neumann and Morgenstern have shown that one
  can derive unique utility functions from a few
  straightforward axioms. Hence, if s is a strategy
  subset of a game ?(N,S,u), we can derive a
  utility subset u(s)that is part of the set P of
  all possible utility combinations
  Pu(s)s?S(u1(s), ...., un(s)) for all s?S
• Actors differ in the way in which they deal with
  uncertainty. They can generally belong to three
  types - risk acceptant - risk neutral - risk
  averse
• The shape of the utility function describes
  whether an actor prefer lotteries with a smaller
  variance (risk aversion), risky choices (risk
  acceptance) or do not differentiate between such
  options (risk neutral).

26
Cooperative vs. non-cooperative game theory

• Cooperative game theory allows players to make
  binding agreements, while in noncooperative games
  players cannote make such agreements.
• In both branches of game theory, researchers can
  make use of solution concepts to assess the
  outcome of the game. An outcome that is chosen by
  the players is an equilibrium.

27
The Logic of Solution Concepts I

• Solution concepts are introduced to reduce the
  number of potential equilibria. A typical example
  is the so-called Battle of the Sexes. The
  simultaneous-move version of the game contains
  three Nash equilibria. This can be illustrated
  with a situation where two countries struggle
  over the harmonization of some regulations.

Country 2
Your way My way
0
1
My way
2
0
Country 1
0
2
The Frank Sinatra Game
Your way
1
0
28
The logic of solution concepts II

• The first two Nash equilibria are the two My
  way/Your way- combinations. The third
  equilibrium is an equlibrium in mixed strategy.
  This means that Country 1 mixes between the two
  options in a way that Country 2 becomes
  indifferent. Country 2 does the same. This
  meansEU2(Your way)1pEU2(My
  Way)(1-p)23p1 ? p1/3 and (1-p)2/3
• Equivalently, we receive for the probability q
  that country 2 opts for Your way a probability
  of 2/3 so that the third equilibrium can be
  described with the following mixed strategy
  pairsCountry 1 1/3 My way 2/3 Your way
  Country 2/3 Your way1/3 My way.
• In general, a Nash Equilibrium is an outcome from
  which no player has an incentive to depart. Mixed
  strategies involve a random choice among pure
  strategies.

29
Dominant strategy

• A strictly dominant strategy is a dominant
  strategy that leads to abetter outcome in every
  contingency. A dominant strategy equilibrium
  contains the set of dominant strategies.
• Example PDG

C
D
3
4
C
3
1
Player 1
1
2
D
2
4
30
Example The Madame Cresson Dilemma

• Define a matrix game for the interaction that
  went on for a long time between Madame Cresson
  (of special fame among all dentists) and
  President Santer. The Commission President had
  two options Tell her publically to resign or
  refrain from doing so. Madame Cresson in return
  could resign voluntarily or try to stay on.

Do nothing S Resign
Resign
C
Hold out
31
Incomplete information

• In a game of incomplete information, Nature moves
  first and is unobsorved by at least one of the
  players. Otherwise the game is one of complete
  information.. Eric Rasmusen 1989. Games and
  Information. An Introduction to Game Theory.
  Blackwell, p. 53.

32
Solution concepts Perfect Bayesian Equilibrium
and Sequential Equilibrium

• Perfect Bayesian EquilibriumA PBE is a pair (s,
  ?) of a strategy combination s and of a set of
  beliefs ? that satisfies (i) every strategy
  choice of a player is Nash, given the strategies
  and beliefs of the other players (ii) the
  beliefs of the players at each information set
  are rational, given the information available so
  far in the game.Rational beliefs means that
  players update their beliefs according to Bayes
  rule.
• Sequential equilibriumThis concept adds a third
  condition (iii) All the beliefs have to be
  consistent.Consistent means that if players
  reach a node out of equilibrium, the subsequent
  game has also to be a sequential equilibrium.
  Bayes rule cannot be used at such nodes because
  the probability of such events is 0.

33
Bayes rule

• Applying Bayesrule means that players learn from
  observing events and update their beliefs in a
  rationalist fashion. If we can only distinguish
  two states of the world A and A, we can
  calculate the posterior belief

Example The probability that you can convict a
liar by using a lie detector is 0.9. How large is
the probablity of convicting a politician of
lying if you know that 10 per cent of all
politicians are liars?
34
Example
Tells a lie
Tells the truth
Liar
Tells a lie
No liar
Tells the truth
Adapted from James D. Morrow. Game Theory for
Political Scientists. PUP 1994.
35
Spatial theory of voting I

• Median voter theorem If three alternatives
  exists, the centrist position will be chosen in
  issue-by-issue voting or decisions along a single
  dimension (Black).
• Extension Candidates will converge to the
  position of the median voter if preferences of
  the electorate are normally distributed (Downs).
• Yet, no median voter exists in multidimensional
  settings, and no convergence to the median voter
  position can be observed in more complex
  electoral competitions.

Source Univ. of Virginia
36
Spatial theory of voting II Chaos theorems
(McKelvey, Schofield)

• So long as a society preserves democratic
  institutions, its members can expect that some of
  their social choices will be unordered and hence
  inconsistent. And when thisis true, no
  meaningful choice can be made.
• W. Riker, Liberalism Against Populism

2
1
3
37
Spatial theory of voting III

• Ideological differences can be described with
  spatial preference models. (X??m, with m being
  the number of dimensions in the Euclidean space
  ?).
• Decision makers have quasi-concave utility
  functions Ui(x) over the options in X. The
  utility of an option is decreasing with the
  distance from the position one prefers (bliss
  point) xi(xi1, xi2). Consequence indifference
  curves are ellipses or circles with the bliss
  points as centres.
• The options that an actor prefers over the
  current policy are described by circles through
  the status quo (Preferred-to set).
• The contract curve links the bliss points of two
  decision makers.

38
Spatial theory of voting IV

• Decision makers are either sincere or
  sophisticated. In the former case, they are
  short-sighted and just opt for the immediatley
  available option. In the latter case, the act
  rather future oriented and optimize their actions
  over the whole interaction (the whole game).
• Early contributions assume perfect information.
  More recent research does not rule out the
  possibility of uncertainty about the decision
  making process or about the preferences of other
  actors.
• The Nash equilibrium is the dominant method to
  describe stable decisions if there are any. The
  Nash solution is reached if none of the players
  has an incentive to deviate.

39
Spatial representation of utility function
40
Multi-peaked preferences
41
Preferred-to-sets, win sets and the Pareto set
(unanimity core)
42
The unanimity and the Q- core
Dimension 2
Q-Core
Dimension 1
43
Minimal quorum for the existence of a q-core
(Greenberg 1979)
Kramer number q W/(W1) whereby W stands for
the number of dimensions in the decision
making space W1
q1/2 W2
q2/3 W3 q3/4 Consequence
Multidimensionality is not necessarily a
problemfor the Council of Ministers. Relatively
strict rules about the dimensionality of
proposals and the separation of power between
the different legislative bodies are also
helpful in preventing legislative chaos.
44
General implications of the agenda setting
literature
1. If a commitee votes sequentially on single
issues, the median position wins (only valid
under the assumption of separable
preferences). 2. If some actor possses special
agenda setting rights (closed rule), equlibrium
solutions (structure-induced equilibrium or SIE)
lie in the interval between the bliss point of
the median voter and the bliss point of the
agenda setter. 3. If all voting bodies possess
agenda setting rights, the median solution
persists. 4. The larger the number of veto
players, the greater the stability in the
decision body.
45
General implications of the spatial theory of
voting for majority decisions
46
Non-spatial voting theory Voting power indices
Weighted voting Situation of three states
(majority threshold 70) Pivot
state 1 2 3 Portion of voters 50 30 20
Permutations (1,2,3) - 1 - (1,3,2) - - 1 (
2,1,3) 1 - - (2,3,1) 1 - - (3,1,2) 1 -
- (3,2,1) 1 - - Number of pivots 4 1 1 Sha
pley value 2/3 1/6 1/6
47
Implications of the division of power Garrett
and Tsebelis
G. Tsebelis/G. Garrett IO, The Institutional
Foundations of Intergovernmentalism and
Supranationalism in the European Union.
International Organization 55/2 357-390 2001
48
Continued.
49
Continued.
50
Franchinos qualification
Fabio Franchino 2004, British Journal of
Political Science