Summary
Summary Experimental Economics
- Course
- Institution
Summary of the course Experimental Economics. Written in English
[Show more]
Seller
Follow
nienkeweijers
Reviews received
Content preview
Summary Experimental EconomicsLecture 1: Methodology
1. What is Experimental Economics
A discipline in which data on economic phenomena are collected in a controlled laboratory or field
environment. Like theory and empirics: experimental economics is used to explain and/or describe
economic activity.
Example of Experiment
We are interested in how auction rules affect bidding. We are gonna discuss 3 experiments.
1) Putting up a hypothetical good for auction. Why would you want to buy a hypothetical good?
Because we make it valuable for you. Handing out a piece of paper to some people. This has a
number. This number is the value of the good to you. If you manage to buy the hypothetical good, I
will pay you the amount on your piece of paper.
Whoever wins the auction: pays her bid and receives the amount specified. The values handed out
differ from one student to another: they are private values. These values differ from one bidder to
another, nobody knows others’ values.
2) We will use the English auction. You may take turns to shout out a bid. Any bid must be higher
than the previous one. You may bid more than once. We stop when there is no more bidding.
3) Putting up for auction a box of chocolate, that is bought for 6 euro. Whoever wins the auction: pays
her bid and receives the chocolate. We will use auction 2 rules: you may shout out a bid.
Comparing the Experiments
What is the difference between auctioning hypothetical good and auctioning chocolate?
We don’t know how each one of you likes chocolate, so we cannot study the relationship between
value and bid. For example, in the hypothetical good, I gave you an induced value, ‘using a reward
structure to induce prescribed monetary value on actions’. This is an essential element of the
experimental control. It allows us to test price theory propositions conditional upon known
valuations. All that is needed for the analysis is the assumption that utility is a monotone increasing
function of the monetary reward.
What is the difference between auctions 1 and 2?
Different auction institutions: distinct theoretical predictions (lecture 2). We have date on bidding
behaviour with (controlled!) same values, so a direct comparison is possible. Controlling the values
and controlling the institutions allows to directly test the predictions of auction theory. This is not
possible with observational field data.
2. A Brief History
A first: Thurstone, 1931: indifference curves.
(1) Market experiments: Chamberlin, 1948: induced demand and cost structure; decentralized market
place too many transactions.
Vernon Smith, 1962, 1964: double auction, centralized bids, asks and transaction prices efficient,
competitive outcomes.
(2) Game experiments: Motivated by von Neumann and Morgenstern (1947), prisoners’ dilemma and
oligopoly games (Reinhard Selten, 1959).
(3) Individual choice experiments: Motivated by von Neumann and Morgenstern (1947), test
theories of choice under uncertainty (Savage, 1954) Allais paradox.
1
,3. The Experimental Method
What we do: we control institutions (market rules, etc.) and incentives (payoffs) and study behaviour.
Advantages compared to observational field studies:
1. Control (ceteris paribus) – Control the factors that may be affecting observed behaviour, so
that extraneous factors are held constant as the treatment variable changes. The most common
cause of loss of control is changing more than one factor at the same time, so that it is difficult
to determine the cause of any observed change in behaviour. Control is also lost when
procedures make it difficult to determine the incentives that participants actually faced in an
experiment.
So control induce preferences, obtains additional measures and gives possibility of
randomization.
2. Replication – The ability to repeat the same setup numerous times in order to determine
average tendencies that are relatively insensitive to individual or group effects.
Goals:
1. Theory testing/selecting – Example is the prisoner’s dilemma. Theory says that the N.E. is
(D,D), in the experiments subjects often play (C,C);
2. Stress testing – Example: Assume that we observe that the market equilibrium is obtained in
experimental market with perfect information about transaction prices. Stress test: is
equilibrium obtained without price information?
3. Search for empirical regularities – Example: We often observe that subjects follow social
norms, but we have no theory Collect data to find regularities that might help us to develop
a theory.
4. Advice policy makers – Examples:
Test mechanism to auction spectrum rights;
Test ways to regulate a privatized electricity market.
the laboratory as a wind-tunnel.
4. Methodological Discussions
Until recently, methodological discussion scarce in experimental economics and large consensus on
important issues. In the last 15 years, increasing interest in methodological issues.
The Social Situation ‘Experiment’
A major problem is what do subjects think when participating in an experiment. In experimental
economics, standards of experimentation have been developed aimed at reducing participants’
uncertainty and specific precisely the game or contest.
In psychology other standards hold. This has led to various disputes.
Disputed Issues
1. Monetary incentives (dependent on decisions)
Almost all economists use performance based financial incentives in experiments. Only 15-
25% of psychologists use performance based financial incentives in experiments. Monetary
incentives could:
Induce preferences;
Changes in decision prominent effect on rewards (saliency)’
Dominance (reward outweighs other motivations like boredom or trying to help the
experimenter);
Is easier to implement than alternative incentives;
2
, No satiation, compared to chocolate;
Most economic experiments are about economics, scarce resources;
Optimal behaviour in psychology experiments not as easily defined, how do you
reward subjects in experiment on mate choice?
Test shows:
Subjects appear to participate more seriously with incentives;
More consistency;
Less ‘sociably desirable’ behaviour;
Effect on average decisions disputed, subject make more effort sometimes better
performance.
2. Deceiving subjects
Experimental economists have a ‘ban’ on deceiving subjects. Psychologists have often
discusses the merits and dangers of using deception. American Psychological Association’s
ethical guidelines proposes to deceive only as a last resort strategy. Yet, approximately half
of psychologists’ experiments use deception.
Why use deception?
1) Subjects might guess the study’s purpose and respond to that. This is more of an issue
with psychology experiments than with economics;
2) Deception might be used to create situations of special interest, that will not occur
‘naturally’ very often.
3. Role of Questionnaires
Questionnaires are hypothetical. Economists cannot learn much from costless choices (no
allocation of scarce resources involved). In Experimental Economics questionnaires are
usually only used for background information.
4. Choice of Wording (abstract or concrete)
If you use words like ‘manager of a firm’, subjects might behave according to how they
perceive a manager should behave. You do not know their perception and you lose control
over incentives and information. Experimental economists give precise instructions describing
the possible actions and their payoff consequences, not ‘roles’ to be played script enactment.
This is that experimental economists provide participants with scripts (instructions) that
supply descriptions of:
The players;
Their action choices;
The restrictions and rules they face;
The possibly payoffs resulting from actions.
And ask participants to enact these scripts.
Script Enactment
Psychologists typically do not provide participants with a script enact (subjects have to ‘guess’ what to
do; no specified roles). Participants are forced to ad-lib. The way they ad-lib can easily be influenced
by their expectations of what the experimenter is looking for.
Social psychologists sometimes use role playing. In some cases:
Roles actually played out;
Scripted;
Behaviour studied (as opposed to statements).
3
,This is close to economists’ script enactment. Difference: in economics experiments, participants do
not simulate real agents but are real agents whose choices have tangible economic consequences for
them.
Supplying a script for participants to enact can make important difference in results obtained. Script
provision demands more elaborate and transparent instructions, but reduces ambiguity of
experimental situation. So it increases researchers’ control over participants’ interpretation and the
replicability of results.
Repeated Trials
A typical experiment in economics: participants make a decision in the same stationary environment
multiple times (for example: play a game 10 time with varying partners). Two reasons:
1. Give participants chance to adapt to the environment;
2. Give participants chance to learn about others.
Underlying both: interest in equilibrium situations, but equilibrium is assumed not to be reached
immediately. No focus on any type of equilibrium.
Two ways to implement repeated trials:
1. Stationary replication of one-shot decisions:
Individual decision making: repeat decision;
Games (strategic interaction): strangers.
2. Repeated games: partners match repeatedly (allows for strategic behaviour).
In both cases: environment (and often, parameterization) remain constant: allows for learning.
The Artificiality of a Laboratory
Often heard question: Do laboratory results care over to the ‘real world’? = What can we learn from
the behaviour of students in such an artificial setting? = 986 variations of the same question = What is
the external validity of this research?
Internal vs. External Validity
Internal validity: The extent to which conclusion follow logically and consistently from research.
External validity: Possibility to generalize conclusions to situations that prompted the research.
Tension: Interval validity requires abstraction and simplification to make research tractable, but this
decreases external validity.
So theory, high score on internal, low score on external. Experiments show, high score on internal,
high/low score on external. This depends on topic. Field data shows low score on internal, high score
on external. Field experiments have a position in between.
Artificiality:
- Major obstacles to external validity;
- If laboratory institutions/incentives insufficiently mirror the object of study, significant loss of
external validity.
Why do economists focus more on internal validity?
- Deductive science;
- Tradition of modelling;
- Theory basis for empirics.
4
,Review of the Discussion
Artificiality an issue from the beginning because of ‘attacks’. Charlie Plott (early 1980s) link between
experiments and theory claimed artificiality cannot be falsified. Vernon Smith (since the 1970s)
defines ‘precepts’ for internal and external validity shifts towards more attention for external validity.
Vernon Smith’s precepts:
1. Non-satiation: more reward is always better;
2. Saliency: the amount of an individual’s reward is linked to the decisions the individual makes;
3. Dominance: the reward structure dominates any subjective costs;
4. Privacy: each subject only has information about the own payoffs.
Smith: these 4 precepts are sufficient to create a small scale microeconomic environment where real
economic agents make real decisions. This leads to an internally valid situation to test economic
theories.
5. Parallelism: the propositions that are being tested in the laboratory are applicable to non-
laboratory microeconomics where similar ceteris paribus conditions hold.
Parallelism is needed for external validity. So, Smith:
- For testing formal theories, precepts 1-4 suffice;
- (early Smith:) it is up to critics to falsify parallelism;
- (later Smith:) experimentalists should not only test theory but should pay more attention to
parallelism.
Procedural and Design Issues
Definitions:
Session: sequence of tasks with group of subjects, 1 day
Treatment: unique environment with respect to parameters of interest
Design: a specification of sessions in treatments to evaluate research question
Experiment: collection of sessions to evaluate research question
5. Theory and Experiments
Three types of theory most commonly used:
- Decision Theory – Expected utility maximization;
- Non-cooperative Game Theory – Nash equilibrium;
- Price Theory – Market equilibrium.
5
, Lecture 2 – Industrial Organization
1. Market Types Studied in the Laboratory
Goal: understanding of market institutions. There are different types:
1. Double Auction:
Trading pit (Chamberlin) – decentralized, poor performance;
Centralized (Smith) – public information, repetition, competitive outcomes.
2. Posted Offer Market – company posts take-it-or-leave-it price;
3. Cournot Competition;
4. Bertrand Competition.
Most markets are posted offer. Why?
Transaction costs (reliance on clerks);
Government regulation;
Higher prices/profits.
Both types can be studied in the laboratory. Double Auctions allow for richer analysis of competitive
equilibrium.
Instructions
Instructions need to:
Explain induced values;
Explain trading institution.
In this case, common to use words like: buyer, seller, price, trade, market.
No suggestions as to type of behaviour expected (like ‘making trades’). Details are important.
Instructions should enable replication.
2. Double Auction Markets
Information
Important choice: what information to provide. Complete information (payoff of others) is not
needed. Trading prices are usually common information. Bid and ask queues:
Usually only highest bid, lowest ask;
Sometimes complete queues;
Sometimes nothing;
Market theory does not specify criteria (it is not game theory).
Procedures
Typical procedures:
Sequence of trading intervals (periods);
Each period lasts preset time (3-5 minutes);
In each period: new valuations (private info);
Bidders buy high value units first;
Sellers sell low cost units first;
Buyers place bids, sellers place asks;
Subjects can accept bid/ask or place new one;
No unprofitable trades allowed (?);
6
The benefits of buying summaries with Stuvia:
Guaranteed quality through customer reviews
Stuvia customers have reviewed more than 700,000 summaries. This how you know that you are buying the best documents.
Quick and easy check-out
You can quickly pay through credit card or Stuvia-credit for the summaries. There is no membership needed.
Focus on what matters
Your fellow students write the study notes themselves, which is why the documents are always reliable and up-to-date. This ensures you quickly get to the core!
Frequently asked questions
What do I get when I buy this document?
You get a PDF, available immediately after your purchase. The purchased document is accessible anytime, anywhere and indefinitely through your profile.
Satisfaction guarantee: how does it work?
Our satisfaction guarantee ensures that you always find a study document that suits you well. You fill out a form, and our customer service team takes care of the rest.
Who am I buying these notes from?
Stuvia is a marketplace, so you are not buying this document from us, but from seller nienkeweijers. Stuvia facilitates payment to the seller.
Will I be stuck with a subscription?
No, you only buy these notes for $8.59. You're not tied to anything after your purchase.
Can Stuvia be trusted?
4.6 stars on Google & Trustpilot (+1000 reviews)
52510 documents were sold in the last 30 days
Founded in 2010, the go-to place to buy study notes for 14 years now