Neural Computation Underlying Individual and Social Decision-Making Ming Hsu Haas School of Business University of California, Berkeley Ming Hsu Haas School.

Neural Computation Underlying Individual and Social Decision-Making

Ming HsuHaas School of Business

University of California, Berkeley

Neesweek, 09.August 2004Forbes, 01.September 2002

The Big Picture

Human Behavior

Economics: formal, axiomatic, global

Psychology: intuitive, empirical, local

Neuroscience:biological, computational evolutionary

The Big Picture

Human Behavior

Economics: formal, axiomatic, global.

Psychology: intuitive, empirical, local.

Neuroscience:biological, circuitry, evolutionary.

Neuroeconomics

“A mechanistic, behavioral, and

mathematical explanation of choice that transcends [each field separately].”

- Glimcher and Rustichini. Science (2004)

The Big Picture

Human Behavior

Economics: formal, axiomatic, global.

Psychology: intuitive, empirical, local.

Neuroscience:biological, circuitry, evolutionary.

Neuroeconomics

Studies how the brain encodes and computes

values that guide behavior.

Allows us to improve models, design markets/AI, create new diagnostic tools

Tools That We Used

Special Populations Functional Magnetic Resonance Imaging (fMRI)

7

fMRI Scanner

fMRI: Changes in Magnetization

Basal State

Activated State

Agenda

• Individual Decision-Making– Ambiguity aversion– fMRI and brain lesion

• Sociopaths– Social preferences– Special population

• Take-aways

Simple Decisions: Blackjack

Simple Decisions: Blackjack

Stock?Bond?

Domestic?Foreign?

Stock?Bond?

Domestic?Foreign?

DiversifyThink long-term

More Complicated: Investing

Whether?Who?When?Where?

37% Rule (Mosteller, 1987)

“Dozen” Rule (Todd, 1997)

Complicated: Love/Marriage

Little knowledge of probabilities

Simple Complex

Most of life’s decisions

Precise knowledge of probabilities

Uncertainty about uncertainty?

Ellsberg Paradox

1961

Urn I: Risk

Most people indifferent between betting on red versus blue

5 Red5 Blue

?

Urn II: Ambiguity

Most people indifferent between betting on red versus blue

? ? ? ??? ???

10 - x Redx Blue

Choose Between Urns

Many people prefer betting on Urn I over Urn II.

? ? ? ? ??? ???

Urn II(Ambiguous)

Urn I(Risk)

Where Is The Paradox?

P(RedII)=P(BlueII)

P(RedII) < 0.5

P(BlueII) < 0.5? ? ? ? ??? ???

P(RedI) = P(BlueI)

P(RedI) = 0.5

P(BlueI) = 0.5

P(RedI) + P(BlueI) = 1

P(RedII) + P(BlueII) = 1

Urn II(Ambiguous)

Urn I(Risk)

Simple Complex

Verizonor

Deutsche Telekom

Jenniferor

Angelina

Not ambiguityaverse

Portfolio Weights: U.S., Japan, and U.K. Investors

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

U.S. Japan U.K.

Proportion of portfolio

CanadaGermanyFranceU.K.JapanU.S.

Verizon or Deutsche Telecom?

French & Poterba, American Economic Review (1991).

fMRI Experiment

Hsu, Bhatt, Adolphs, Tranel, and Camerer. Science. (2005)

fMRI Experiment

Hsu, Bhatt, Adolphs, Tranel, and Camerer. Science. (2005)

Expected Reward Region

€

y i, jt,v = α + β amb A(i, j, t) + β riskR(i, j, t)

+δE(i, j, t) + πW (i, j, t,v) + ε i, jt,v

y - Brain response A(.) - Ambiguity trialsR(.) - Risk trialsE(.) - Expected value of choicesW(.) - Nuisance parameters

Lower Activity under Ambiguity%

Sig

na

l Ch

an

ge

Region Reacting to Uncertainty

€

β amb > β risk

N.B. This region does not correlate with expected reward.

Orbitofrontal Cortex

€

y i, jt,v = α + β amb A(i, j, t) + β riskR(i, j, t)

+δE(i, j, t) + πW (i, j, t,v) + ε i, jt,v

y - Brain response A(.) - Ambiguity trialsR(.) - Risk trialsE(.) - Expected value of choicesW(.) - Nuisance parameters

Brain Imaging Data

Behavioral Choice Data Stochastic Choice Model

Link Between Brain and Behavior

Early

Late?

A Signal for Uncertainty?

Lesion Subjects

Orbitofrontal Control

Lesion Experiment

100 Cards

50 Red50 Black

100 Cards

x Red100-x Black

Choose between gamble worth 100 points OR

Sure payoffs of 15, 25, 30, 40 and 60 points.

Estimated Risk and Ambiguity Attitudes

Orbitofrontal Lesion

Control Lesion

Orbitofrontal lesion patients more rational!

Linking Neural, Behavioral, and Lesion Data

Brain Imaging Data

Behavioral Choice Data Stochastic Choice Model

Imputed value

OFC lesion estimate = 0.82

Agenda

• Individual Decision-Making– Ambiguity aversion– fMRI and brain lesion

• Sociopaths– Social preferences– Special population

How neurosciencecan help economics

How economics can help neuroscience

Norman BatesPsycho, 1960

Criminality

• Estimated psychopathy rates among prisoners (various times after 1990)– North American: 20.5% (2003 PCL-R

manual)– Canada: 15 – 25% (federal prison)– Iran: 23%– UK: 26%

• Younger beginnings (14 y.o. vs. 28 y.o. )• “Instrumental” homicides

Measuring Psychopathy

• Psychopathy Checklist-Revised, Screening version (PCL-R SV)– 24 point scale: 12 traits scored 0, 1, 2

• Two factors– Interpersonal-affective factor (6 traits)– Impulsivity-social deviance (6 traits)

• Impulsivity-social deviance (Factor 2) is less important for us– Except for safety concerns, of course!

Interpersonal-affective factor

• Callous and unemotional• Superficial charm• Grandiosity• Lack of empathy and shallow affect• Deception and manipulativeness• Lack of remorse• Not accepting responsibility

Characterizing Psychopathy using Economic Games

• What we’re doing– Characterize behavior in these individuals– Provide a quantitative measure of (social)

behavior

• Where we want to go– Use this measure to search for neural and genetic

correlates of psychopathy– And other psychiatric and neurological diseases

Responder Game

Your payoff

Other’s payoff

Your payoff

Other’s payoff

B: Costless punishment

Generous

Selfish

B: Costly Reward

Generous

Selfish

Responder Game: Intentions Matter

Responder Game: Intentions Matter

Power matters?SPs (only): Refuse to let Player B choose

Responder Game: Intentions Matter

Power matters

I would not give control over to another person, even for more money.

Responder Game: Intentions Matter

Power matters?

I would not give control over to another person, even for more money.

Seems like A1 is the more “dominant.”

Take-aways

• Neuroeconomics is possible– Studying neural mechanisms of economic decision-making– Nascent field, only about 10 years old– Much progress during that time

• Many open questions, opportunities– Moral decision-making– Strategic thinking– Financial bubbles– http://neuroecon.berkeley.edu

Eric Set

Edelyn Verona

Colin Camerer

Ralph Adolphs

Daniel Tranel

Steve Quartz

Peter Bossaerts

Meghana Bhatt

Cédric Anen

Shreesh Mysore

Acknowledgements