Economists and psychologists. Different empirical methods, reflecting their distinct focus
Economists: field studies
Psychologists: experimental studies
Types of Empirical Studies
Field Studies
Real-world setting. Observation
Preferred to analyze market behavior, policy impact and group dynamics
Experimental Studies
Controlled environment, manipulation of variables to test hypothesis
to study individual motivation and decision-making process
Either type
Between-subjects:
Compare different group of subject
each group with a different condition/taskks
e.g. group 1: immediate gratification, group 2: delayed gratification
Within-subjects:
Examine same subject under different conditions or tasks
e.g. Same subject for both immediate and delayed gratification
Preference for the Types of Empirical Studies
Economists rely more on field studies
Focus on Behavior
\(E:\) Study what they do rather than why they do?
\(P:\) focus on why people behave as they do
Group Behavior and Markets
\(E:\) prioritize studying group behavior over individual behavior
Challenges with experiments
\(E:\) impractical and unethical
e.g. govt may avoid experimenting with tax policies and subsidies
Overcome challenges of observational studies (manipulation of recent variables) by using sophisticated econometric techniques.
Field Studies in Economics and BE
Observe real-world decision made by individuals in natural settings
e.g.
Choice between energy-efficient and cheaper alternatives
Betting choices
Investment decisions
Consumers responses to promotional offers
1998 study in Israel: Day care: fine for coming late... Now... lateness has a price... you can purchase their time... market like incentives crowd out social preferences.
After the fine ended, the parents were habituated to coming late
Reason as to why wasn't known. As we didn't have the data for that
Advantages of Field Studies
High ecological validity: Results reflective real world behavior
No concerns about artificiality
Challenges of Field Studies
Conceptual Validity: may not always measure what they intend to measure
Confounds: results depend on two or more effects which cannot be isolated from each other
Early morning classes. Students are inattentive:
Due to content being boring?
Due to content being available somewhere else?
Due to feeling sleepy early in the morning?
Due to not having food in the morning?
You cannot isolate these causes, how do you know how much effect is being contributed by each?
Factors for choosing an inefficient alternative
Ignorance of future cost savings (not aware)
Disbelief or uncertainty about future savings
Cash constraints preventing higher upfront costs
Hidden costs related to more expensive appliances (e.g. maintainence)
Random decision or inability to grasp future advantages
The Rapprochement
Psychologists took interests in economics
Experimental economics is recognized, pioneered largely by Vernon Smith
Experimental Studies in Psychology and BE
Expt studies involving manipulating real or hypothetical scenarios to observe subject behavior
Advantages over field / observational studies
Greater control over variables
e.g. Subjects bet more against a shabbily dressed, diffident opponent than a confident professionally dressed one, even though outcomes are based on chance.
Types of rewards
Real rewards: motivate subjects to behave as in real life to increase predictive accuracy
Hypothetical rewards: "Imagine you are...", offers flexibility, enabling studies of an otherwise difficult to perform nature.
Kirby & Marakovic compared discounting under both situations:
For hypothetical rewards, discount rates were lower.
Design and Interpretation of Experiment
Experiments can reveal insights into factors that are impractical to study observationally
But economists criticize Behavioral experiments
Reason for Criticism
Flawed Experimental Design
Use of financial incentives
\(E:\) financial incentives ensure real-world like behavior
\(P:\) fin incentives might overshadow intrinsic interest
Evidence: mixed evidence
Use of Deception
Common in psychology
Justified
Directly telling the purpose of the experiment might influence the behavior
Experiments that would otherwise be not possible to observe
Problem: Subjects become cynical... need naive participants (e.g., freshman students)
Lack of Control
Insufficient control \(\implies\) confounded effects
e.g. Endowment effect
Issues with Interpretation of Experiments
Can lab findings be generalized to real-world behavior?
No! It is influenced by context... so extrapolation is difficult
Morals and Ethics
Nature and extent of scrutiny of one's actions by others ("I am being watched")
context ("not natural")
self-selection of the individuals making the decisions
the stakes of the game (large rewards vs small rewards will lead to different behaviors)
Experimental Economists are criticized for making broad claims based on limited data
Cherry-picking of results: report only ones that support your theory/purpose and don't report contradictions
Issues with Treatment of Assumptions
Relaxing one assumption can have a knock-off effects
Modelers need to take all assumptions as a set and check how many need to be modified so that the new set is self-consistent
Purpose of study: Field vs Experimental
Parameter Estimation
\(F:\) for estimating real world params (e.g. price elasticity of demand)
Advantage: high ecological validity
Theory testing
\(E:\) Test competing theories
Advantage: Control, replication and manipulation... Identify the best fitting theory
Extra-Laboratory experiments
Recently proposed
Hybrid approach incorporating elements of field studies too
Conducted in the field while controlling for various factors to investigate the effect
Without subject's awareness (ecological validity)
Follow-up to laboratory experiments (to highlight the drawbacks of lab experiments... helps us understand thevalidity)
This enables Larger Subject Pools and more diverse participation than lab experiments
Useful for evaluating the Effectiveness of policy in a real-world setting.