ECON 282: Introduction to Experimental Economics

1
ECON 282 Introduction to Experimental Economics

• Professor Graeme Walker
• office WMC 2696
• office hours 300-400 (Monday) or
  by appointment
• email gmw1_at_sfu.ca

2
Taking a course in experimental economics is a
little like going to dinner at a cannibal's
house. Sometimes you will be the diner, sometimes
you will be part of the dinner, sometimes both.
Bergstrom and Miller
3
Brief Introduction

• If you take a laboratory course in the physical
  sciences, you get to mix smelly chemicals, or
  monkey with pulleys, or dissect a frog, but you
  are always the experimenter and never the subject
  of the experiment. In the market experiments
  conducted in this class, you and your classmates
  will be the participants in the markets as well
  as the scientific observers who try to understand
  the results.
• It is hard to imagine that a chemist can put
  herself in the place of a hydrogen molecule. A
  biologist who studies animal behavior is not
  likely to know what it feels like to be a duck.
  You are more fortunate. You are studying the
  behavior and interactions of people in
  economically interesting situations. And as one
  of these interacting economic agents, you will be
  able to experience the problems faced by such an
  agent first hand. We suspect that you will learn
  nearly as much about economic principles from
  your experience as a participant as you will from
  your analysis as an observer.
• Bergstrom and Miller

4
Course Outline

• The goal of this course is to learn about the
  experimental methodology, as well as some of the
  analysis tools needed to evaluate the outcomes of
  experiments
• In doing this, you will be exposed to a wide
  range of topics in economics including
• Markets and market design
• Game theory
• Econometrics
• Social economics (e.g. identity, learning,
  networks)
• Crime and Punishment
• Development Economics

5
Course Outline

• This is not an exhaustive list of topics, but is
  meant to you a bit of an idea as to what to
  expect in 300- and 400-level course that focus
  solely on one of these areas.
• Incomplete lecture slides will be available on
  the course webpage
• You should print these out before coming to the
  lecture (try printing 4 slides to a page
• Pre-requisites are ECON 103 and ECON 105, so it
  is assumed that students have a basic
  understanding of economics.

6
Grading

• Participation 10
• There will be 4 experimental sessions run in
  tutorials
• Attendance will be taken
• The full participation grade will be given
  randomly for one of the experimental sessions
• This forces you to come to all experimental
  sessions
• If you know you cannot make it to one of the
  tutorials, come see me during my office hours
• Assignments 20
• There will be 2 assignments of equal weight
• Assignments will involve data analysis
• You can work in groups of up to 3 members
• Midterm 30
• Tentatively set for June 27th
• Final Exam 40
• August 13th

7
Course Schedule
Date Lecture Tutorial
May 9 Introduction NO TUTORIALS
May 16 Experimental Design NO TUTORIALS
May 23 NO CLASSES Experiment 1 Markets
May 30 Data Analysis Qualitative NO TUTORIALS
June 6 Data Analysis Quantitative Experiment 2 Public Good
June 13 Markets How to use excel, Sample questions
June 20 Public Goods ASS1 due
June 27 MIDTERM NO TUTORIALS
July 4 Game Theory Experiment 3 Ultimatum
July 11 Social Preferences Experiment 4 Crime
July 18 Social ID Sample questions
July 25 Crime ASS 2 due
August 1 NO CLASSES NO TUTORIALS
August 8 Development NO TUTORIALS
8
Textbooks

• Textbooks are recommended not required
• There will be a number of papers covered in
  lecturesthe relevant parts should be read
• I will post what is relevant and from which
  article

9
Now its Time for our 1st Experiment!
10
Experimental Instructions

• Players
• 1 and 2
• Actions
• each player chooses whether to attend a hockey
  game or the ballet
• Payoffs
• Player 1
• If you choose hockey and the other player chooses
  hockey, you receive 3 pts
• If you choose hockey and the other player chooses
  ballet, you receive 1 pt
• If you choose ballet and the other player chooses
  ballet, you receive 3 pts
• If you choose ballet and the other player chooses
  hockey, you receive 0 pts
• Player 2
• If you choose hockey and the other player chooses
  hockey, you receive 3 pts
• If you choose hockey and the other player chooses
  ballet, you receive 0 pts
• If you choose ballet and the other player chooses
  ballet, you receive 3 pts
• If you choose ballet and the other player chooses
  hockey, you receive 1 pt

11
Payoff Matrix
Player 2 Player 1 Hockey Ballet
Hockey 3 3 1 1
Ballet 0 0 3 3
12
Any guesses on the name of this game?
13
Battle of the Sexes
14
What was the point of this experiment?

1. To see if people prefer hockey to ballet?
2. To see if people can coordinate?
3. To see who is more likely to coordinate?
4. Did people learn to coordinate?
5. To see if people behave the way John Nash
   theorizes they should?

15
How do we answer these questions?

• Need some basic statistical tools
• Need to understand the design of the experiment
• In order to understand how an experiment is
  designed, we first need to
• Define what an experiment is
• List the ingredients to an experiment

16
What is an experiment?

• An experiment is an effective procedure for the
  discovery of, and selection between, different
  possible explanations that are of equivalent or
  greater or lesser importance to us

17
What is the goal of experimental economics?

• The goal of experimental economics is an
  increased understanding of real world phenomena
  by designing effective experiments to
  systematically break model assertions

18
Why are experiments useful?
19
Why are experiments useful?

1. Refine theories
2. Construct new theories
3. Design/redesign/wreck institutions
4. Test predictions

Suggests/modifies
EMPIRICS
THEORY
Tests/modifies
20
Some Reasons Economists Run Experiments

• Test a theory or discriminate between theories
• Public goods and free-riding
• Social identity and favouritism
• Explore the causes of a theorys failure
• Learning
• often theories that initially perform poorly show
  improvement if subjects get more experience
• Subject Motivation
• often theories show improvement when payoffs are
  increased

21
Some Reasons Economists Run Experiments

• 3. Compare institutions
• Dutch vs. English Auction
• 4. Evaluate policy proposals
• Corruption monitoring ? top-down vs. bottom-up
• 5. The lab as a testing ground for institutional
  design
• Smith and market efficiency

22
What are the ingredients to an experiment?

• Subjects
• Often undergrad students (e.g. CalTech)
• People from developing countries
• Policy-based experiments
• Cheaper, lower opportunity cost
• Environment
• Initial endowment
• Preferences
• Costs that motivate exchange
• Institutions
• Market communication
• e.g. bids, offers, acceptances
• Rules that govern exchange of information
• Rules under which messages become binding

Observed behaviour of participants constitute the
controlled variables
Controlled by monetary rewards
Defined by experimental instructions
23
What were the ingredients to our experiment?

• Subjects
• 20 students from ECON 282
• 10 male, 10 female
• Environment
• Player 1(2) weakly preferred hockey (ballet)
• Did not give monetary incentive (Big Problem)
• Institutions
• Players were paired up and remained in these
  pairs for both periods in the experiment
• Players saw who their partner was
• Players simultaneously chose either hockey or
  ballet
• Players were not allowed to communicate between
  periods

24
How do these ingredients allow us to answer our
questions?
25
1. Which activity do people prefer?

• This was not the goal of the experiment, as
  subjects were to have had their preferences
  induced by the associated payoffs
• Is there a problem with labeling the two choices
  as hockey and ballet? YES!

26
1. Which activity do people prefer?
27
2. Can people Coordinate?

• For this game, coordinating is considered when
  both players either choose hockey or ballet
• How should we define coordinate?
• If one group coordinates at least once in either
  round?
• If there are more instances of groups
  coordinating than not (in either round)

28
2. Can people Coordinate?
29
3. Were there groups that were more or less
likely to coordinate?

• Need to compare coordination across treatments
• Female-male
• Male-male
• Female-female

30
3. Were there groups that were more or less
likely to coordinate?
31
4. Did people learn to coordinate?

• Understanding learning involves comparing
  behaviour over time
• In our experiment, this means comparing
  coordination between the 1st round and the 2nd

32
4. Did people learn to coordinate?
33
5. Was Nash correct?

• Nashs theory says that players should either
  both play hockey or both play ballet or play some
  mixture of the two, with the preferred activity
  played more often than the other
• Previous experiments suggest that after many
  rounds of play, subjects converge to a pattern of
  coordinating on both hockey and ballet
• What do the results from our experiment indicate?
• Not enough observations to make a statistically
  significant comment

34
What are some General things Economists have
Learned from Experiments?

• Institutions matter
• Because information and incentives matter
• The Endowment Effect
• People value a good more once their property
  right to it has been established
• This is inconsistent with standard economic
  theory that says a persons willingness-to-pay
  for a good should equal their willingness-to-accep
  t compensation to be deprived of the good

35
What are some General things Economists have
Learned from Experiments?

• 3. Common information is not sufficient to yield
  common expectations or knowledge
• There is no assurance that publicly announcing
  instructions will yield common expectations among
  players since each person may still be uncertain
  about how others will use the information
• As subjects gain experience, they tend to foster
  common expectations
• 4. Information less can be better
• e.g. continuous double auction private vs.
  complete and common information
• Private information treatments converge to the
  equilibrium outcome quicker and more dependently
• This is because when people have complete
  information they can identity more
  self-interested outcomes than competitive
  equilibria and use punishing strategies in an
  attempt to achieve them, which delays reaching
  equilibrium

36
What are some General things Economists have
Learned from Experiments?

• 5. Dominated strategies are for playing, not
  eliminating
• Prisoners dilemma it is optimal for both to
  cooperate, but it is dominant for both to defect
• 6. Fairness taste or expectation
• People think it is unfair for firms to increase
  price unless it is because costs increase
• But theory predicts price can increase due to
  shortages (caused by an increase in demand)
• e.g. price of shovels increase after a blizzard

37
What are some General things Economists have
Learned from Experiments?

• 7. Unconscious optimization in market
  interactions
• Economic agents can achieve efficient outcomes
  which are not part of their intentions
• Cannot rule out the possibility the unconscious
  is a better decision make than the conscious

38
Now its Time for our 2nd Experiment!
39
Experimental Instructions

• Think long and hard at how far it is (in km) from
  Vancouver to HCMC
• Write your best guess down on a piece of paper
  (and hand it in)
• Now, take your next best guess, write it down,
  and hand it in

40
ANSWER

• Approximate distance from Vancouver Canada to Ho
  Chi Minh City Vietnam is 11763 km

41
Results to be announced at the beginning of next
class.