You are on page 1of 57

American Economic Association

Neuroeconomics: How Neuroscience Can Inform Economics Author(s): Colin Camerer, George Loewenstein, Drazen Prelec Source: Journal of Economic Literature, Vol. 43, No. 1 (Mar., 2005), pp. 9-64 Published by: American Economic Association

Accessed: 22/05/2011 23:58

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=aea.

.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Economic Association Neuroeconomics: How Neuroscience Can Inform Economics Author(s): Colin Camerer, George Loewenstein, Drazen Prelec, Vol. 43, No. 1 (Mar., 2005), pp. 9-64 Published b y : American Economic Association Stable URL: http://www.jstor.org/stable/4129306 . Accessed: 22/05/2011 23:58 Your use of the JSTOR archive indicates y our acce p tance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the p ublisher re g ardin g an y further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=aea . . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to Journal of Economic Literature. http://www.jstor.org " id="pdf-obj-0-54" src="pdf-obj-0-54.jpg">

American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to Journal of Economic Literature.

http://www.jstor.org

Journalof EconomicLiterature Vol.XLIII (March2005), pp. 9-64

Neuroeconomics: How

Neuroscience

Can

Inform

Economics

COLIN CAMERER,

GEORGE LOEWENSTEIN,

and DRAZENPRELEC*

Who knowswhat I want to do? Who knowswhat anyone wantsto do? How can

be sure about back and going something is
be sure about
back and
going
something is

something

forth,

like that? Isn't it

energy

in

electrical

you

really brain. Some minor one of the brain

what

want to

all a

question

of

you

brain chemistry,signals

you

know whether

the cortex? How do

some kind of

do or just

nerve impulse in the in

unimportantplace

little activity takes place

somewherein this

hemispheres

and suddenly I want to go to Montanaor I don'twant

to go to Montana. (WhiteNoise, Don DeLillo)

1. Introduction

In the last two decades,

following almosta

century of separation, economics has begun

to

import

insights

from

psychology.

prominent

"Behavioral

fixture on the

economics" is now a

intellectual landscape and has

in

economics,

spawnedapplications to topics

*

Camerer:

California

Institute

Mellon

Technology.

of

Technology.

Loewenstein: Carnegie Massachusetts Institute of

University. Prelec:

We

thank partici-

confer-

1997) at

on Neural

pants at the Russell Sage Foundation-sponsored

ence on Neurobehavioral

Carnegie

Mellon, the

Economics

(May

Princeton workshop

Economics December 8-9, 2000, and the Arizona confer-

ence in March 2001. This research was

supported

by NSF

grant SBR-9601236 and by the Center for Advanced Study

in Behavioral Sciences, where the authors visited during

1997-98. David Laibson's presentations have been partic-

ularlyhelpful, as were comments and suggestions from the

editor and referees, and conversationsand comments from

Ralph Adolphs, John Allman, Warren Bickel, Greg Berns,

Meghana Bhatt, Jonathan Cohen, Angus Deaton, John Dickhaut, Paul Glimcher, Dave Grether, Ming Hsu, David

Laibson, Danica

Mijovic-Prelec, Read Montague, Charlie

Rangel, Peter

Shizgal, Esther
Shizgal,
Esther

Plott, Matthew Rabin, Antonio

Steve Quartz, and Paul Zak. Albert Bollard,

Hwang, and Karen Kerbs provided editorial assistance.

such as finance, game theory, laboreconom-

ics, public

(see Colin

2004).

informed

finance,law,

Camererand

and macroeconomics Loewenstein

George

Behavioraleconomics has mostly been

by

a branch of

psychology

called

"behavioraldecision research," but other cognitive sciences are ripe for harvest.Some

importantinsights will

roscience,

either

science will

surely come fromneu- or because neuro-

directly what is believed about

reshape

psychology which

in turninforms economics.

brain activity

Neuroscienceuses imaging of

techniques how the brainworks. The brainis the ultimate

and other

to infer details about

"blackbox." The foundationsof economic

theory were constructed assuming that details

aboutthe functioning of the brain'sblack box

would not be known. This

pessimism was

expressedby William Jevons in 1871:

  • I hesitateto say thatmen will everhave the means

of measuring directly the feelings of the human

heart. It is from the

quantitative effects of the

feelings that we must estimate their comparative amounts.

9

  • 10 Journal of Economic Literature, Vol. XLIII (March 2005)

Since

feelings were meant to predict behav-

only

be assessedfrom

behavior,

ior but could

economists realized that, without direct

measurement,

vening

cepts

feelings

were useless inter-

constructs. In

of

ordinal

the 1940s, the con-

utility

and revealed

superfluous inter-

preference eliminatedthe

mediate

step

ings. Revealed

equates

unobserved

of positing immeasurablefeel-

preference theory simply

preferences

with

by

avoided

behave consistently,

observed choices. Circularity is

assuming

which

that

people

makes the

theory falsifiable; once

they prefer

A to B,

choose B

they have revealed that

people

over A.

should not

Later

subsequently

extensions-discounted,

expected, and subjective expected utility,

and

"as if"

Bayesian updating-provided tools which

long a black box. The

not

be

in other human sciences.

now

neuroscience

has

similar

sidesteppedpsychological

approach

made

good

detail. The "as if"

sense as

tially

as the brainremained substan-

development of eco-

held

hostage to

proved

the

nomics could

progress

But

Jevons's pessimistic prediction wrong;

study of the

beginning

thoughts

are,

in

brain and nervous

system

is

to allow direct measurement of

and feelings. These measurements

turn,

challenging our understanding

action,

of the relation between mind and

leading calling old ones into

new

findings

ries they

ic

to new theoretical constructs and

question.

How can the

and the theo-

of neuroscience,

spawned,

have

that

inform an econom-

theory

developed so impressively in

their absence?

In

thinking

about the

that neuro-

it is useful to

ways

science can

inform economics,

distinguish two types of contributions, which we term incrementaland radical approach-

es. In the incremental science adds variables

approach,

to

neuro-

conventional

accountsof decision making or suggestsspe- cific functional forms to replace "as if"

assumptions that have never been well sup- ported empirically. For example, researchon the neurobiology of addiction suggests how

drug future
drug
future

consumption

limits

pleasure from

consumption of other goods (dynam-

commod-

ic cross-partial effects in utility for

ity bundles) and how environmentalcues

trigger unpleasant

craving and increase

can be approximated

theory

and

then

(see

Douglas

demand.These effects

by extending

applying

standard

conventional tools

Bernheim and Antonio Rangel 2004; David

Laibson

2001;

Ted

O'Donoghue and

involves

turning

eco-

asking how

differently if it

insights

neuro-

argue,points

MatthewRabin 1997). The radical

approach back the hands of time and

nomics

might

have evolved

had been informedfrom the start by

and

findings now available from

Neuroscience,we will

science.

to an entirely new set of constructsto under-

lie economic decision making. The standard

economic

theory of

constrained utility maxi-

mizationis most as the result of

naturallyinterpreted either

learning based

on

consump-

tion experiences (whichis of little help when

prices,

income,

and

opportunity

sets

change),

ing

options-as

or careful deliberation-a

balanc-

of the costs and benefits of different

might

characterize

complex

buy-

planning for retirement,

hammering

out a contract.

decisions like

ing

a house, or

Although economists mayprivately acknowl-

edge

beings

tion,

that actual flesh-and-blood human often choose without much delibera-

the economic models as written invari-

ably represent

decisions in a "deliberative

where

equilibrium," i.e., that are at a stage

further

deliberation, computation, reflec-

by

itself alter the

agent's

tion, etc. would not

choice. The variablesthat enter into the for-

mulationof the decision

problem-the

pref-

erences, information, and constraints-are precisely the variablesthat should affect the

decision, if the person had unlimited time and

computingability. While not denying that deliberationis part

of human decision making, neuroscience

points out two generic inadequacies of this approach-its inability to handle the crucial roles of automaticand emotional processing.

Camerer,Loewenstein, and Prelec: Neuroeconomics

11

First, much of the brain

"automatic" processes,

implements

which are faster than

conscious deliberations and which occur

with little or no awarenessor feeling of effort

(John Bargh

et al. 1996;

Bargh

and

Tanya

and

Chartrand 1999; Walter Schneider

Richard Shiffrin 1977;

Shiffrin and

Schneider 1977). Because people have little

introspective access to these processes, or volitional control over them, and these

or no

were evolved to solve

problems of

processes

evolutionaryimportance ratherthan respect

logical dicta,

the behavior these

processes

generate of inference and choice. Second, our behavior is

enced by finely

systems

need not follow normativeaxioms

strongly influ-

(emotion)

tuned affective

whose basic design is common to

(Joseph

LeDoux

Edmund Rolls

daily

or

humans and many animals

1996; Jaak

1999). These

Panksepp

systems

1998;

are essential for

functioning, and when they are damaged

perturbed, by

ances in

brain

injury,

stress,

imbal-

neurotransmitters, or the "heat of

the

moment," the logical-deliberativesys-

if

completely

intact-cannot

a fluid

tem-even

regulate behavior appropriately. Human behavior thus

requires interactionbetween controlledand automat-

ic

processes, and between cognitive

systems.

However,

many

and

affective

that

behaviors

from this interplay are routine-

emerge

ly and

uct of

falsely interpreted as being the prod- deliberationalone

cognitive

Michael Miller, and

(George

Michael

Wolford,

Gazzaniga

which

2000). These results (some of

suggest

that

for choice

are described below)

introspective accountsof the basis

should be takenwith a grain of salt. Because

automatic

processes

are

designed to keep

behavior"off-line" and below consciousness,

we have far more

introspective access to

controlled than to automatic processes. Since we see only the top of the automatic

iceberg, we

naturally tend to exaggerate the

importance of control.

Neuroscience

findings and methods will

undoubtedly play an increasingly prominent

role in economics and other social sciences

(e.g.,

law, Terrence Chorvat, Kevin McCabe,

and VernonSmith 2004). Indeed, a new area

of economics that has been branded "neu-

roeconomics"has

already

formed the basis

of numerousacademic

brought neuroscientists

together

(see

gatherings that have

and economists

press).1

also Paul Zak, in

Participating in the development of a shared

intellectual

enterprise

will

help

us ensure

that the neuroscience informs economic

questions

authors'

such

we care about. Stimulated by the

own participation in a number of

our

goal

in this

paper is to and how

meetings,

describe what neuroscientistsdo

their discoveriesand views of humanbehav-

ior

might

inform economic

analysis.

In the

next section (2), we describe the

diversity of

3

tools that neuroscientists use. Section

introducesa

simplified account of how cog-

nition and affect on the one hand, and auto-

matic and controlled processes

on the other,

work separately, and interact. Section 4 dis-

cusses general Section 5 goes

implications for economics.

into

greater

detail about

top-

implications of neuroeconomicsfor four

ics in

sion

economics: intertemporal choice, deci-

making

under

risk,

game theory,

and

labor-marketdiscrimination. Through most

of the paper, our focus is

largely on

how neu-

roscience can inform models of microfoun- dations of individual decision making. Section 6 discusses some broader macro

implications and concludes.

2. NeuroscienceMethods

Scientific

technologies are not just

tools

scientists use to

explore New tools also define new scientific fields

areas of interest.

and erase old boundaries.The

ated astronomyby elevating the

telescope

cre-

science from

1 The first meeting was held at Carnegie Mellon in

1997.

Later

meetings

were

held

in

Arizona and

Princeton, in 2001, in Minnesota in 2002, and on Martha's Vineyard in 2003 and Kiawah Island in 2004.

Occasional sessions devoted to this rapidly growing topic

are now common

at large annual meetings

economics and neuroscience.

in both

  • 12 Journal of Economic Literature, Vol. XLIII (March 2005)

pure cosmologicalspeculation.

made

scope

ogy.

possible

The same is

The micro-

similaradvances in biol-

true of economics. Its

constantlyreshaped by

boundarieshave been

tools such as mathematical,econometric, and

simulation methods. Likewise, the current

surge of interestin neuroscience by

ogists emerged largely

and the methods

psychol-

from new methods,

may productively blur the

psychology.

boundaries of economics and

This section reviewssome of these methods.

2.1

Brain Imaging

Brain imaging is currently the most popu-

lar neuroscientifictool. Most brain imaging

involves a

comparison of people performing

"experimental" task and

difference between

different tasks-an

a "control"task. The

images

two

takenwhile

subject is performing the

of

regions

of the

tasks provides a picture

brain that are

differentially activated by the

experimental task.

The oldest,

There are three basic imaging methods.

electro-encephalogram (or EEG)

scalp

to meas-

uses electrodesattached to the

ure electrical

activitysynchronized to stimu-

(known as

ERPs). Like

lus events or behavioral responses

Event Related Potentials or

EEG, positron emission topography (PET)

scanning

is an old

technique

in the

rapidly

is

changing time-frameof

still a useful

technique.

neuroscience, but

PET measuresblood

flow in the brain, which is a reasonable proxy

for neural

activity, since neural activity in a

region leads to increasedblood flow to that

region.

ular,

The newest, and

currently most pop-

imaging method is functional magnetic

imaging

(fMRI), which tracks

brain usingchanges in mag-

due to

blood

oxygenation

resonance

blood flow in the

netic

properties

(the "BOLD signal"). Simultaneousdirect

recording

responses

of neural

processing

and fMRI

signal

confirms that the

BOLD

reflects input to neuronsand their processing

(Nikos Logothetis et al. 2001).

Although

the

fMRI is

increasinglybecoming

each

method has

method of choice,

pros and cons. EEG has excellent temporal

resolution (on the order of one millisecond)

and is the

only

method used with humans

monitors neural

blood flow. But

only

activity, as

spatial reso-

measureactivi-

that directly

opposed to, e.g.,

lution is

ty

in

the

poor

and it can

part

outer

of the brain. EEG

improving

resolution has, however, been

through

of

the use of

ever-increasing numbers

electrodes. Interpolation methods, andthe

combineduse of EEG and fMRI for measur-

ing

outer-brain signals and inner-brain sig-

promise

to create

nals at the same time,

statistical methods which make reasonable

inferences

about activity throughout the

signals.

EEG

For

economics, a

is its relativeunob-

brain from EEG

majoradvantage of

trusiveness and

eventually reach

possible to

from

people

portability.Portability will

the

point

where it will be

take unobtrusivemeasurements

as

they go

about their

daily

provide better spatial

poorer temporal

neurally

affairs.PET and fMRI

resolution than EEG, but

resolution because blood-flow to

active areasoccurs with a stochastic lag from

a few seconds (fMRI) to a minute (PET).

Brain imaging

of brain

still

provides

only a crude

snapshot

are

activity. Neural processes

a 0.1 millimeter

thought

to occur on

scale in 100 milliseconds (msec), but the

spatial and temporal

scanneris

onds.

only

aged

constrains

the

continueto

combine the

resolution of a

typical

3

millimetersand severalsec-

Multiple trials per subject can be aver-

to form

compositeimages, but doing so

experimental design.

However,

technology has improvedrapidly and will

improve.Hybrid techniques that

strengths

of different methods

for

brains

are particularlypromising. Techniques

simultaneously scanning multiple

("hyperscanning") have alsobeen developed,

which can be used

level differences in

to study multiple-brain-

activity

in

games

al. 2002).

and

markets (Read Montague et

2.2 Single-Neuron Measurement

Even the finest-grained brain imaging

techniques only measure activity of "cir-

cuits" consisting of thousandsof neurons. In

Camerer,Loewenstein, and Prelec: Neuroeconomics

13

single

neuron

measurement, tiny electrodes

measuring a

below,

are insertedinto the brain, each

single neuron's firing. As we discuss

single

neuron measurement studies have

produced

believe,

some

striking findings

that, we

are relevantto economics. A limita-

tion of

because

single

neuron measurementis that,

damages

insertion of the wires

neurons, it is

Studying

largely restrictedto animals.

animals is informative about

many

brain structuresand

humans because

functionsof non-humanmammals are simi-

lar to those of humans

(e.g.,

we are more

of mon-

species).

genetically similarto many species

keys

than those

species

are to other

Neuroscientists commonly divide the brain

into crude

of

regions

that reflect a combination

evolutionarydevelopment, functions, and

most

common, triune divi-

"reptil-

basic

physiology. The

sion drawsa distinctionbetween the

ian brain," which is

responsible for

survival functions, such as

breathing,sleep-

ing, eating, the "mammalian brain," which

neural units associated with

encompasses

social

emotions, and the "hominid" brain,

which is

unique

to humans and includes

much of our oversizedcortex-the thin, fold-

ed,

layer covering the brain that is responsi-

ble for such "higher" functions as language,

consciousnessand

MacLean 1990).

measurementis

long-termplanning (Paul

Because

single

neuron

largely

restrictedto nonhu-

far shed far more

light

man animals, it has so

on the basic emotional and motivational

that humans share with other

higher-level processes

consciousness.

processes

mammals than on

such as language and

2.3

ElectricalBrain Stimulation (EBS)

Electrical brain stimulation is another

method that is largely restrictedto animals.

In 1954, two

Peter Milner

psychologists(James

1954)

Olds and

discovered that rats

would learn and execute novel behaviorsif

rewarded by brief pulses of electrical brain

stimulation (EBS)

to certain sites in the

brain. Rats (and many other vertebrates,

includinghumans) will work hard for EBS.

For a

big

series of EBS

pulses,

ratswill leap

grids,

and

over hurdles, cross electrified

forego their only daily opportunities to eat,

drink,

or mate. Animals also trade EBS off

against smaller rewards in a sensible fash-

ion-e.g.,

they demand more EBS to

they

are

hungry.

forego

food when

Unlike more

naturalistic rewards, EBS does not satiate.

And electrical brain stimulationat

sites often elicits behaviors such as

specific

eating,

drinking(Joseph Mendelson 1967), or copu-

lation

Hoebel

(Anthony Caggiula

1966). Many

and

abused drugs,

Bartley

such as

cocaine,

amphetamine,

heroin, cannabis,

and nicotine,

lower the threshold at which

animalswill lever-press for EBS (Roy

1996). Despite

its

obvious

economics,

only

Wise

applications to

explored

a few studies have

substitutability of EBS and other reinforcers

(Steven

Hursh

and B. H. Natelson; Leonard

Green and Howard Rachlin 1991; Peter

Shizgal 1999).

2.4

Psychopathology and Brain Damage in

Humans

Chronicmental illnesses

(e.g., schizophre-

autism),

nia),

developmental disorders (e.g.,

degenerative diseases of the nervous system,

and accidentsand strokesthat

damage local-

ized brain regions help us understandhow

the brain works

1994).

an areaX

(e.g.,

Antonio Damasio

When patients with known damage to

perform

a

special

task more

poor-

and do other tasks

ly than "normal" patients,

equally well, one can

to

do

the

special

infer that areaX is used

task. Patients who have

undergoneneurosurgical procedures such as

lobotomy

sion) or

(used

in the

past

to treat

depres-

(an

radical bisection of the brain

extreme

remedy for epilepsy,

provided

now

rarely

(see

used) have also

valuabledata

Walter Freeman and

James Watts 1942;

Gazzaniga and LeDoux 1978).

Finally,

a relatively new method called

transcranial magnetic stimulation (TMS)

uses pulsed magnetic fields to temporarily

disrupt brain function in specific regions.

The difference in cognitive and behavioral

  • 14 Journal of

Economic Literature, Vol. XLIII (March 2005)

functioning that results from such disrup-

tions

provides

clues about which

regions

controlwhich neuralfunctions. The theoret-

ical

advantage of TMS over brain imaging is

that TMS directly leads to causalinferences

about brain

functioning

rather than the

provided by

purely associationalevidence

imaging techniques. Unfortunately, the use

of TMS is

currently limited to

the cortex (it

visual

in the backof

is particularly useful for studying

in the

processes

the

occipital lobe,

head). It is also

controversialbecause it

can cause seizures and may have other bad

long-run effects.

  • 2.5 Psychophysical Measurement

An old and

ment of

heart rate,

response

(GSR,

pupil

simple technique

is measure-

psychophysiological indicatorslike

blood

pressure, galvanic

the

skin

sweating in

palms), and

response

to

dilation (pupils dilate in

arousal, including monetary reward). These

measurements

portable,

and

are easy,

very rapid

not too obtrusive,

in time. The draw-

back is that measurementscan fluctuatefor

many reasons

many

lead

just

ing

(e.g., body

movement) and

different combinations of emotions

to similar psychophysiologicaloutput,

as

being pulled over by a cop

and meet-

similar

a blind date

may produce very

anxiety responses.

Often these

combination

emotional

measurements are useful in

with

other techniques

likely

or in lesion

patients

who are

logical

to have very different physio-

(e.g.,

sociopaths

do

not

reactions

show normal GSR fear reactions before a

possible

monetary loss). Facial musculature

attaching

small

(on the

can also be measured by

electrodes to

smiling

cheekbones) and

the eyebrows).

muscles

frowning muscles (between

2.6

Difusion Tensor Imaging(DTI)

Diffusion tensor imaging (DTI) is a new

technique which exploits the fact that water

flows rapidlythough myleinated(sheathed)

neural axons (Dennis Bihan et al. 2001).

Imaging the water flow can therefore reveal

the trajectories which project from one neu-

ral

region

to others

(like watching patterns

helicopter

can tell

you

of economic and

region's

in

useful

circuitry and is an

of car traffic from a

roughly the ebb and flow

social

neurons project

understanding

neural

activity). Knowing where a

is

extremely

important complement to simply imaging

in

activity

with little

multiple regions (using fMRI)

ability to pin

down which

activity

Furthermore, the

technique

is

an

autopsies, which

occurs earliest.

can be used after

obvious advantage.

2.7

Is Neuroscience Just About Where

ThingsHappen in the Brain?

Neuroscience is sometimes criticized as

providing

little more than a

picture

of

"where things happen in the brain"or, more

cynically, as simplyshowing

caused

by

action in

that behavioris

system

the nervous

(which was never in doubt). Indeed, some

neuroscientistswho are

purely

interested in

regions

the functionality of different brain

would

endorse

such

a characterization

unapologetically.

However,

the

long-run goal

more than a

what

different

parts

for

of neuro-

map of the

of the brain

espe-

science is to provide

mind.

By tracking

by

are activated

cially

tasks, and

by looking

tasks,

overlap between

gaining

an

of the

diverse

neuroscientistsare

understanding of what different parts

brain

do,

how the

parts

ry," and, hence, how

ent

types

because

of

interact in "circuit-

of the brain are

the brain solves differ-

For

example,

problems.

parts

different

more or less associated with affective or

cognitive processing (a distinction we

define

ple

more precisely

they

below),

imagingpeo-

types

of

clues

while

are doing different

economic

tasks

provides important

affective and

about the mix of

cognitive

processes in those tasks. For example,

Sanfey et al. (2003) find that the insula cor-

tex-a

region in the temporal lobe that

encodes bodily sensations like pain and

odor disgust-is active when people receive

Camerer,Loewenstein, and Prelec: Neuroeconomics

15

low offers in an ultimatum

game.

low

This means that even if

bargaining

rejecting a

adapted

tough-

the future),

offer is done because of an

up a reputation

get

more in

for

instinct to build

ness (in order to

the circuitry that encodes this instinct clear-

ly has an

affective component, which

is not

purely cognitive.

For

about functional

regions

neuroeconomists, knowing

specialization,

more

and how

collaboratein different

tasks, could

substitutefamiliar distinctions between cat-

egories

of economic behavior (sometimes

arbitrarilyby suggestions

conventions)

which

with new

established

become modeling

ones

ple,

in

grounded

the insula

in neural detail. For exam-

activity noted by Sanfey et al.

when

subjects

guess

(sec-

Bhatt and

bargaining is also present

playing

matrix games are asked to

subjects

think they will do

Meghana

what other

ond-order beliefs) (see

Camerer, in press).

the

subjects

This

ultimatum study are not

simply

about what

suggests that maybe

who receive low offers in the

disgusted,they

are

evaluating a second-order belief

proposersexpect

them to do, as

an input into an emotionalevaluation. This is

just

a

speculation,

of course, but it shows

how direct

understanding of neural circuitry

and the search for

can inspire theorizing

new data.

In

light

of

the

long

reviewed earlier in this

list of methods

section, it is also

worth emphasizing that neuroscience isn't

only

ies

about brain

imaging.

Brainlesion stud-

(as well as TMS)

allow one to examine

the

impact

of

disablingspecific parts

of the

brain. They have often provided clearerevi-

dence

on functionality

with

of

specific

brain

regions-even

rare

types

ods have.

of

tiny sample

sizes from

meth-

damage-than imaging

Single

neuron measurement pro-

just

about what

parts

but about the specif-

vides informationnot

of the brain "lightup,"

ic conditions that cause specific neurons to fire at accelerated or decelerated rates. And

electrical brain stimulation, though it is largely off limits to studies involving

humans,

provides

a

level of

experimental

field

pudding."

many

control that

economists conducting

envy.

is in the

researchshould

As always, "the proof

Here, our

goal

ways

is not to review the

in which neuroscience will

rapidly

change economic theory, because that is not

where we are

some

key

yet.

Our

goal

is to showcase

findings in neuroscience, and stim-

ulate the reader's curiosity about what these

findingsmight mean for economics.

3.

Basic Lessons from Neuroscience

Because most of these

techniques

involve

localizationof brain

activity, this can easily

neuroscience is

"geography

of the

foster a misperception that

merely developing

brain," a

part

of the

map

job.

a

of which brain bits do what

If that were indeed so, then

there would be little reason for economists

to

attention. In reality, however, neuro-

beginning

to use

regional activity

pay

science is

differences and other clues to elucidate the

principles

tioning,

of brain

organization and func-

is

radicallychanging

which in turn,

our understanding of how the brain works.

Our

of the

goal

in this section is to

highlight

some

findings

from neuroscience that may

prove relevantto economics.

3.1 A Two-DimensionalTheoretical

Framework

Our organizing theme, depicted

in Table 1,

emphasizes

the two distinctionsmentioned

in the introduction, between controlledand

automatic processes, and between cognition

and affect.

3.1.1

Automaticand ControlledProcesses

The distinction between automatic and

controlled

was first

processes

proposed by

Many

others

Schneider and Shiffrin (1977).

have developed similar two-system models

since then, with different labels: rule-based and associative (Steven Sloman 1996);

rational and experiential systems (Lee Kirkpatrick and Seymour Epstein 1992);

  • 16 Journal of Economic Literature, Vol. XLIII (March 2005)

Two

TABLE1

DIMENSIONS OF NEURAL FUNCTIONING

Cognitive

Affective

Controlled Processes

"

serial

"

effortful

"

evoked deliberately

"

good introspective

access

Automatic Processes

"

"

parallel

effortless

"

reflexive

"

no introspective

access

reflective

and

reflexive

(Matthew

Automatic processes

are the

opposite

of

Lieberman et

al. 2002); deliberative and

implementive systems

Kentaro

Fujita,

and

(Peter Gollwitzer,

Gabriele

Oettingen

controlled

processes
dimensions-

not accessible to

on each

of

these

they operate in parallel, are

consciousness,

and

are

2004); assessment and locomotion

(Arie

Kruglanski et al. 2000), and type

  • II processes

I and

type

(Daniel Kahnemanand Shane

Frederick 2002).

relatively

rapid

tasking,

power

effortless. Parallelism facilitates

response, allows for massive multi-

and

gives

the brain remarkable

types

of

comes to certain

when it

Controlled processes,

as described by the

tasks,

such

as

visual

identification.

two rows of Table

1,

are serial

(they

use

Parallelism also

provides redundancy that

step-by-step logic or computations), tend to

be invoked

deliberatelyby

her or she encounters a

the

agent when

or sur-

decreases the

When neurons

in a region,

the

brain's vulnerability to injury.

are progressivelydestroyed

challenge

and

prise

(Reid Hastie 1984),

are often

feeling

of

associated with a

subjective

effort. People can typicallyprovide a good

introspectiveaccount

of controlled

process-

consequences are typically

than sudden

("graceful

gradual

rather

degradation").3 "Connectionist"neural net-

work models formulated

chologists

by cognitive psy-

James

(David Rumelhart and

es. Thus,

if asked how

they

solved a math

McClelland 1986)

capture these features

problem

or choose

a new

car,

they

can

and have been applied to

many domains,

often

recall the considerations and the

steps leading up

to the choice.2 Standard

tools of economics, such as decision trees

3The

damage

brain's

ability

to recover from environmental

property called plasticity. In

is also facilitated by a

and dynamic programming, can be viewed

as stylized representations

of controlled

one study that illustrates

the power

of

plasticity,

the

optic

nerves of ferrets (which are born when they are still at a

relatively immature state of development

when the brain is

processes.

2 Elaborate methods have been

developed that