Experimental Economics 20001

1
Experimental Economics 2000/1

• John Hey
• jdh1_at_york.ac.uk

2
Experimental Economics

• General Reading
• Lecture Programme
• Specific Reading
• Seminar Programme
• Examination
• Other - web page, EXEC, ESA

3
Lecture Programme

• Prisoners Dilemma Games
• Co-ordination Games
• Centipede Games
• Public Goods Problems
• Beauty Contests
• Herding Models
• Markets
• Asset Markets
• Herding in Markets

4
Seminar Programme

• Introduction to Z-tree (week 4)
• Illustration of Z-tree experiments (week 6)
• Outlining the design of an experiment (week 8)
• Implementing the experiment (week 10)
• Z-tree www.iew.unizh.ch/home/fischbacher/
• ESA www.econ.nyu.edu/Dept/ESA/index.htm

5
The Methodology of Experimental Economics (1)
Testing theories

• We reproduce the conditions of the theory in the
  laboratory
• We have one subject for each agent in the theory
• We give the subjects the same incentive as the
  agents in the theory
• We then see how the agents behave.

6
Prisoners Dilemma Games

• The Basic Structure and why it is of interest
• One-off prisoners dilemmas
• Testing the theory in the lab
• Finitely repeated prisoners dilemmas
• Randomly repeated prisoners dilemmas
• Infinitely repeated prisoners dilemmas

7
An example of a Prisoners Dilemma
8
General Structureagtc dgtb bgta
9
Another Example
10
A further example
11
Repetition?

• Finite
• Infinite
• Random

12
Test in the lab for the one-off game

• Two Subjects
• Given Payoff Matrix
• Asked to Decide
• Paid off the amount in the realised cell of the
  table
• Sent away

13
Alternative lab scenarios

• Several subjects (more than 2)
• Play several games
• Matched with different or same people each game,
  depending on objective of experiment
• Paid the payoff in all the games or the payoff in
  a randomly chosen game

14
Existing Experiments

• See Handbook, pages 26 to 28 variations on the
  PD have been the subject of virtually continuous
  experimental interest since the 1950 experiment
  of
• Selten and Stoecker (1986) - note copies of all
  references are with Helen Hawksby - finitely
  repeated supergames

15
Selten and Stoeckers Basic Game
16
Selten Stoeckers Treatment

• 26 subjects
• each subject participated in 25 supergames - each
  with 10 repetitions of basic game
• (actually only 12 subjects)

17
Basic Results

• Always some co-operation (not the Nash outcome)
• Extent of co-operation first grew...
• then diminished
• but did not disappear entirely
• Authors fitted Learning Model to data

18
A synopsis of their results
19
Conclusions

• Straightforward to test Game Theoretic
  predictions of Prisoners Dilemma Game in
  laboratory - though care needed
• Clear subjects do not behave as Game Theory
  predicts
• Experiment enables a Theory of Learning to be
  constructed.