A Theory of Fads, Fashion, Custom, and Cultural Change as Informational Cascades
Author(s): Sushil Bikhchandani, David Hirshleifer, Ivo Welch
Source:
The Journal of Political Economy,
Vol. 100, No. 5 (Oct., 1992), pp. 992-1026
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/2138632
Accessed: 10/12/2010 20:54
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=ucpress.
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.
The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to The
Journal of Political Economy.
http://www.jstor.org
A
Theory
of
Fads,
Fashion, Custom,
and
Cultural
Change
as Informational
Cascades
Sushil
Bikhchandani,
David
Hirshleifer,
and
Ivo
Welch
University of California,
Los
Angeles
An
informational cascade
occurs when
it
is
optimal
for
an
individual,
having
observed the
actions
of
those ahead
of
him,
to follow
the
behavior
of the
preceding
individual
without
regard
to
his
own in-
formation. We
argue
that localized
conformity
of
behavior and
the
fragility
of
mass
behaviors
can
be
explained by
informational cas-
cades.
Let
them
alone:
they
be
blind
leaders
of the
blind.
And
if
the
blind
lead
the
blind,
both
shall
fall into
the ditch.
[Matthew
15:14]
I.
Introduction
One
of
the most
striking regularities
of human
society
is localized
conformity.
Americans
act
American,
Germans
act
German,
and
Indi-
We thank Daniel
Asquith, Christopher
Barry, Gary
S.
Becker,
Michael J. Brennan,
Colin
Camerer,
Darrell
Duffie,
Robert
Ellickson,
Dan
Friedman, Stephen
J. Gould,
Jack Hirshleifer,
David
Hull,
Timur
Kuran,
Steve
Lippman, John
Mamer, Hubert
Markl,
Robert
May,
David
Nachman, Jonathan
Paul,
Tom
Philipson,
Ivan P. L. Png,
Thomas
Schelling,
Max
Steuer,
Sheridan
Titman,
Gordon
Tullock,
the editor, two
anonymous referees,
the members of the Rational
Choice seminar at
the University
of
Chicago,
and seminar
participants
at
University
of
California,
Berkeley, Hebrew
University, Jerusalem,
the
National Bureau of
Economic
Research,
National Taiwan
University,
Northwestern
University,
the
University
of
Pittsburgh,
and
the
University
of Utah for
helpful
comments
and
discussions. Parts of this
paper
subsume
parts of
an
earlier
paper,
"Sequential Sales,
Path
Dependence,
and Cascades"
(September
1989
version) by
Welch.
[Journal
of
Political
Economy,
1992,
vol.
100,
no.
5]
?
1992
by
The
University
of
Chicago.
All
rights
reserved.
0022-3808/92/0005-0001$01.50
992
INFORMATIONAL
CASCADES
993
ans
act Indian.
At
one school
teenagers
take
drugs,
but at
another
they "just say
no."
English
and
American
youths
enthusiastically en-
listed
to
fight
in World War
I,
but
pacifist
sentiments
prevailed prior
to
World War II
and in
the 1960s.
Four
primary
mechanisms have been
suggested
for
uniform
social
behavior:'
(1)
sanctions
on
deviants,2
(2) positive
payoff externalities,3
(3)
conformity
preference,4
and
(4) communication.5 The
first
three
theories
can
explain
why society
may
fix
on
undesirable
choices,
or
at
least
why
the
social
outcome
may
be
history-dependent. Sanctions
can
enforce a malevolent
dictatorship, payoff
externalities can
drive
a better
technology
to extinction
(e.g.,
the beta video
system),
and
people
with a direct
preference
for
conforming
may jump
on
the
bandwagon
for
fairly arbitrary
behavior
(e.g.,
bell-bottom
jeans).
These effects tend
to
bring
about a
rigid
conformity
that cannot
be
broken
by
small
shocks.
Indeed,
the
longer
the
bandwagon
continues,
the
more robust
it
becomes. The
fourth
theory
implies convergence
toward the correct outcome
if
communication is credible
and
costless.
It does
not
explain
why
mass behavior is
error-prone.
None
of these theories
explains
why
mass
behavior
is often
fragile
in the sense
that
small
shocks
can
frequently
lead
to
large
shifts
in
behavior.6
For
example,
cohabitation
of
unmarried
couples was
viewed
as scandalous
in the
1950s,
was flaunted in
the
1960s,
and
was
hardly
noticed
in
the 1980s.
Colleges
in which
students
demonstrated
and
protested
in
the 1960s became
quiet
in
the
1980s. The recent
rejection
of
communism
began
in
Poland
and later
spread
rapidly
among
other Eastern
European
countries.
Religious
movements,
re-
vivals, and
reformations,
started
by
a
few
zealots,
sometimes
sweep
across
populations
with
astonishing rapidity.
Addiction to and
social
attitudes
associated
with
alcohol,
cigarettes,
and
illegal drugs
have
fluctuated
widely.
This
paper
offers
an
explanation
not
only
of
why people
conform
but also
of
why
convergence
of
behavior
can be
idiosyncratic
and
l Boyd
and Richerson
(1985)
examine several
general
models
of cultural transmis-
sion that could
be consistent with these
mechanisms. Becker
(1991) analyzes conformity
in
product
demand
in
a model that
is
also consistent with
several of these mechanisms.
2
See, e.g.,
Akerlof
(1980),
Bendor and
Mookherjee (1987),
Coleman
(1987),
Hirshleifer
and
Rasmusen
(1989),
and Kuran
(1989).
3
See, e.g., Schelling (1960, 1978),
Dybvig
and
Spatt (1983),
Farrell and
Saloner
(1986),
Katz and
Shapiro (1986),
and Arthur
(1989).
For
example, conventions such
as
driving
on
the
right- (or left-)
hand
side
of
the road are
self-enforcing, once a few
individuals follow
the
convention.
4 InJones (1984),
individuals
inherently
wish to
conform
with
the behavior of others.
3 Conformity can
be achieved if
early
individuals explain the benefits of alternatives
to
later
ones
(see, e.g., Rogers 1983).
6
In
Kuran
(1989),
sanction-enforced behavior for
specific
sets of
exogenous parame-
ter values
can
be sensitive
to small shifts.
994
JOURNAL
OF POLITICAL
ECONOMY
fragile.
In
our
model,
individuals
rapidly
converge
on
one
action
on
the basis
of
some
but
very
little information.
If even
a little
new
information
arrives,
suggesting
that
a different
course
of
action
is
optimal,
or
if
people
even suspect
that
underlying
circumstances
have
changed
(whether
or not
they
really
have),
the
social equilibrium
may
radically
shift.
Our
model,
which is
based
on what
we
call
"informa-
tional
cascades,"
explains
not
only
conformity
but also
rapid
and
short-lived
fluctuations
such as
fads,
fashions,
booms,
and crashes.
In
the
theories
of
conformity
discussed earlier, small
shocks lead
to
big
shifts
in
mass behavior
only
if
people
happen
to
be very
close
to
the
borderline
between alternatives.
Informational
cascades
explain
why
society,
on the basis
of little
information,
will
systematically
tend
to
land
close
to the
borderline, causing
fragility.
An informational
cascade
occurs when
it
is
optimal
for an individ-
ual,
having
observed
the actions
of those
ahead
of
him,
to
follow
the
behavior
of the
preceding
individual
without
regard
to
his
own
information.
Consider
the
submission
of this
paper
to
a
journal.
The
referee
will read
the
paper,
assess
its
quality,
and
accept
or
reject
it.
Suppose
that
a referee
at
a
second
journal
learns
that
the
paper
was
previously
rejected.
Under
the
assumption
that
the referee cannot
assess
the
paper's
quality
perfectly,
knowledge
of the prior
rejection
should
tilt
him
toward
rejection.
Suppose
now
that the second
journal
also
rejects
and that
when
the
paper
is submitted
to a third
journal,
the
third
referee
learns
that
the
paper
was
rejected
at
two
previous
journals.
Clearly,
this
further
raises
the chance
of
rejection.
In a
fairly
general
setting
with
sequential
choices,
we show that at
some
stage
a
decision
maker
will
ignore
his
private
information
and
act
only
on the information
obtained
from
previous
decisions.
Once
this
stage
is
reached,
his decision
is
uninformative
to
others.
There-
fore,
the
next
individual
draws
the same inference
from
the
history
of
past
decisions;
thus
if
his
signal
is drawn
independently
from the
same
distribution
as
previous
individuals',
this individual
also
ignores
his
own
information
and
takes
the same action
as the
previous
indi-
vidual.
In the absence
of
external
disturbances,
so
do
all later
indi-
viduals.
The paper
submission example
is
special
in
that
only
one
journal
can
accept
the
paper,
ending
the submission
process.
Thus the
only
possible
cascade
that can
arise is
one
of
rejection.
In
many
situations
cascades
can
be
either
positive-wherein
all
individuals
adopt-or
negative-wherein
all
individuals
reject.
Consider a
teenager
decid-
ing
whether
or
not
to
experiment
with
drugs.
A
strong
motive for
experimenting
with
drugs
is the fact that
friends
are
doing
so.
Con-
versely,
seeing
friends
reject
drugs
could
help
persuade
a
youth
to
stay
clean.
INFORMATIONAL
CASCADES
995
Although
the outcome
may
or may not
be socially
desirable,
a rea-
soning process
that
takes into
account
the
decisions
of others is
en-
tirely
rational
even
if
individuals
place
no value
on conformity
for
its
own
sake. Imitation
is,
of
course,
an
important
social
phenomenon,
as has
been
documented
by
numerous
studies
in
zoology,
sociology,
and social
psychology.
Our contribution
is to model
the
dynamics
of
imitative
decision
processes
as informational cascades.
We examine (1)
how
likely
it
is
that a
cascade
occurs, (2)
how likely
it
is
that
the
wrong
cascade
occurs
(can
a
good paper
be
unpublish-
able?), (3)
how fashions
change
(why
were college
students of
the
1980s pre-business
"achievers,"
whereas those
in
the
1960s flirted
with
"alternative
cultures"?),
and
(4)
how
effective are
public
infor-
mation
releases
(e.g.,
a
campaign
to
publicize
the health
effects
of
smoking).
There
are several
related
papers
in
which
private
information
causes
individuals
to imitate the actions
of others.
In
Conlisk's
(1980)
evolutionary
model,
optimizers-who
incur
a decision
cost-coexist
with
imitators-who
avoid
this cost
but make inferior
decisions be-
cause
of
observational
lags.
Welch
(1992)
examines
the
likelihood of
cascades
and
optimal
pricing
in the market for initial
public
stock
offerings.7
Banerjee
(in press)
independently
models
"herd
behavior"
as
cascades.
Conceptually,
our
paper
differs from
Welch's
and
Baner-
jee's
in
emphasizing
the
fragility
of
cascades
with
respect
to
different
types
of
shocks;
cascades
can
explain
not
only
uniform
behavior but
also
drastic
change
such as
fads.8
The
remainder
of
the
paper
is
structured
as
follows. Section
II
presents
the basic model,
shows
that cascades
can
often be
mistaken,
and
provides
conditions
under which
a cascade will almost
surely
start.
It then
examines how
a
few
early
individuals can have a
dispro-
portionate
effect
and how
small
parameter
shifts can
transform an
imitator
into
a
fashion
dictator.
Section
III
examines the effect
of
prior
disclosure
of
public
information and shows
that cascades
are
fragile
when
new
public
information can
arrive.
Section IV
discusses
several
examples.
Section
V shows how
the
possibility
of
changes
in
the
underlying
value
of alternative decisions can lead to
"fads,"
that
is,
to drastic
and
seemingly
whimsical
swings
in
mass
behavior
without
obvious
external
stimulus.
Section
VI
concludes
the
paper.
7Becker (1991)
examines pricing
decisions under demand externalities that
might
arise from informational
sources.
8 In
Scharfstein
and
Stein
(1990)
and
Zwiebel (1990), conformity
is an agency
phe-
nomenon. Scharfstein
and Stein show
that a manager may
imitate the action
of a
preceding
manager
in order
to
improve
his
reputation
for
high
ability. Zwiebel
shows
that
relative
performance
evaluation
may
cause
managers
to adhere
to
inferior
indus-
try standards.
Bhattacharya,
Chatterjee,
and
Samuelson
(1986)
provide
an
interactive
learning
model
of
research and development.