Citizen Evacuation in Response to Nuclear and Nonnuclear Threats

Abstract

The study compares evacuations in nuclear and nonnuclear threats. Two issues in particular are examined: (1) citizen warning source and perceived credibility of warnings; and (2) citizen evacuation decision-making processes. We review citizens source of first warning, and perceived credibility of different warning sources. Cross-hazard comparisons are made among total evacuees, reasons given for evacuating and not evacuating, and citizen beliefs about the nature of the threat. Three types of hazard are chosen for comparisons; nuclear, volcano and riverine flood. The nuclear emergency used for analysis was the March 28, 1979 reactor accident at Three Mile Island (TMI), Pa. With regard to warning source, in the case of TMI most respondents first heard of the incident via mass media; virtually all others reported they first heard from a personal or nongovernmental source. Almost no respondents cited officials as a first source. The pattern of first information receipt in natural disasters was quite distinct. Most citizens heard first from emergency response authorities, and the next most frequently cited source was personal contacts. The mass media accounted for only a small proportion of first contacts. Citizen belief in real situational danger and advisories from officials were the most frequently cited reasons for leaving among evacuees in both nuclear and nonnuclear incidents. Also, for both TMI and the natural disasters, most of those who chose not to evacuate said that they believed they were in no real danger.

Full text

L L
BHAR-400/81/013
CITIZEN
EVACUATION
IN
RESPONSE
TO
NUCLEAR
AND
NONNUCLEAR
THREATS
FINAL
REPORT
Ronald
W.
Perry
SEPTEMBER
1981
Prepared
under
Contract
EMW-C-0296,
Work
Unit
Number
4821F,
Federal
Emergency
Management
Agency,
Washington,D.C.
20472
')Bduelile
DTIC
0
Human
Affairs
Research
Centers
E
E
CT
E
~OCT
19
1981
4000
N.E.
41st
Street
*
Seattle,
Washington
98105
O 118
Di
STROIfON
FTEMEii
A D
-Approved
for
public
zeleas*
Distribution
Unlimited

Legal
Notice
This
report
was
prepared
by
Battelle
as
an
account
of
sponsored
re
search
activities.
Neither
Sponsor
nor
Battelle
nor
any
person
act
Ing
on
behalf
of
either:
(a)
Paskes
any
warranty
or
representation,
axres
or
implied, with
respect
to
the
accuracy,
completeness
or
ufulness
of
the
information
contained
in
this
report
or
that
t6e
use
of
any
Information,
apparatus,
process,
or
composition
disclosed
In
this
report
may
not
infringe
privately
owned
right.l;
or
(b)
Assumes
any
liabilities
with
respect
to
the
use
of,
or
for
damages
resulting
fromn
the
use
of,
any
information,
apparatus,
process,
of
composition
disclosed
int
this
report.

Acces!:ion
for
NTIS
c.UBHARC/81/013
flhIC
TAB
UnPnnc2:,,e21
D
ivtritiution/'
CITIZEN
EVACUATION
IN
RESPONSE
TO
NUCLEAR
-~-A
n :id/t r
AND
NONNUCLEAR,
THRATS
Dist
speciai
FINAL
REPORT
Work
Unit
4821f
BY
RONALD
W.
PERRY
for
FEDERAL
EMERGENCY
MANAGEMENT
AGENCY
Washington,
D.C. 20472
Prepared
Under
Contract
Number
EMW-C-0296
FEMA
REVIEW
NOTICE
This
report
has
been reviewed
in the
Federal
Emergency
Management
Agency
and
approved
for
publication.
Approval
does
not
signify
that
the
contents
necessarily
reflect
the
views
and
policies
of
the
Federal
Emergency
Management
Agency.
Approved
for
Public
ReleaseOU
19
p
~September,
1981
v
D

REPORT
DOCUMENTATION
PAGE
IRLAD
1NTRtJCT1C'#
MJ~9OH1E
COMP'LETING~
FORM
3.
Govy
ACCESSION4
No,
i
NEC01,?"'CA
ALOSUMUEf
Ciien
Evacuation
in
Response
to
Final
epeo.'-
/
Nuclear
and
Nonnuclear
ThreatsAgyw9Qk~981
S.
O
ANAO~a
3*
VA-4NY"EmofP
i~onald
W./Perry
EMW-C-frF 7~ "
L
K#N~a.ORNO
IZATIC00
aSAAN
AMC
A004955
10
PROGRAM
99.IMaENT.
PPCStCT.
7^54
gU
4CLWo
dSSI
T
IED
ca
Bapproled
forPulic
Setlee
Dirbuio
9nlmi0d.
WahingtIson
e
.C. 707 on.~-/,~~'
ThisC
~tudCAGNY
repest&
DRsSo
an0em
initia
ose
in
Cl~pp&
of.
SRthe
CASochl
epofw
Anpod
ocr
thbicRease;
Dwoistrueuin
partiareexmied
1),
cTITIzN
wtarnTIng
shouc
ad
perstceivefdredibilit
of.
warnngs
an
(2)ucitizn
Evacuationdeionmknprcse.Wrvewctzs
sorcig
epoist
wann7adprevdceiiiyo
ifrn
ann
Threeoer
Mie sln
D0.
ABTRC
(Ca
147
Aorvo *
vt
O
p
ogvw
"d~
*535ol 0,SOLET
ThisSCIT stuyAeprsets
n nitalsteOi
sOpp
of
I PAhe
concet
ofee~
comprehensive
~ ~ ~ ~ ~
~~-
emrgnc maaeetb*oprn autosi ula

20.
Abstract
(continued)
sources.
Cross-hazard comparisons
are
made among
total
evacuees, reasons
given
for
evacuating
and
not
evacuating,
and
citizen
beliefs about
the
nature
of
the
threat.
Three
types
of
hazard
are
chosen
for
comparisons:
nuclear,
volcano
and
riverine
flood.
The
nuclear
emergency
used
for
analysis
was
phe
March
28,
1979
reactor
accident
at
Thrce
Mile
Island
(TMI),Pennsylvania.
With
regard
to
warning
source,
in
the
case
of
TMI
most
respondents
first
heard
of
the
incident
via
mass
media; virtually
all
others reported
they
first
heard
from
a personal
or
nongovernemental
source.
Almost
no
respondents
cited officials
as
a
first
source.
The
pattern
of
first
information
receipt
in
natural
disasters
was quite
distinct.
Most
citizens
heard
first
from
emergency
response authorities,
and
the
next
most frequently
cited
source
was
personal
contacts.
The mass
media
accounted
for
only
a small
proportion
of
first
contacts.
In
comparisons
of
public
confidence
in
information
sources,
the
important
finding
was
that
at
TMI
the
public perceived
the
mass
media
as
most
reliable,
while
in
the
nonnuclear
disasters public confidence
was
highest
in
local
emergency
response authorities.
Citizen
belief
in
real
situational danger
and
advisories
from
officials were
the
most
frequently
cited
reasons
for
leaving among evacuees
in
both
the
nuclear
and
nonnuclear incidents.
Also,
for
both
TMI
and
the
natural
disasters,
most
of
those
who
chose
not
to
evacuate
said
that
they
believed
they
were in
no real
danger.'.Among
nonevacuees
at
TMI,
conflicting
messages
and
the
absence
of
an
official evacuation
order
were
frequently
cited
as
reasons
for
staying.
In
the
natural
disasters, citizens
also
reported
that they
chose
to
not evacuate
in
order
to
protect
their
homes.
Finally,
the
implications
of
the
findings
for
evacuation
planning
and
operations
are
assessed.
t2

ACKNOWLEDGEMENT
The
data
on
natural
disasters
used
in
this
report
were
collected
and
initially
analyzed
in
connection
with
grants
from
the
Division
of
Civil
and
Environmental
Engineering
of
The
National
Science
Foundation.
Grant
number
PIR-77-23697
supported
the
studies
of
riverine
flood
evacuations,
and
a
suppleua-ent
to
that
grant
supported
the
subsequent study
of
the
evacuations
during
the
May
18,
1980
eruption
of
Mt.
St.
Helens
volcano
in
Washington
State. The opinions
and
conclusions
drawn
from
these
data
bases
are those
of
the
author
and
do
not
necessarily
reflect
official
National
Science
Foundation Policy.
i

EXECUTIVE
SUMMARY
The
study
represents
one initial
research
step
in
support
of
the
concept
of
comprehensive emergency
management
(CEM).
Comprehensive
emergency
management
refers
to
the
problem
of
developing
a
capability
for
handling
all
phases
of
activity--mitigation,
preparedness, response,
and
recovery--in
all
types of
disasters
by
coordinating
the
efforts
of
many
different agencies
(cf.
National
Governor's
Association,
1979:11).
Thus,
an
important
aspect
of
CEM
is
the
concept
of
managing
a
variety
of
types
of
disaster:
it
emphasizes
an "all
hazards"
approach
for
FEMA.
In
turn,
this
dictates
a
concern
on
the
part
of
managers with developing methods
and
concepts
which
are
applicable
across
numerous
disasters, both natural
and
man-made.
In
support
of
this
goal,
the
FEMA
sponsored
National
Academy
of
Sciences
Committee
on
United
States
Emergency
Preparedness
has
begun
to
lay the
theoretical
and
conceptual
foundations
for
making
such
cross-hazard
comparisons.
A
large
part
of
this
effort
has
involved
identifying common
or
generic
functions
which must
be
accomplished
in
managing
emergencies-i.e.,
evacuation,
search
and
rescue,
warning
dissemination, sheltering, rehabilitation,
public
information,
etc.--and
discussing, based upon historical
case
studies,
the
applicability
of
each
function
across
different
types
of
disasters.
In
developing
a
capability
for
CEM,
the
logical
extension
of
this
work
is
to
begin
examining
specific functions
and
making
systematic,
data-based comparisons
among
different
types
of
disasters.
In
this
way,
one
can
build
a
body
of
information which
documents
similarities
and
differences
in
human
performance
relative
to
specific
functions
for
numerous
disasters.
iii

The present report
focuses
upon
one
generic
function,
evacuation,
and
makes
comparisons among
two
natural
disasters
and one
nuclear
disaster.
An
important
goal
of
this
work
is
to
begin
laying
the
empirical
groundwork
for
making
further
comparisons
regarding evacuation
response
with
still
other
types
of
disasters. This
report
is
not
directly
aimed
at
providing
incident-specific guidance
for
emergency managers. Instead,
it
is
meant
to
serve
as
an
initial
step toward
building
a
body
of
data-based
comparisons which
can
subsequently
be
examined
and
integrated
into
a
data-bank
for
use
by emergency planners.
At
present
the
scientific
and
technical
literature
contains
no
systematic empirical
comparative
studies
of
human
response
to
different disaster
agents.
In
this
report
a
very
simple
model
for
making
such
comparisons
about
evacuation
behavior
is
used
as
a
starting
point
for
developing more
complex
models
in
subsequent analyses.
It
is
to
be
emphasized
that
this
work
represents
a
beginning
point
in
constructing
what
can
eventually
be
a
larger
body
of
generalizations
about
human
performance
under different
disaster conditions which
can
be used
in
developing comprehensive
emergency
management
strategies.
Its
optimal
value,
therefore,
rests
not
so
much
in
the
results
of
these
comparisons
alone,
but
in
its
integration
with
future
similarly
comparative
studies.
The
purpose
of
this
report
is
to
describe
the
results
of
a
comparative
analysis
of
data
on citizen
evacuation
behavior
in
response
to
nuclear
and
nonnuclear
threats.
Two
issues
in
particular
are
examined:
(1)
citizen
warning
source
and
perceived
credibility
of
warnings;
and
(2)
citizen evacuation decision-making.
We
review
citizens'
source
of
first
warning,
the
relative
utility
of
warning
I
f
Lvm

information,
and
the
perceived reliability
of
different warning
sources.
Then, cross-hazard comparisons
are
made
among
the
total
numbers
of
citizens
who
evacuate,
reasons
given
for
evacuating
or
not
evacuating
and
citizen beliefs
about
the
nature
of
the
threat
to
their
families.
Comparisons
are
made
among
three
types
of
hazard:
nuclear,
volcano,
and
riverine
flood.
The
nuclear
hazard
involves
evacuation behavior
after
the
accident which occurred
at the
Three
Mile
Island
nuclear
plant
on
March
28,
1979.
Three
studies serve
as
our
primary
sources
of
data
on
this
incident:
(1)
a
comprehensive
telephone
survey
of
area residents
sponsored
by the
Nuclear
Regulatory Commission
(Flynn,
1979);
(2)
a
mailed survey sponsored
by
Rutgers University
and
conducted
by
that
institution's
geography
department
(Barnes
et
al.,
1979);
and
(3)
a
mailed
survey sponsored
and
conducted
by
the
geography department
at
Michigan
State
University
(Zeigler
et
al.,
1981).
While
a
number
of
studies
of
Three
Mile Island
were reviewed, data were
selected
from
the
above
three
to
use
in
secondary
analysis.
Each
of
these
studies
is
based
upon
probability
samples
of
citizens
living
within
a
specified
radius
(distance)
of
the
Three Mile
Island
reactor.
The
data
on
volcano hazard
are
drawn from
a
study
of
evacuation
response
in
a
comunity
threatened
by
the
May
18,
1980
eruption
of
Mt.
St.
Helena
in
Washington
State
(Greene
et
al., 1980;
Perry
et al.,
1980a).
Data
on
three
comunity
evacuations conducted
in
response
to
riverine
floods
are
drawn
from
Battelle Institute archives (Perry
et al.,
1980b),
based
upon
research sponsored by'the National
Science
Foundation.
All
of the
natural
hazard
response
studies
are
also
based
upon
probability
samples
of
comunity
residents.
V
.....
._ _ _ _ _
__
_ _ _ _
._ _ _ _ _ _ _ _
__
_ _ _ _ _
_

The
report
is
structured around
five
chapters.
The
first
presents
a
review
of
past
attempts
to
compare
nuclear
and
nonnuclear
threats,
and
develops
an
explicit
logic
for
r
king
such
comparisons.
This
is
followed
by
a
section which
gives
a
brief
overview
of
each
of the
disaster
events
to
be
compared.
The
next
two
chapters
address,
in
turn,
comparisons
of
warning
source
credibility
and
evacuation decision-making
among
the
nuclear
and
nonnuclear
threats.
Finally,
the
last
chapter examines
the
implications
of
the
findings
for
the
problem
of
evacuation
planning.
Comparing Nuclear
and
Nonnuclear
Threats
This
report argues
that
emergency
management
and
citizen
response
to
nuclear
threats
can
appropriately
be
examined
within
the
same
conceptual
and
analytic
framework
as
any
other
disaster
agent,
whether
natural
or
man-made. The
same
basic
definition
of
disaster encompasses
a
broad
spectrum
of
disaster
events
and
when
definite characteristics
of
disasters
are
examined,
it
was
found
that
natural
hazards
differ
as
much
among themselves
as
from
nuclear hazards.
Thus,
there
appears
to
be
no
substantial theoretical reason
for
treating
nuclear
disasters
as
a
phenomenon which
is
incomparable
to
other
events
characterized
as
disasters
in
the
research literature.
It
is
not
argued
that
all
disasters
are
basically
the
same
or that
they
have
similar
consequences.
It
is
acknowledged
that
nuclear
disasters,
like
all
disaster
agents,
possess
some
unique
characteristics.
The most
unique aspect.of
a
nuclear power plant
accident
is
that
a
threat
which cannot
imsediately
be
seen,
heard,
or
felt-radiation--is
involved. Thus,
some
attention
is
necessary because,
in
terms
of the
way people
perceive
the
situation,
such
circumstances
are
different
from
those
which
accompany
other disaster
agents.
Research
vi
-
_
- -
_ _ _ _
_

indicates
that some
of
the
public
view
nuclear
energy,
and most
applications
of
it,
as
a
particularly threatening
hazard
with
the
potential
for
extraordinarily long-term
negative
effects. The
idea
that
people
have
a
different
"mind
set"
for
nuclear
disasters
certainly
does
not preclude
comparisons
with
nonnuclear disasters. Instead,
it
simply
requires
that
this
perceptual
dimension
and
the
emotional response
to
it
be
acknowledged
and
that
necessary
qualifications
be made
when
such
differences may
have
some
bearing
upon
human performance.
The
Disaster Events
This
chapter
gives
a
review
of
five threats:
the
May
18,
1980
eruption
of
the Mt.
St.
Helens
volcano
in
Washington
State; three
riverine
floods
which
occurred
in
the
Western
United
States
between
December,
1977, and
March,
1978;
and the
nuclear
reactor accident
at
Three Mile Island,
Pennsylvania,
which
began
March
28,
1979.
Descriptive
data
are
presented
on
each
of
the
disaster
events
and
a
largely
nontechnical
account
of
the
circumstances surrounding disaster
impact
is
related. The purpose
of these
descriptions
is
twofoid:
to
provide
an
overview
of
the
disaster
incident
and
to
convey general information which
will
make
interpretations
and
comparisons
between
events
more
meaningful.
Comparisons
of
Warning
Source
and
Credibility
This
chapter
presents
an
analysis
of
the
sources
from
which
citizens
first
received disaster
information
and
their
evaluation
of
the
credibility
of
different
sources
of
information about
the
threat.
The
majority
of
respondents--69%--reported
that they first
heard
of the
TMI
accident
from
the
mass
media;
virtually
all
others
first
heard
from
a
vii

personal,
non-governmental
contact--primarily
friends,
neighbors,
relatives
or
job
colleagues. Almost
no
one
cited
an
official--emergency
management
authority
or
general
government--source
as
the
place
from
which
they
initially heard
about
the
accident.
The
pattern
of
first
information
receipt
in
natural
disasters
was
quite distinct: most
people--about 50Z--heard
first
from
emergency
response
authorities,
and
the
next
most frequently
cited
first
source was
personal contacts.
The
mass
media accounted
for
only
a
small
proportion
of
the
first
contacts.
To
a
certain
extent,
the
differences
in
patterns
of
first
warning
source
between
THI and
the
natural disasters
may
be
understood
in
terms
of
the
low
forewarning
at
TI.
However,
the
differences
point
to
an
important
distinction
in
the
pattern
of
who
controls
the
emergency
response
to
the
natural
disasters
versus
the
TMI case. In
the
natural
disasters control
and
communication
tend
to
remain
with
local
authorities
and
the
mass
media
play
a
less
distinct
role
during
the
emergency
period. Two
important
factors
in
this
control
are
that
in
natural
disasters:
(I)
technical
status
reports
on
the
hazard
go
from
experts
to
emergency
response authorities
who
incorporate
the
information
into
their
planning
and
interpret
the
data
for
the
public,
and
(2)
those emergency
response
authorities
are
traditionally visible
to
and
recognized
by
the
public
as
responsible
for
protecting
the
citizenry.
The
consequence
of
having
visible emergency
response
authorities
in
control
is
that
it
enables
the
public
to
define more
easily
the
disaster
as
an
event
which
can
be
managed
to
an
acceptable
outcome.
viii
-'4

The finding
that
most respondents
first
heard
about
TMI
via
the
mass
media
foreshadowed
the
subsequent reliance
on
the
mass
media
as
a
primary
comnunication
channel
to
the
public.
Many
factors
influenced
public
perceptions
of
the
emergency
response
efforts
at
TMI,
including
high
visibility
of
political
figures
coupled with
lower
visibility
of
traditional
emergency
response personnel,
and
real
conflict among
responder
agencies.
The use
of
the
media,
however,
as
a
main
channel
of
comunication
to
the
public
probably exacerbated
(and
no
doubt
sometimes
exaggerated)
problems
of
control.
In
the
comparisons
of
public
confidence
in
information
sources,
the
important finding
was
that
at
ThI
the
public
perceived
the
mass
media
as
the
most
reliable
source,
while
in
the
nonnuclear disasters
the
public
placed
highest confidence
in
local
emergency
response authorities.
Evacuation Decision-Making
Citizen
belief
in
real
situational danger
and
advisories
from
officials
were cited most
frequently
as
the
critical
reasons
for
evacuating
in
both the
nuclear
and
nonnuclear
incidents.
Indeed,
these
two
reasons
alone
account
for
more
than
55%
of
the
volcano
evacuees,
69%
of the
flood
evacuees,
and
nearly
45% of
the
evacuees
at
TMI.
I
Interestingly,
mass
media
warnings
were
infrequently
chosen
as
the
most
important
reason
for
evacuating
in
all
three
types
of
hazard.
It
was
found,
however,
that
social
network
contacts
were relatively
more
important
to
evacuation
decision-making
in
the
natural
disasters
than
at
Three
Mile Island.
ix
-------------------------------..
.,.*.
.~

For both
TMI
and
the
natural disasters,
most
of
those
citizens
who
did
not
evacuate
chose not
to
because
they
did
not
believe
that
a
real
danger existed.
Among non-evacuees
at
TMI,
the
presence
of
conflicting
messages
and
the
absence
of
an
official evacuation
order
were
frequently
cited
reasons
for
staying.
In
the
natural
disasters,
people
also
reported
that they
chose
to
stay
so
that
they
could
protect
their
homes
from
the
environmental
threat.
Unlike
the
natural
disasters,
fear
of
looting
was
given
as
a
reason
for
not
evacuating
at
TMI.
Finally,
this
chapter concludes
with
a
detailed
examination
of
why
so
many
people
spontaneously
evacuated
at
TMI.
It
is
argued
that
the
evacuations
can
be
explained
in
terms of
two
general
categories
of
reasons:
(1)
largely
circumstantial
factors
related
to
the
way
the
emergency
was managed;
and
(2)
factors
related
to
the
public's
perception
of
the
risks involved
in
nuclear accidents.
The
evidence
marshalled
in
this
report suggests
that
once
we allow
for
the
fear
or
dread
characteristics associated
with
nuclear
disasters,
the
evacuation
response
at
TMI
can be
explained
using
the
same
model
developed
to
understand evacuation
behavior
in
other natural
and
man-made
hazards.
The
rudimentary
model
suggests
that
citizens
evacuate
when
four
conditions
are met:
(1)
they
have
accounted
for
the
safety
of
their
immediate
household;
(2)
they
have
been given--by
authorities-or
have
personally developed
a
plan
for
protective
action;
(3)
they
believe
that
a
threat does exist
in
the
environment;
and
(4)
they
perceive
that
upon
impact
this
threat
could result
in
some
level
of
damage
to
their
person,
family
and
property.
At
Three
Mile Island
the
nuclear
nature
of
the
x
I-
~___
.
_______77

threat
meant
that
people
perceived
personal risk
to
be
very high
(condition
four),
but
in
general
those
who
evacuated
were
people
for
whom
all four
conditions were
met.
Implications
for
Evacuation
Planning
Nine
general
conclusions
are
elaborated
in
this
chapter which
have
implications
for
the
conduct
of
evacuation planning:
"
During
the
course
of
a
nuclear
reactor
emergency,
local
emergency
response authorities
should be
integrated
into
the
public
information
system
and
should
constitute
the
public's
primary
source
of
accident-relevant
information.
"
When
an
emergency--either
nuclear
or
nonnuclear--is
in
progress,
the
mass
media
should
not
be
relied upon
as
a
primary
commnication
channel
to
the
public.
"
When
an
emergency
is
in
progress,
officials
should
distinguish
the
function
of
providing
public information
about
the
emergency
from
the
function of
sending
messages
which
direct
some
emergency
response.
*
In
all
disasters,
particularly
nuclear disasters,
rumor control
is
a
critically
important function.
"
The public information
function
is a
particularly
important
component
of
emergency
response
plans
for
dealing
with
nuclear
power plant
accidents.
"
The
"dual
use"
philosophy
appears
-to
be
founded
upon
reasonable
assumptions
in
that
the
basic
principles
of
human
response
to
natural hazards
also
describe
human
response
to
nuclear
threats.
xi
I

9
Inter-organizational
and
inter-agency
coordination
and
preparedness
for
ordering
and
over-seeing
a
mass
evacuation
are
crucial problems
in
both
nuclear
and
nonnuclear disasters.
e
The
high
level
of
spontaneous
evacuations
around
ThI
appears
to
be
related
to
the
public's
elevated
perceptions
of
levels
of
personal
risk
associated
with
radiation
threats.
9
Citizen
evacuation response
during
nuclear disasters
may
be
understood
in
terms
of
the
same
variables which
explain
evacuation
decision-making
in
nonnuclear disasters.
This chapter elaborates
the
reasoning behind
each
of
the
general
conclusions
and
derives
specific corollaries applicable
to
the
conduct
of
emergency
planning
and
operations.
Finally,
a
number
of
implications
of
the
present
study
for
further
research
are
discussed.
Three
general,
and
several
specific
studies
are
suggested:
(1)
a
study
of
how
resesrch
results
are
disseminated
from
researchers
to
planners
and
policy-makers,
as
well
as
how
these
latter
actors evaluate
and
incorporate
research
information
into
the
emergency
management
process;
(2) a
study
of
the
calculus
used
by
citizens
in
assessing
risks
associated with
radiation
relative
to
other
hazards;
and
(3)
studies
of
the
design
and
implementation
of
both
public
information
programs
regarding nuclear disasters
and
dissemination
programs
for
specific
emergency
response
plans.
xii
-----. =
-

TABLE OF
CONTENTS
Page
ABSTRACT
............... ............................ ii
EXECUTIVE
SUMMARY
............ ....................... iii
LIST
OF
TABLES
......... ......................... ...
xv
CHAPTER
ONE:
Introduction
......
...................
.1
CHAPTER
TWO:
Comparing Nuclear
and
Nonnuclear
Threats
.....
6
Classifying
Disaster Events
.......
..
................
8
Unique
Aspects
of
Disaster Events
...
.............
.15
CHAPTER
THREE:
The
Disaster Events
....
..............
.19
Volcanic
Eruption
.......
.....................
.19
Flood
Events
........
........................
.23
Valley
........ ........................
.23
Fillmore
........ .......................
.25
Snoqualmie
....... ...................... .
27
Nuclear Power
Plant
Accident
..... ............... ...
28
CHAPTER
FOUR:
Comparisons
of
Warning
Source
and
Credibility
.
33
Source
of
First
Information
.....
................
.34
Source
Credibility
.......
.....................
... 40
Summary
.........
..........................
.48
CHAPTER
FIVE:
Evacuation Decision-Making
.. ...........
.52
Reasons
for
Evacuating
...... ................... ...
52
Reasons
for
Not
Evacuating
.....
.................
. 59
The
Overall
Evacuation
Response
.................
.. 63
xiii

CHAPTER
SIX:
Sumary-
Implications
for
Evacuation
Planning
72
Implications
Arising
from
Source
Credibility.
.. ......... 73
Evacuation
Decision-Making
.. .................... 76
'Implications
for
Further
Study
.. .................. 81
BIBLIOGRAPHY
AND
REFERENCES. ...................... 85
xiv

LIST
OF TABLES
Table
Number Page
I
Disaster Agents Classified
by
Selected
Defining
Characteristics
......
...................
.12
2
First Source
of Information:
Three
Mile
Island
.. . .35
3
First Source
of
Information:
Nonnuclear
Disasters
. . .37
4
Utility
of
Information
from
Sources:
Three
Mile
Island
.........
........................ .
41
5
Most
Reliable
Source:
Three
Mile Island
.........
.44
6
Most
Reliable
Source:
Nonnuclear
Disasters
......
.45
7
Reasons
for
Evacuating:
Three
Mile Island
.......
.53
8
Critical Information
in
Decision
to
Evacuate:
Three
Mile Island
.......
.....................
.55
9
Most
Important
Reason
for
Evacuating:
Nonnuclear
Disaster
........
.......................
.57
10
Reasons
for
Not
Evacuating: Three
Mile
Island
.....
. 60
11
Most Important
Reason
for
Not
Evacuating:
Nonnuclear
Disasters
........
......................
.62
1.2
Perceived Threat
to
Family During
TMI
Accident
.....
.67
13
Perceived
Threat
from
Nonnuclear
Disasters
.......
.68
xv
, -- =- __

CHAPTER
ONE
INTRODUCTION
The
study
represents
one initial
research
step
in
support
of
the
concept
of
comprehensive emergency
management
(CEM).
Comprehensive
emergency management
refers
to
the
problem
of
developing
a
capability
for
handling
all
phases
of
activity--mitigation,
preparedness,
response,
and
recovery--in
all
types
of
disasters
by
coordinating
the
efforts
of
many
different
agencies
(cf.
National
Governor's
Association,
1979:11).
Thus,
an
important aspect
of
CEM
is
the
concept
of
managing
a
variety
of types
of
disaster:
it
emphasizes
an
"all
hazards" approach
for
FEMA.
In turn,
this
dictates
a
concern
on
the
part
of
managers
with
developing methods
and
concepts which
are
applicable
across
numerous disasters,
both
natural
and
man-made.
In
support
of
this
goal,
the
FEMA
sponsored
National
Academy
of
Sciences
Committee
on
United
States
Emergency
Preparedness
has
begun
to
lay the
theoretical
and
conceptual foundations
for
making
such
cross-hazard comparisons.
A
large
part
of
this
effort
has
involved
identifying
coamon
or
generic
functions
which
must
be
accomplished
in
managing
emergencies--i.e., evacuation,
search
and
rescue,
warning
dissemination,
sheltering, rehabilitation,
public
information, etc.--and
discussing,
based upon
historical
case
studies,
the
applicability
of
each
function
across
different
types
of
disasters.
In
developing
a
capability
for
CEM,
the
logical
extension
of
this
work
is
to
begin examining
specific functions
and
making
systematic,
data-based
comparisons
among
.'~ .~ *

2
different
types of
disasters.
In this
way,
one
can
build
a
body
of
information which
documents similarities
and
differences
in
human
performance
relative
to
specific
functions
for
numerous disasters.
The present report
focuses
upon
one
generic function,
evacuation,
and
makes comparisons among
two
natural
disasters
and
one
nuclear
emergency.
An
important
goal
of
this
work
is
to
begin
laying
the
empirical
groundwork
for
making
further
comparisons
regarding
evacuation
response
with
still
other
types
of
disasters.
This report
is
not
directly
aimed
at
providing
incident-specific guidance
for
emergency managers.
Instead,
it
is
meant
to
serve
as
an initial
step
toward
building
a
larger
body
of
data-based
comparisons
which
can
subsequently
be
examined
and
integrated
into
a
data-bank
for
use by
emergency planners. At
present
the
scientific
and
technical
literature
contains
no
systematic
empirical
comparative
studies
of
human response
to
different disaster
agents.
In
this
report
a
very
simple
model
for
making
such
comparisons
about
evacuation
behavior
is
used
as
a
starting
point
for
developing more
complex
models
in
subsequent analyses.
It
is
to
be
emphasized
that
this
work represents
a
beginning
point
in
constructing
what
can
eventually
be
a
larger
body of
generalizations
about
human
performance under
different
disaster conditions which
can
be
used
in
developing comprehensive
emergency
management
strategies.
Its
optimal
value,
therefore,
rests
not
so
mach
in
the
results
of
these
comparisons
alone, but
in
its
integration
with
future
similarly
comparative
studies.
To
date,
social
scientific examinations
of
the
question
of
the
comparability
of
human
response
to
nuclear
and
nonnuclear
hazards
have
been
largely
at
a
theoretical
level.
Those who
believe
that
response
to

3
nuclear
disasters
will
be
different have based
their
arguments upon
the
presumed
unique
nature
of
the
disaster
agent:
the
radiological
component,
the
potentially
huge
magnitude
of
negative consequences,
extended secondary
effects,
and
lack of
public
experience
with
such
disasters.
Thus,
the
emphasis
is
upon
the
disaster
agent
itself.
On
the
other
hand,
it
has
been
demonstrated
that
one can
focus
upon
generic
functions
in
disasters
and
develop
a
strategy
for
making
comparisons
among
different disaster
agents
based upon
the
ways
in
which
these
functions
are
conducted.
In
adopting
this
approach,
research
emphasis
shifts
away
from
concern with
cataloguing
ways
in
which disaster
agents
are alike
or
different,
and
focuses
upon
assessing
ways
in
which
function-related
human behaviors
compare
between
nuclear
and
nonnuclear
disasters. This emphasis
upon
analyzing
"functions" allows
the
investigator
to
focus
upon
the
relevant
issue
of
commonalities
or
differences
in
human
response.
In
this
report,
we have
adopted
the
latter strategy and
focus
upon
the
function
of
population evacuation. This
report examines
two
issues
in
particular:
(1)
citizen
warning
source and
perceived
credibility
of
warnings;
and
(2)
citizen evacuation
decision-making.
We
review
citizens'
source
of first
warning,
the
relative utility
of
warning
information,
and
the
perceived
reliability
of
different
warning
sources.
In
looking
at
evacuation decision-making, cross-disaster
comparisons
are
made
among
the
total
numbers
of
citizens
who
evacuate,
reasons
given
for
evacuating
or
not
evacuating
and
citizen beliefs
about
the
nature
of
the
threat
to
their
families.

4
Comparisons
are
made among
three types of
disaster,
one
nuclear
and
two
nonnuclear.
The
nuclear hazard
studied
involves
evacuation
behavior
after
the
accident which occurred
at
the
Three Mile
Island
nuclear
power
plant
on
March
28,
1979.
Three
studies
serve
as
our
primary
sources
of
data on
this
incident:
(1) a
comprehensive
telephone survey
of
area
residents
sponsored
by
the
Nuclear
Regulatory
Commission
(Flynn,
1979);
(2) a
mailed
survey
sponsored
by
Rutgers University
and
conducted
by
that
institution's
geography department
(Barnes
et al.,
1979);
and
(3) a
mailed
survey
sponsored
and
conducted
by
the
geography
department
at
Michigan
State
University
(Zeigler
et
al.,
1981).
While
a
number
of
studies
of
Three
Mile Island were reviewed,
data
were selected
from
the
above
three
to
use
in
secondary
analysis.
Each of these
studies
is
based
upon
probability
samples
of
citizens
living
within
a
specified radius
(distance)
of
the
Three
Mile
Island reactor.
The
two
types
of
nonnuclear
disasters
examined
in
this
report
are
volcanoes
and
riverine
floods.
The
data
on
volcano hazard
are
drawn
from
a
study
of
evacuation
response
in
a
community
threatened
by
the
May
18,
1980
eruption
of
Mt.
St.
Helens
in
Washington
State
(Greene
et
al.,
1980;
Perry
et
al.,
1980a).
Data
on
three
community
evacuations
conducted
in
response
to
riverine
floods
is
drawn
from
Battelle
Institute
archives
(Perry
et
al.,
1980b)
to be
used
in
the
comparative analyses.
All
of
the
natural
hazard
response
studies
are
also
based
upon
probability
samples
of
community
residents.
The
remainder
of
this
report
is
structured
around
five
chapters.
The
first
presents
a
review
of
past
attempts
to
compare
nuclear
and
nonnuclear
threats,
and
explicitly develops
a
logic
for
making
such

comparisons.
This
is
followed by
a
section which
gives
a
brief
overview
of
each
of
the
disaster
events
to
be
compared.
The next
two
chapters
address,
in
turn,
comparisons
of
warning
source
credibility
and
evacuation decision-making
among
the
nuclear
and
nonnuclear
threats.
Finally,
the
last
chapter
examines
the
implications
of
the
findings
for
the
problem
of
evacuation
planning.

6
CHAPTER
TWO
COMPARING
NUCLEAR
AND
NONNUCLEAR
THREATS
To
date,
there
have been
very
few
attempts
to
make systematic
comparisons
of
human
response
to
different
types of
disaster
agents.
Indeed,
there
has
been
a
general
reluctance
to
apply
findings
about
human
behavior
from one
type
of
natural
disaster
to
another;
the
matter
of
comparing nuclear
with
nonnuclear
threats
did not
begin
to
appear
in
the
professional literature
at
all
until
the
late
1970's.
In
part,
this
condition
reflects
the fact
that
historically
a
large
component
of
disaster
studies has
been
journalistic
and
descriptive
in
nature
(Gillespie
and
Perry,
1976:303).
Hence, attention
has
often
been
focused
upon
the
disaster
event
itself--the
hurricane
or
the
earthquake--and descriptions
of
specific
consequences
of
the
disaster
for
victims.
The
literature reported,
then,
on
earthquake victims
crushed
under
rubble
or
burned
by fires
and
hurricane
victims
drowned
in
the
storm
surge.
In
this
context,
many
disaster researchers
have argued
that
different disaster
agents
have
different
characteristics
and
impose
different
demands upon
a
coimunity
social
system;
thus,
human
reaction
to
different disasters
is
likely
to
be
different.
Such
reasoning
concentrates
upon
the
disaster
event
itself and
specifically
focuses
on
the
uniqueness
of
different
events.
It
is
of
course
correct
that
disaster
events
at
this
level
are
all
different;
particularly
in
terms
of
the
precise
agent
which
imposes
physical damage.
However,
this
approach
involves
essentially
a
phenotypic
classification
system
for
disaster
events,
focusing
upon
the
_ __ _ _
*1
.. .
-----
-

7
surface
or
visible properties
of
each
event.
Carried
to
its
logical
extreme,
such
an
approach would
conclude
that
even
all
riverine
floods
possess
certain
unique
characteristics,
which
technically
implies
that
they
are not
fully
comparable with
one
another.
In
the
past
decade,
there
has
been
a
transition
in
disaster
studies
toward
an
increased
concern
with
the
development
of
conceptual
schemes
for
understanding
and
explaining human
response
to
disaster.
In
so
doing,
research
attention
has
turned from
describing
disaster
events
to
understanding
the
demands
and
stresses
resulting
from
disaster
impact
and
cataloguing
various strategies
for
coping with
such
demands
and
stresses. To effect
this
shift from
examining
the
event
to
focusing upon
human
response
requires
that
(1) a
more systematic
means
of
classifying
disaster
events
be
devised
to
promote
(2)
the
delineation
of
comnon
functions
or
demands
imposed
upon
individuals
and
social
systems
as
a
consequence
of
disaster
impact.
The
purpose
of the
classification system
is
to
characterize
disasters,
not
in
phenotypic
terms,
but
in
terms
of
features
which
have
an
impact
on
the
kinds
of
protective
or
ameliorative measures
that
might
be
used
in
a
mitigation
program.
In
this
way,
one
may
choose
a
given
function--for
example,
population
warning--and
examine
the
ways
in
which
the
task
varies
across
different
disaster
events
because
of
different
disaster
characteristics--such
as
the
presence
of
a
technology
to
detect
the
pending
threat
in
advance
or the
speed
of
disaster
onset
once
detected.
S__ I

The
following paragraphs
develop
a
logic
for
classifying
disasters
in
terms
that
facilitate
effective comparisons
of
human
response
across
different
disaster
agents. This
review
draws
upon
the
classification
schemes
devised
by
Kreps
(1979)
and
Perry
et al.
(198
0c)
for
comparing
natural
disasters
with nuclear
attack. The
scheme
presented here
is
devised
by
examining
the
definition
of.disaster
and
isolating
crucial
dimensions
for
comparison.
Finally,
each
of
the
three
hazards
of
interest
here--volcanoes,
floods,
and
nuclear
power
plant
accidents--are
classified
using
the
selected
dimensions
as
the
basis
for
comparison.
Classifying
Disaster
Events
Disasters
are
usually
thought
of
as
catastrophic
events, frequently
associated
with
the
forces of nature:
earthquakes,
tornadoes,
hurricanes,
etc.
Yet other
events,
such
as
explosions, chemical
spills
or
industrial
accidents,
are also
described
as
disasters.
In
establishing parameters
for
the
social
scientific
study
of
disaster, Charles Fritz
(1961:655)
has
advanced
a
definition which concentrates
on
important
distinguishing
features
of
disaster
events. He suggests
that
a
disaster
is
any event:
. . .
concentrated
in
time and
space,
in
which
a
society
or
a
relatively
self-sufficient subdivision
of
society, undergoes
severe
danger
and
incurs
such losses
to
its
members
and
physical
appurtenances
that
the
social
structure
is
disrupted
and
the
fulfillment
of all
or
some
of the
essential functions
of
the
society
is
prevented.
This classic
definition
stresses
that
disesters
occur
at
a
definite
time
and
place
and
that
they
disrupt
social
intercourse
for
some
period
of
time.
Allen Barton
(1970:38) proposes
a
similar
definition,
but
chooses
to
focus
upon
social
systems,
arguing
that
disasters
exist
"when
many
_ __ _ _ _ _ _ _ _
_

9
members
of
a
social
system
fail
to
receive
expected
conditions
of
life
from
the
system". Both
Fritz
and
Barton
agree
that
any event
which
results
in
a
significant
change
in
inputs
or
outputs
for
a
given
social
system
is
accurately characterized
as
a
disaster.
The
important
point
to
be
derived
from
inspecting
these
definitions
is
that
volcanoes,
hurricanes,
floods,
chemical
spills,
explosions,
or
nuclear
power plant
accidents
all
fit
equally
well
into
either
definition. Hence,
at
this
level
of
abstraction,
both nuclear
and
nonnuclear disasters
may
be
treated
under
the
same
conceptual
rubric.
Given
that
nuclear
and
nonnuclear disasters
may
be
subsumed
under
the
same
definitional umbrella,
one can
further
specify
the
links
betwen
the
two
classes
of
events
by
comparing
them
in
relation
to
known
disaster
characteristics
in
general.
That
is,
one can
specify
how
nuclear
and
nonnuclear disasters
compare
relative
to
important
defining
characteristics
of
disaster
events.
There
has
been
some
discussion
of
how
nuclear
and
nonnuclear
disasters differ
in
the
early literature
on
human
response
to
natural
disasters. Most
of
this
work
was
done
at
the
Ohio
State
University
Disaster
Research
Center between
1963
and
1972
and
focused
upon
the
problem
of
assessing
the
implications
of
studies
of
natural
disaster
for
the
problem
of
nuclear
attack
(Kreps,
1979).
One
study,
conducted
by
Anderson
(1969)
examines
the
functioning
.of
civil
defense offices
in
natural disasters
and
applies
his findings
to
the
nuclear
attack
environment.
In
developing
his
analysis
Anderson
argued
that
in
spite
of
various
differences
between
nuclear
and
nonnuclear
disasters:

10
..
[these
differences]
can
be
visualized
as
primarily
ones
of
degree. With
the
exception
of
the
specific
form
of
secondary
threat,
i.e.
radiation,
and
the
probability
that
a
wider
geographic
area will
be
involved,
a
nuclear
[disaster]
would
not
create
essentially
different problems
for
community
response
(1969:55).
Therefore, Anderson
began
laying
the
basis
of
a
scheme
to
compare nuclear
with
natural
disasters
by
examining
two
important
distinguishing
features
of
disasters:
the
form
of
secondary
impacts and
the
scope
of
impact.
Allen
Barton
(1970)
advanced
a
classification
scheme
for
disasters
which
builds upon
the two
distinguishing
features
used by
Anierson.
In
his
attempt
to
characterize
the
nature
of
social
system
stress
Barton
chose
four
basic dimensions:
scope
of
impact, speed
of
onset,
duration
of
impact,
and
social
preparedness
(1970:40-47). Scope
of
impact
is
a
geographic
reference
categorizing
impact
as
involving
either
a
small
area
or
only
a
few
people
(narrow
impact),
or
as
encompassing
a
large
area
or
number
of
people
(widespread
impact).
Speed
of
onset refers
to
the
suddenness
of
impact
or
to
the
time
period
between
detection
of
a
hazard
and
its
impact
on
the
social
system.
This
dimension
is
usually
classified
as
either
sudden
or
gradual.
Duration
of
the
impact
itself
refers
to
the
time
that
elapses
between
initial
onset
of
impact
and
the
point
at
which
it
subsides.
This
can
be
a
few
minutes
(short)
in
the
case
of
a
tornado,
or
several
hours
(long)
in
the
case
of
some
riverine
floods.
Finally,
social
preparedness
is
used
in
the
context
of
possible
forewarning
to
indicate
whether
or not
the
current
state
of
technology
permits
authorities
to
anticipate
or
predict
a
threatened
disaster
impact.
In
addition
to
the
dimensions discussed
by
Barton,
we
will
also
retain
Anderson's
concept
of
secondary
impacts
in
our
scheme.
Virtually
all
hazards,
whether
nuclear
or
nonnuclear,
entail
some
secondary
I- -

11
impacts;
in
some
cases
the
secondary
impact
is
even more
devastating
than
the
initial or
primary
impact.
Riverine
floods
tend
to
deposit
silt
and
debris
over
inundated
areas,
earthquakes
involve
aftershocks
and
often
result
in
urban
fires,
tropical
cyclones
leave
great
physical
destruction,
often creating
public
health
risks.
Nuclear
power
plant
accidents potentialy
involve
radioactive atmospheric
releases
thereby
producing
a
possibly very
lingering
secondary
impact
in
the
form
of
residual
radiation.
By
assembling
lists
of
distinguishing characteristics
such
as
those
elaborated above,
one
can
classify
a
range
of
disaster
agents
and
be
alerted
to
important
distinctions
among
them. Table
1
classifies
the
three
agents
of
interest
in
this
report
in
terms of
the
five
important
defining characteristics.
It
is
interesting
to
note
at
the
outset
that
volcanoes
and
nuclear
power
plant
accidents
are
identically
classified
on
all
five
dimensions
for
comparison.
Both
hazards
involve
a
variable
scope
of
impact,
with
volcanoes' negative effects usually
extending
a
maximm
of
a
few
miles
from
the
crater,
and
plume
inhalation hazards associated with power
plant
accidents
extending
to an
approximate
10
mile
radius
from the
plant
(cf.
U.S.
Nuclear
Regulatory
Comission,
1981).
Under
special
conditions,
however,
the
scope
of
impact
may
be
considerably
greater.
The
May
18,
1980
eruption
of
the
Mt.
St.
Helena
volcano
spread
volcanic
ash
over
a
three state
area
and
a
"worst-case"
reactor
accident
involving
a
core
melt
could
affect
an
entire
region
of
the
United
States.
The
speed
of
onset
for
volcanoes
and
powerplant
accidents
is
sudden,
with
no
long
period
of
threat
before
the
initial impact.
For both
cases,
the
duration

12
1n
n
-1 mP 0
P.. 0 C a
~
-
n.C
0 0 0. 0v
0. ro
Q.
-0 '4 -
go~
11
m
0 91 0
0- (A
Ca -.
0
Ca0
to
a
.
4
P- W.4
(D "
a1C
0 o. 14 Ca-4P
Do~
A.
..
00 IA
I-
m
1 0
as -( m
.Ca 0
Ca aCal
0.0
m
a
00
11. w. 0Pa
W.f
n
~
0
Ca Ia
n
tCf
nf
0(0
0
'
0
a0
Ft .- Cao
c'

13
of impact
is
long.
A
volcanic eruptive
sequence
usually
involves
multiple
eruptive
events
over
a
period
of five
to
twenty
years
(MacDonald,
1972).
The
duration
of
impact
for
powerplant accidents
is
highly
variable,
but could involve
several
days
(the
danger
period
at
THI
extended
about
6
days). In
absolute
time,
this
is
shorter
than
a
volcanic eruption
sequence, but both
are
of
long
duration compared
to
other hazards
such
as
tornadoes,
hurricanes,
or
tsunamis.
Both
volcanic
eruptions
and power
plant
accidents
generate
secondary
impacts.
Human
settlements near
a
volcano
may experience
lasting
physical
damage
from any
of
several
agents--lava
flows,
mud
flows,
large
tephra,
ash
fall, or
flooding--and
the
aftermath
of this
type
of
damage
can
create
public
health
hazards
due
to
polluted
water
supplies,
waste
disposal,
etc.
Power
plant
accidents which
involve
atmospheric
releases
of
radiation
produce potential
secondary hazards associated
with
human
inhalation
and
possible entry
into
the
flood
chain
via
animal
ingestion.
Finally,
with regard
to
social
preparedness
or
predictability,
the
present
state
of
technology
is
such
that
neither
volcanic
eruptions
nor
power
plant
accidents may
be
forecast
in
advance. There
is,
in
both
cases
however,
a
technology
for
detecting
events once
they
have
occurred.
In
the
case
of
some
volcanoes,
once
an
eruptive sequence
has
begun,
either seismic
or
geochemical
clues
may
be
used
to
make
approximate
forecasts
of
eruptive
events..
With nuclear power
plants,
available technology
is
designed
to
detect
minor aberrations early
in
the
hope
of
taking
correction
action
before more
serious
difficulties
develop.
Thus,
while strictly speaking
one
cannot
predict power plant
accidents,
the
nature
of
the
detection
function
is
such that
by
detecting

14
malfunctions
early, subsequent
malfunctions
which
may
eventually
result
in
a
serious
atmospheric release
may
be
anticipated
(and
perhaps
prevented).
Riverine
floods
differ
from
volcanoes
and
power plant
accidents
primarily
in
terms
of
two
of
the
defining
characteristics:
floods
are
predictable,
usually
some time
in
advance,
and speed
of
onset
is
gradual,
requiring
6
hours
or
more
to
reach
a
flood
crest
(Owen,
1977).
Also,
another general
point
of
distinction
is
that
floods
occur
more
frequently
than
either
volcanic
eruptions
or
nuclear
power
plant
accidents. Thus,
from
the
standpoint
of
both
the
authorities
and
the
public,
riverine
floods
are
a
relatively
familiar
hazard,
which
can
be
predicted
in
advance,
and
that
develop
at
a
slow
pace.
Like
volcanic
eruptions
and
power
plant
accidents,
floods
have
a
variable
scope
of
impact,
usually
affecting
only
a
few
square
miles,
but
potentially
a
much
larger area.
Riverine
floods
are
characterized
by
a
long
duration
of
impact,
usually
a
few
days.
Secondary
impacts
associated with riverine
floods
include physical
damage
to
dwellings,
damage
to
arable
land due
to
silt
and
sand
deposits,
and
associated
public health
hazards.
It
has
been
argued above
that
one
can
appropriately
examine
a
variety
of
disasters--specifically
riverine
floods,
volcanoes
and
nuclear
power
plant
accidents-within
the
same
conceptual
and
analytic
framework.
The
same
basic definition
subsumes
all
of
the
events,
and they
may
be
described using
a
single
scheme
for
defining characteristics
of
disasters.
Thus,
a
careful
examination
of
the
problem
reveals
no
I --
-w
-. ,. - - . - ..

15
significant
conceptual
reason
for
treating
nuclear
and
nonnuclear
hazards
as
fundamntally
different
such
that
they
must
be
separated
and
examined
using
different
frameworks
in
social
scientific
analysis.
Unique Aspects
of
Disaster
Events
The
preceding discussion
was
meant
to
demonstrate
that
logical
and
appropriate
comparisons
can
be
made among
nuclear
and
nonnuclear
threats;
analytically,
in
terms
of the
present
state
of
disaster
research,
there
is
no
justification
for
isolating
nuclear disasters
in
a
class
by
themselves.
This
is
not
to
say,
however,
that
all
hazards--whether
nuclear
or
nonnuclear--do
not involve
some
unique
characteristics.
In
conducting
a
comparative analysis,
one
must
review
and
examine
the
implications
of
unique
hazard
characteristics
for
the
human
response
variables
of
interest.
In
this
case, our
concern
focuses
upon
one
generic
function
performed
in
disasters:
population
evacuation.
More
specifically, we
are
interested
in
people's perceptions
of
warning
source
credibility
and
their
reasons
for
evacuating.
The
following
paragraphs
briefly
highlight
several unique aspects
of
the
nuclear
hazard
as
a
means
of
facilitating
our
comparative
analysis
by
noting
specific
qualifiers
which
may
be
incorporated
into
subsequent data
analyses.
As
a
disaster
event,
the
most
unique
aspect
of
a
nuclear
power
plant
accident
is
that
a
nuclear component
is
involved.
Thus,
some
attention
is
necessary
because,
in
terms
of the
way
people
perceive
the
situation,
such
circumstances
are
different
from
those
which
accompany other
disaster
agents.
Research
indicates
that
some of
the
public
views
nuclear energy,
and
most applications
of
it,
as
a
particularly
1-
"T-m

16
threatening
hazard with
the
potential
for
extraordinarily
long-term
negative
effects--literally
the
power
to
irreversibly destroy generations
(Lindell,
et
al.,
1978).
Of
course,
the
idea
that
people
have
a
different "mind
set"
for
nuclear
disasters
certainly
does
not
preclude
comparisons with nonnuclear
disasters.
Instead
it
only
requires
that
this
"emotional"
dimension
be
acknowledged
and
that
the
necessary
qualifications
be
made
when
such
perceptual
differences
may
have
some
bearing
upon
human performance.
Two
aspects
of
this
emotional
dimension
should
be
mentioned
here:
risk
perception
and
experience.
The
agent of
threat
to
the
human
population
in a
nuclear
power
plant
accident
is
nuclear radiation.
In
contemporary
American
society,
this
agent
is a
high
fear-generating
mechanism
regarding which
the
public
at
large
is
poorly
informed
(Kaplan,
1978;
Rankin
et
al.,
1978).
Furthermore,
surveys indicate
that
much
of
the
information
that
the
public
does
hold
about
nuclear
power plants
is
technically
incorrect
(Earle,
1981).
This
situation
produces
an
environment where
some
people
potentially have exaggerated conceptions
of
the
destructive
potential
of
an
accident,
while
others
may
believe
that
negative consequences
are
of
less
concern.
Also,
there
is
widespread
disagreement
on
what constitutes
a
source
of
acceptable
("accurate")
information
about
nuclear
hazards,
particularly
power plants
(Martin,
1980).
Thus, public
perception
of
danger
associated with nuclear
power
plants
is
highly
variable,
and
there
are
few
sources
of
information
perceived
to
be
acceptable
which might
serve
to
promote
a
more
homogeneous definition
of
threat.
That
is,
through
selective choice
of
information,
individuals
with
extreme attitudes,
whether exaggerating
or

17
minimizing
risks,
can
locate
sources
which
reinforce
their point
of ii
view.
Such
circumstances
tend
to
exacerbate
the
problems
associated
with
emergency planning
and
response.
The
second
aspect
of
the
emotional
response
to
nuclear
disasters
is
that
most
citizens
lack
a
reference
point
in
their
experience
for
such
events.
Only
one
reactor
accident
involving
potential threat
to
offsite
populations
has
occurred
in
the
United
States,
and
this
involved
an
area
of
comparatively
small
size
around
Harrisburg,
Pennsylvania.
While
the
media coverage
was
extensive,
the
majority
of
the
population
has
at
best
only
vicariously experienced
the
power
plant
accident.
Consequently,
unlike
the
situation which
prevails
with natural
disasters,
one
cannot
expect people's
"prior experience"
with
nuclear disasters
to
help
them
arrive
at
a
definition
of
threat
associated
with
a
given
nuclear disaster.
Indeed,
the
effects
of the
accident
at
Three
Mile
Island
upon
public
perception
of
risks
associated with
nuclear
power
plant
accidents
are
unclear.
Three Mile
Island
was
a
localized
threat,
characterized
by
apparent
confusion
of
all
parties
involved,
a
shortage
of
visible,
strong
official leadership and
shrouded
in
conflicting
accounts
in
the
mass
media
(cf.
Flynn,
1979;
Chenault,
et
al.,
1979;
Sandman
and
Paden,
1979).
In
the
short
run,
the
incident
produced
two
general
consequences:
(1)
it
resulted
in
intensive
dissemination
of
a
variety
of
information
(some
technically accurate
and
some
not)
regarding nuclear
power
plant
safety;
and
(2)
the
apparent
confusion
and
slow
action
initially
on
the
part of
officials
raised doubts about
the
capability
of
authorities
to
handle nuclear
disasters.
On
the
other hand,
in
spite
of
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_

InI
18
the
attention
given
the
incident
and
whatever
its
seriousness
uay
have
been,
no
documented negative health
effects have
been observed
in
the
local
population.
In
closing
this
section
on
the
unique
aspects
of
hazards,
it is
important
to
point
out
that,
from
the
public's
point
of
view,
volcanoes
share
some
of
the
emergency response
problems
associated
with
nuclear
power
plant
accidents.
Volcanic
eruptions
are
not
common, particularly
in
the
continental
United
States,
and
public
experience
with them
is
almost
nil.
Furthermore,
public
knowledge
of
the
risks
associated
with
volcanoes
is
limited and
sometimes
technically inaccurate
(Perry
et
al.,
1980a).
In
the
case of
volcanoes,
however,
there
is
an
identifiable
body
of
publicly accepted
sources
of
information
about
the
hazard.
Thus,
there
is
an
available source
of
threat
relevant
data
which
the
public
may
use
in
devising
or
arriving
at
situational
definitions
of
threat.
Finally,
the
purpose
of
this
discussion
has
been
to
document
special
aspects
of
hazards
which
may
be
helpful
in
interpreting
human
response
data.
As
it
was
pointed
out,
the
simple
presence
of some
unique
characteristics
does
not
justify
separating
the
analysis
of
nuclear
and
nonnuclear
disasters.
Instead,
such
distinguishing
features
should
be
acknowledged
and
treated
as
factors
deserving
special
attention
in the
context
of
comparing human
response
to
nuclear
and
natural disasters.
The following
chapter
presents
descriptiv.e
data
on
each
of
the
disaster
events
to
be
compared
and
relates
a
largely
nontechnical
account
of
the
circumstances surrounding disaster
impact.
The
purpose
of
these
descriptions
is
twofold:
to
provide
an
overview
of
the
disaster
incident
and
to
convey
general
information which
will
make interpretations
of
between-event
comparisons more meaningful.
._ _ _ _ _ _ _ _ _ . . . . . .-= , _ . _ .:
-.-
.= . . . ..

19
CHAPTER THREE
THE
DISASTER
EVENTS
This report compares
selected
aspects
of
human
response
to
volcanic
eruption,
riverine flooding
and
a
nuclear
power
plant
accident. The
following
descriptions
provide
an
overview
of the
disaster
in
each
relevant
community.
Volcanic
Eruption
Late
in
March,
1980,
Mt.
St.
Helens,
Washington,
resumed volcanic
activity
after
123
years
of
dormancy.
In
general
the
public
responded
with excitement
and
curiosity
to
this
activity.
News
media
devoted
much
attention
to
the
small
steam
and ash
eruptions.
As
it
became
apparent
that
the
volcano
was
not
going
to
settle
quietly back
into
dormancy,
public
officials
in
the
surrounding
counties
and
in
several
federal
agencies
developed
or
strengthened existing emergency
plans.
Scientists,
particularly
those from
the
U.S.
Geological Survey, provided
information
to
the
media
and
officials
concerning
likely scenarios
of
future
volcanic
activity.
The
Cowlitz County
Sheriff's
Office prepared
a
pamphlet
describing
their
warning system
and
distributed
it
to
residents
along
the
Toutle
and
Lewis
River drainage
areas.
The
public
maintained
a
high
level
of
interest
throughout
this
six
week
period
from
initial
activity
to
the
cataclysmic eruption,
fostered
in
part
by
the
media's
attention
on
the
volcano.
While
there
is
some
evidence
that
citizens
in
the
vicinity
of
the
mountain
were
concerned
---
-------
'I

20
that
specific
contingency
plans be
developed
and
that
officials
be
prepared
for
a
major
eruption (Perkins,
1980),
there
is
also
evidence
that
the
public
felt
officials
were
too
restrictive
in
their
policies
concerning
access
to
the
volcano.
Cougar residents,
for
example, were
reported
as
being
angry
by
the
roadblocks which
cut
their
town
off
from
a
"booming" volcano
business
(The Columbian,
1980:20).
The
cataclysmic eruption began
when
an
earthquake
of
approximately
magnitude
4.9
was
recorded
at
8:32
a.m.
on
Sunday, May
18th
(Rosenfield,
1980:498).
This
earthquake apparently
triggered
a
tremendous landslide
on the
north
side
of
the
volcano
which
led
imediately
to
the
explosion
(Geophysics
Program,
1980:530).
A
member
of
the
U.S.
Geological
Survey
volcano
team
described
the
eruption
in
detail,
writing
that
this
avalanche
was,
within
seconds,
overtaken
by
a
large
laterally directed
blast
that
exploded
out,
with
hurricane
force
winds,
more
than twenty
kilometers
from
the
volcano's
sumit
(Christiansen,
1980:532).
The
avalanche
then formed
a
debris
flow
that
mainly
turned and flowed
down
the
valley
of
the
North Toutle
River
for
18
kilometers.
The
displaced
water
of
Spirit
Lake,
the
melting
blocks
of
ice
from
the
former
glaciers
on
the
volcano's north
flank,
water
from
the
displaced
river
bed,
and
melting
snow
and
ice
on
the
volcano's remaining
slopes
produced
mudflows
that
flooded
the
debris
flow
and
generated
floods
all
the
way
down
the
Toutle River,
the
Cowlitz River
and
eventually
the
Columbia River
(cf.
Christiansen,
1980:532).
These mudflows
and floods
destroyed
bridges,
roads
and
homes
and
filled
the
channel
of the
Columbia River,
temporarily stranding
ocean-going
ships
upstream
in
the
Port
of
Portland.

21
The
effects
of
the
eruption
were tremendous.
The
once
symmetrical
9,671-foot
peak
now
has
a
rim
that
reaches
a
reported 8,400
feet
at
its
highest
point.
The north
flank
opening
to
the
crater
is
now
at
about
the
4,400
foot
elevation (Korosec
et
al.,
1980:16).
The
blast
destroyed
150
square
miles
of forest,
killing vegetation
and
wildlife.
Sixty-eight
people
have
been
listed
as
killed
or
missing. Three
billion
board
feet
of
timber
valued
at
approximately $400,000,000
were
damaged
or
destroyed
(U.S.
Senate
Hearings,
1980:151),
169
lakes
were either
moderately
damaged
or
destroyed,
and
over
3,000
miles
of
streams
are
either
marginally
damaged
or
destroyed
(U.S. Senate
Hearings,
1980:139).
In
total,
after
the
first
two
major
eruptions
(May
18
and
May
25)
it
was
estimated
that
damages
totaled
more
than
$1.8
billion
in
property
and
crops;
this
included
damages
in
the
vicinity
of
the
volcano
as
well
as
those
areas
that
suffered
from
the
ash
fall
(U.S. Senate
Hearings,
1980:18).
Toutle
and
Silverlake,
Washington,
constitute
adjacent
unincorporated
areas
in
Cowlitz County approximately
25
miles northwest
of
Mt.
St.
Helens,
situated
along
the
Spirit
Lake
Highway. Year-round
area
residents
are
for the
most
part
involved
in
some
aspect
of
the
logging
industry. The other
mainstay
of
the
local
economy
is
tourism.
Toutle
and
Silverlake
are
located
just
north
of
the
point
at
which
the
north
and
south
forks of
the
Toutle
River
join. The
area's
population
is
relatively
small,
approximately
1,500.
Few
people
in
the
Toutle/Silverlake
area
reported
hearing
any
noise
from
the
initial
eruption
at
8:32
a.m.
on
May
18.
For
most
the
first
evidence
of
the
eruption
was
the
huge
mushroom-shaped
ash
cloud which
_-
_
_ _
. . . .

22
filled
the
horizon
to
the
south.
Residents
reported feeling
a
dramatic
increase
in
temperature;
with
it
came
the
sounds
of
trees
and
automobile
windshields
cracking
from
the
heat.
The
area also
experienced
light
ash
fall;
the
ash
cloud
reached
Silverlake
about
one
and
a
half hours
after
the
eruption (Korosec
et
al.,
1980:14).
The
most
serious
threat,
however,
was from
mudflows
and
flooding.
After
the
blast,
the
water temperature
in
the
Toutle River
rose
above
80
degrees
farenheit;
these
temperatures
and
the
mudflows
contributed
to
the
destruction
of
most
of
the
anadramous
fish
in
the
river. The
mudflows
and
floods
caused
the
river
to
rise
well above
its
banks.
Seven
state
highway
bridges
and
numerous
county
and
private
bridges
over
the
Toutle were
destroyed,
as
well
as
almost
300
homes
in
low-lying
areas
nearby
the
river.
Fortunately, most
of
the
communities
of
Toutle
and
Silverlake
lie
on
the
slopes
above
the
Toutle
River
and
were
minimally
affected
by
the
mudflows
and
floods.
Official
concern
about
flood
danger along
the
Toutle
remained
high
for the
several
days
immediately
following
the
eruption.
The
eruption
had raised
and
reshaped
Spirit
Lake which
fed
into
the
north
fork
of
the
Toutle River. Down valley
from
Spirit
Lake,
a
large
debris
flow
raised
the
valley
floor
of
the
South
Fork Toutle
River
by
several
hundred
feet
for
a
distance
of
about
14
miles.
At
first
the
massive
debris
flow
was
thought
to
be
only
marginally
stable,
but
a
study
of the
deposit
by
soils
engineers concluded
that
there
was
virtually
no
possibility
that
it
would
become
remobilized
and
move
on
down
valley.
Most residents
were,4lerted
by
Cowlitz County
Sheriff's officials
of
the
initial
eruption.
The
deputies drove
predesignated
routes,
using
their
high-low
sirens
and
their
public
address
systems.
A
telephone
1-
------

23
ring-down system
was also
implemented, again
in
predesignated
areas
that
had
a
high
probability
of
flooding.
Although
the
Toutle
Fire
Department
did
not
receive official
notification
of
the
eruption
from
the
County
Sheriff's
Office,
as
had been
arranged
in
pre-eruption planning meetings,
once
there was
physical
evidence
of
the
eruption
Fire
Department
volunteers
assisted
in
the
warning. They
also
helped
man
the
roadblocks
to
keep sightseers
out
of
the
area.
A
large
proportion
of
the
residents
evacuated,
a
process
which
was
facilitated
by
unfounded
rumors
that
a
cloud
of
poison
gas
was
moving
toward
Toutle
and
Silverlake.
Flood
Events
Our
data
on flood
events
are
drawn
from
three
cases.
The
floods
affected
three
communities
of
similar
size.
Each incident also
involved
similar
disaster
characteristics;
in
all
three
cases,
authorities
were
forewarned
and issued
pre-impact
warnings
to
citizens,
evacuations
successfully occurred,
and
duration
of
impact
was
approximately
the
same.
In
subsequent
comparative analyses,
data
from these
three
events
will
be
pooled
to
represent
flood
response information.
In
this
chapter,
however,
a
separate
overview
is
given
for
each event.
Valley
Valley
is a
small
comunity
on
the
Platte
river
a
few
miles
northwest
of
a
major
midwestern
rail
and
commercial
center.
The
community
is
sustained
by
railroad
interests,
a
large
manufacturing
firm,
and
some

24
agricultural
and
livestock
enterprises.
Valley
has
a
long
history
of
spring
floods.
After
a
severe incident
in
March
1912,
construction
began
on
a
levee
system
along
the
Platte
which
was
completed
in
1919.
Levee
failures
that
resulted
in
flood
waters
reaching
Valley
have occurred
only
three
times
since
the 1912 flood;
these
were
during
1948,
1960
and
1962.
In
mid-March
of
1978,
the
national
Weather
Service
issued
a
flood
watch
for
the
lower
Platte,
bringing
to
the
attention
of
the
news
media
the
presence
of
ice
jamming
and
lowland flooding
along
the
river.
Although attempts
were
made
to
break
up
the
ice,
rising
water resulted
in
the
erosion
of
Union
Dike located
approximately
three
miles
north
of
Valley
on
the
evening
of
March
19.
This
marked
the
beginning
of
the
most
severe
flood
in
the
town's
history.
Although
there
were
no
deaths,
property
damages were extensive; damages
to
railroad
equipment
alone
exceeded
two
million
dollars.
Most
of
the
private
residences
in
Valley
experienced
some
water
damage, ranging
from
basement
flonding
to
major
structural
failure.
Water,
sewer
and
natural
gas
lines
were damaged
and
services interrupted.
These
problems
kept
most
residents
from
their
homes
at
least
48
hours
and
many
could
return
only
after
four
to
five
days.
Virtually
all
of the
town's
residents received
advance
warning
of
the
flood.
On
Saturday, March
18,
the
Volunteer
Fire
Department
initiated
patrols
of the
levee
which
protects
Valley.
Thus,
when
Union
Dike began
to
crumble
on
the
evening
of
the
19th,
the
problem
was
detected
promptly
and
warning
radioed
to
Valley
as
well
as
other
nearby communities.
Water
did
not
reach
Valley
for
approximately
three
hours.
In town,
a
command
post
was
organized
at
City
Hall.
The
fire
and
police
departments

25
contacted appropriate outside
agencies
for
help
and
coordinated
all
emergency
services
in
Valley.
Warnings
to
evacuate were
issued
via
public
address
systems
on
emergency
vehicles
and
door-to-door.
Civil
defense
sirens
were
also used
to
issue an
evacuation
alert.
Most
residents
were
warned
a
minimum
of
30
minutes
prior
to
flood
impact
and
some
had
as
much
as
3
hours
notice.
These
remarks
refer
to
a
warning
to
evacuate
because
of
imminent danger;
the
mass media
had
"warned"
that
the
ice
jams
could
produce
flooding
for
two
days
prior
to
impact.
Approximately
90
percent
of
the
households
and
one large
nursing
home
were
evacuated
by
the
next
morning.
About three-fourths
of
these
evacuations
were
accomplished
prior
to
impact.
The
Red
Cross
and
Salvation
Army
provided shelter
for
evacuees,
first
in
Valley
itself,
then
in
nearby
towns
when
high water
required
relocation. More
than
650 families
registered
at
the
Boys
Town
shelter
established
by
the
Red Cross. Length
of
stay
at
the
shelter
tended
to
be
quite
short.
Many
families
stayed
only
long
enough
to
assess
damages
to
their
homes
and
arrange
to
stay
with
relatives
or
friends
in
nearby
comunities.
Moat people
were
gone
from their
residence
at
least
four
days,
the
period
necessary
to
reestablish basic
services
in
the
community.
Fillmore
Fillmore
is a
Western
community
of
about 8,500.
The
citrus
and
railroad
industry
are
major
local
employers.
The
couusnity
is
located
near
the
Santa
Clara River, where
it is
joined by
a
tributary,
Saspe
Creek. The
Sespe has
flooded
at
least
six
-imes
since
1962, the
greatest

26
damage being
inflicted
in
1969.
Although
flood
control
plans
for
the
Sespe
are
currently
being
considered,
at
the
time of the
flood
no
man-made
levees
existed
in
the
area
around
Fillmore.
Early
on
the
morning
of
March
4,
1978,
the
Sespe,
swollen
by nearly
nine
inches of
rain
in
a
24
hour
period,
began
to
overbank.
As the
banks
began
to
fail,
the
Sespe
in
effect
was
diverted
through
the
west
end
of
Fillmore.
To
make
matters worse,
Highway
126,
which
connects
Fillmore
with
nearby
towns,
is
a
raised
highway.
Debris
accumulating
under
the
bridge dammed
the
Sespe,
creating
a
lake
in
the
low-lying
areas.
When
the
highway
was
bulldozed
to
stop
the
formation
of
the
lake, at least one
main
phone
line was
severed,
considerably
increasing
the
extent
of
Fillmore's
isolation.
Flood
damages
in
Fillmore exceeded
six
million
dollars.
Nearly
200
homes
sustained
major
structural
damage
and
approximately
1,200
people
evacuated
from
their
homes. Most
of
the
damage
and
half
of
the
evacuations
occurred
in
the
extreme
west
end
of town.
One
man
was
killed
when
his
home collapsed
due
to
water
erosion.
Warnings
to
evacuate
in
Fillmore were delivered
by
police
and
fire
department
personnel both
door-to-door
and
by
public address
systems
from
patrol
cars
and
helicopters. The
process
of
warning
residents
began
at
approximately
6:00
a.m.
on
March
4;
the
flood
waters
reached
a
peak
at
about
2:00
p.m.
the
same
day.
Most
evacuees
were directed
to
a
Red Cross
shelter
located
in
a
nearby
school
gymnasium.
By
the
evening
of
March
4,
police
cordoned
off
the
flooded area
to
maintain
security,
and
evacuees
were
prevented
from
returning
to
their
homes
until
the
following day.
--
_ _
_

27
Snoqualmie
Approximately
1,300
people
reside
in
the
northwestern
U.S.
town
of
Snoqualmie which
is
situated
along
the
south
bank
of
the
Snoqualmie
River.
The
town
is
supported
primarily
by
the
timber
products
industry
and
some
tourism. Snoqualmie
and
surrounding
communities
have
historically
been
subject
to
late
fall
and
winter
floods.
An
extensive
monitoring
system, operated
by
the
county, exists
on
the
Snoqualmie
River
and
advance
warning
of
imminent
flooding
is
provided
directly
to
emergency
services offices
in
threatened
communities.
On
December
1,
the
County
Flood Control
office
informed
Snoqualmie
officials
that
the
river
was
rising
and
that
flooding was
very
likely
to
occur.
Fire
Department
volunteers
began
a
twenty-four hour
"river
watch"
immediately,
and
just
after
midnight
on
December
2, a
warning
to
evacuate
was
issued
to
residents
of
low-lying
areas.
The
area
to
be
warned
contained
approximately
200
households.
Warnings
were
delivered by
fire
and
police
department
personnel
using
several
methods. Initially,
street public
address
systems
and
door-to-door
contacts
were
used.
Some
telephone
contacts
were
also
made
and
authorities acknowledged
that
considerable
informal
"word
of
mouth"
warnings
were
exchanged among neighbors.
In
Snoqualmie
the
local
high
school and
a
large
church
are
designated
in
emergency
plans
as
shelters.
Families
for
whom
official
transportation
was
provided were
taken
to
one
of
these
locations. Most
evacuees provided
their
own
transportation,
however,
and
tended
to
go
to
the
homes
of
friends
or
relatives. Most
evacuees
could return
to
their

28
homes
within
twenty-four hours.
A
Red
Cross shelter,
where evacuees
could
obtain
vouchers
to
pay
for
lodging
and
meals,
was
established
at
noon
on
December
2,
but was
sparsely
utilized.
Flood
damages
in
Snoqualmie
were
comparatively
low:
approximately
$500,000.
About
80
dwellings were damaged
by flood
waters,
as
well
as
local
bridges
and
roads.
The
damage
figures
are low
because
most
commercial
and
industrial
enterprises
are
located on
high ground.
There
were
no
deaths
or
injuries
as
a
result
of
the
flooding.
Nuclear Power Plant
Accident
The
reactor
accident
at
Three Mile
Island
(THI)
is
probably
best
described
as
an
extremely
complex
event
which
has
been
the
subject
of
volumes
of
description
in
the
print
and
broadcast
media,
as
well
as
a
number
of
technical
and
scientific
studies
(cf.
Kemeny,
1979;
Martin,
1980).
Technically,
of
course,
ThI
was
not
a
disaster;
the
major
environmental release
of
radiation which
would constitute
a
disaster
was
precluded.
The
situation
may
be
technically characterized
as
an
emergency,
however,
and
the
evacuations
which
occurred
in
connection
with
the
nuclear
threat
may
be
compared
with evacuations
in
the
face
of
other
threats.
To
attempt
a
brief
overview
of
the
event
itself
is
a
difficult
undertaking,
and
by
necessity
must
focus
upon
a
few
milestones
rather
than trying
to
portray
each
facet
of
the
incident.
This overview
concentrates
on
milestones associated
with
three
general
human
response
issues:
the
nature
of
warning
information disseminated;
the
commanication system
for
dealing
with
the
public;
and
the
outcomes
of
the
incident itself.
. ...

29
The
accident
at
Unit
Two
of
ThI
began
at
approximately
4
a.m.
on
March
28,
1979,
with
a
malfunction
which disabled
the
reactor's pneumatic
control
system. The
accompanying
heat
and
pressure, coupled with
a
mechanical
failure, resulted
in
hundreds
of
thousands
of
gallons
of
radioactive water being
pumped
into
the
containment
building,
and
then
into an
adjacent
auxillary building.
The
ventilation
system
in
this
auxillary building
pumped
some of
the
highly
radioactive
gases
which
accompanied
the
water
into
the
atmosphere.
At
approximately
6:50
a.m.
radiation
alarms sounded
and
reactor operators declared
a
site
emergency.
After
the
site
emergency
was
declared,
the
notification
process
was
initiated
and
contacts were
made
with
local,
county
and
Pennsylvania
State
authorities,
as
well
as
regional
and
headquarters
offices
of
various
federal agencies:
the
Nuclear
Regulatory
Commission,
Department
of
Energy,
Defense
Civil
Preparedness
Agency,
Environmental Protection
Agency
and
Food
and
Drug
Administration.
For
the
rest
of the
day,
Wednesday
and
Thursday,
contacts
and
information
exchanges, involving
many
conflicting
and
garbled messages,
took
place among
Metropolitan Edison
officials,
reactor operators,
county,
state
and
federal
agencies
and
the
Governor's
Office
(cf.
Martin,
1980:47-130).
Representatives
of the
national
and
international mass
media
converged
on
the
site
(Sandman
and
Paden,
1979).
Pennsylvania
Emergency Management Agency
and
county officials
were
advised
of
a
possible
need
for
evacuation
of
civilians
and
began
preparing
for
such
an
eventuality. Although
"reactor technicians suggested
that
the
machine
was
under
control
and
slowly
returning
to
normal," some
local
residents
began
to
leave
the
area
(Chenault,
1979:5).

30
On
Friday morning
at
approximately
8
a.m.,
a
significant release
of
radiation
was
detected
and
a
general
emergency
was
declared
at the site
(Donnelly
and
Kramer,
1979:23).
This release
was
apparently
"uncontrolled"
and
further
uncontrolled
releases
were
believed
to
be
possible. Information
was
released
to
the
NRC
regarding
the
presence
of
.a
hydrogen
gas
bubble
in
the
reactor
which
was
growing
in
volume
and
making
the
task
of
cooling
the
core
more
difficult
(Martin,
1980:229).
At
approximately
noon,
Governor Richard
Thornburgh
issued
an
advisory
that
pregnant women
and small
children
living
within
five
miles
of
TMI
evacuate
and
people living
within
a
ten
mile
radius
should
stay
indoors
(American
Nuclear
Society,
1979:4).
Following
this
Friday
evacuation
advisory,
approximately
12,180
persons living
within
five
miles
of
ThI
and
31,360 persons
living
within
a
five
to
ten
mile
ring evacuated
(Flynn,
1980:16).
These
figures
represent
35%
of
the
total
population
within
five
miles
and
252
of
the
total
population
of
the
five
to
10
mile
"ing.
The
number
of
representatives
of
Federal agencies
at the
site
continued
to
grow;
by
Friday
evening
83
NRC
personnel were
either
on
site
or
in
the
area
(Donnelly
and
Kramer,
1979:23).
By
Saturday official concern about
the
hyrdrogen
bubble
was
increasing.
At
approximately
2:30
p.m. Chairman
Joseph
Rendrie
of
the
NRC
held
a
news
conference
and
announced
that
the
hydrogen bubble
could
potentially
explode. Federal
and state
officials
discussed
the
possibility
of
extending
the
plans
for
potential
evacuation
to
a
twenty
mile
radius
around
the
reactor
site.
The spontaneous
(that
is,
not
officially
ordered) evacuation
of
citizens
living near
the
power plant
continued.
By
late
afternoon,
NRC
staff
determined
that
the
hydrogen
•, ,,,•_______________-

31
bubble
could
not explode; NRC
representatives
Harold Denton
and
Governor
Thornburgh
held
an
11
p.m.
news
conference
to
announce
this
and
President
Carter's
visit
on
Sunday
(cf.
Martin,
1980:230).
By
Sunday, April
1, it
was
determined
that
the
hydrogen
bubble
was
shrinking
and that
the
reactor appeared
to
be stable
(Donnelly
and
Kramer,
1979:22).
President
Carter
made
a
well
publicized
visit
to the
reactor
site.
Evacuation
readiness
preparations
were
continued
in
nearby
counties.
On
Monday
it
was
announced
that
the
hydrogen
bubble had shrunk
to
150
cubic
feet
and
was
still
diminishing
(Martin,
1980:231).
Civil
Defense
officials
noted
that large
numbers
of
citizens
had
already evacuated
the
area
and
absenteeism
was
creating
labor
difficulties
in
Harrisburg
(Donnelly
and
Kramr,
1979:22).
County
and
State
authorities continued
to
formalize
plans
for
possible
evacuations
and
the
Food
and
Drug
Administration
recomiuended
that
potassium
iodide
tablets
be
distributed.
Late
Monday
evening
the
situation
at
the
reactor
had
stabilized enough
that
the
NRC
agreed
to
let
Metropolitan Edison
allow
the
reactor
to
cool
without
depressurization.
By
Tuesday,
the
crisis
had
begun
to
subside.
The
hydrogen
bubble
had
significantly
reduced
in
size,
thereby
reducing
the
likelihood
that any
evacuation
of
the
general
population
would
be
necessary.
Schools
located
near
the
THI
site
were
reopened
on
Wednesday. People who had
left
the
area
began
to
return
home.
It
is
estimated
that
144,000
people
living
within
a
fifteen
mile
radius
evacuated
their
homes
at
some
point
between
March
28
and
April
3;
this
is
approximately
39
per cent
of
the
total
population
(Flynn,
1979:14).
On
April
9,
Governor
Thornburgh
advised
pregnant women
and
young children
to
return
to
their
homes.
--w---

32
The
precise
severity
and
consequences
of
the
reactor
accident
are
difficult
to
assess,
even
two
years after
the
event.
While
the
potential
for
human
deaths and
environmental
contamination
is
very high
in
reactor
accidents,
there
were
no
deaths
at
TMI
and
comparatively
little
environmental
contamination.
It
was
estimated
that,
as
a
function
of
atmospheric releases,
persons
living
within
a
50
mile
radius
of
TMI
received
an
average
radiation
dose
equal
to
about
one per
cent
of
the
annual
background
radiation
level;
persons living
within
five
miles
received
an
average
dose
of
about
10
percent
of
the
annual
background
level
(Kemeny,
1979:34).
Even allowing
for
errors
in
measurement,
these
doses
are
so
small
that
the
President's
Comission
on
the
accident
reported
that
there
will
be
no
detectable
physical
health
effects.
Three
TMI
employees received
larger
doses
during
the
course
of
the
accident,
but
even
these
doses
were
not
major.
Transient
mental
health
disorders
were
believed
to
be
"the
major
health
effect"
due
to
the
accident
(Kemeny,
1979:35). No
evacuations
were
officially
"ordered";
an
advisory
was
issued
for
pregnant
women
and
school
children. While physical damage
to
the
ThI
reactor
was
extensive,
the
major consequences
of the
accident
appear
to
be
related
to
identifying
specific improvements
necessary
in
the
capability
to
respond
to
power
plant
emergencies
at all
levels.
AA

33
CHAPTER
FOUR
COMPARISONS OF
WARNING
SOURCE AND
CREDIBILITY
The
sources
from
which
citizens
receive information
regarding
the
disaster
event
and
their
assessment
of
these
sources
are
important
in
understanding
patterns
of
evacuation behavior. Disaster
research
in
general
has
shown
that
the
source from
which
disaster
warning
information
is
received
is
related
to
how
the
warning
and
the
hazard
are
evaluated
and
what
immediate
reaction
is
undertaken (Perry,
et
al.,
1980a:73;
Windham
et
al.,
1977:39;
Mileti
and
Harvey,
1977:5;
McLuckie,
1970:38).
Furthermore,
the
importance
of
a
warning
source
shows
up
in
a
variety
of
contexts.
For
example,
knowing
the
source
from
which
individuals
first
received information regarding
a
disaster
event
can
be
used
to
draw
inferences about
(1)
the
response capacity
of
the
emergency preparedness
systems,
and
(2)
the
relative
speed
with
which
different
comnunication
channels
to
the
public
operate.
Also,
research
shows
that
a
first
step
toward
getting citizens
to
evacuate
is
accomplished
when
the
individual
receives
a
warning message
from
a
source
perceived
to
be
credible (Perry,
1979;
Mileti,
1975:210;
Anderson,
1968:299;
Williams,
1964:94;
Janis,
1962:59). Studies
of
the
differential
credibility
of
warning
sources provide feedback
regarding
whether
or
not
the
official
emergency
response system
itself
is
credible,
whether
the
way
in
which
a
particular
warning
was
handled
affected
credibility,
and
to
what extent other sources
already
viewed
as
highly
credible
might
be
incorporated
into
the
emergency response system.

34
Understanding
the
differential
credibility
of
sources
also
allows
authorities
to
evaluate which
ones
are
most
useful
for
delivering
imediate
warning
information
regarding
a
disaster
in
progress
and
which
ones
are
most
effective
for
communicating
information
about ways
of
planning
for
and
coping with
a
hazard
or
risk on
a
longer
range
basis.
In
this
chapter
we
will
compare
source
of
first
disaster information
and
source
credibility
among
one
nuclear
and
two
nonnuclear
events.
Source
of
First Information
Table
2
shows
the
source from
which residents
nearby
the
reactor
at
Three
Mile
Island
first
learned
of the
accident.
The
majority
of
respondents--69 percent--cited
the
mass
media
as
first
source,
with
most
of
these
mentioning
radio
as
the
specific source.
Virtually
all
other
respondents--29 percent--first
heard
about
the
accident
from
a
social
contact,
primarily
friends,
neighbors,
relatives
or job
colleagues.
Interestingly,
almost
no
one
reported
that they
first
heard
of
the
accident
from
an
"official source"--that
is,
from
a
contact
with
emergency
management
personnel
or
a
responsible
local
or
state
governmental
official.
At
TMI, then,
the
mass
media
provided
initial
information
to
the
largest
proportion
of
citizens
and
social
network
contacts
accounted
for
most
of
the
remainder.
This
finding
is in
part
a
function
of
the
nature
of
the
accident:
it
was
completely unanticipated
and
the
seriousness
and
likely
consequences
of
the
accident
for
the
public
were unclear.
Thus,
the
absence
of
forewarning meant no
apparent
time
for
emergency response officials
to
confer
and
communicate
directly
with
the
public.
In
performing
a
"notification
function",
the
newsmedia

35
TABLE
2
FIRST
SOURCE
OF
INFORMATION:
THREE
MILE
ISLAND*
Source
N %
Radio
186
52.0
Television
50 14.0
Newspaper
11
*3.O
Radio
Truck
(Authority)
7 2.0
Friends/Neighbors
46
13.0
Job
Colleagues/Employer
18
5.0
Other
(Relatives,
etc.) 41
11.0
*Adapted
from
Barnes
et
al.
(1979:13).

36
broadcast
routine
notification
to
the
public.
Hence
the
first
announcement made
by
Lt.
Governor
William
Scranton
in
effect
simply
informed
the
media
of
the
accident,
without describing
the
role
to
be
played
by state
or
local
emergency response professionals.
Table
3
shows
source
of
first
information
for
the
two
nonnuclear
disasters:
the
May
18
eruption
of
Mt.
St.
Helens volcano
and
the
floods.
These
data
show
a
very different pattern by which citizens
first
heard
of the
disaster
events.
In
each
type
of
disaster,
most
respondents--nearly half--first
heard
of
the
event
from
local
emergency
response authorities.
The next
largest
proportion
of
citizens, again
in
both
disaster
events, cited
social
networks
(neighbors,
friends
or
relatives)
as the
first
source
of
information. The
mass
media accounted
for
only
a
small
proportion
of
the
first
contacts.
These
findings
represent
a
reasonably
common
pattern
of
first
source
contacts
in
natural
disasters
(Perry
et
al.,
1980a):
emergency response authorities
constitute
the
first and
primary
sources
of
information, supplemented
by
informal
contacts
in
overlapping
social
networks.
The pattern
sometimes
e
varies with higher
dependence
on
social
networks
when
the
disaster
occurs
with
no
forewarning.
In
natural disasters,
however,
little
forewarning
rarely
results
in
heavy dependence
on
mass
media.
To
a
certain
extent,
the
differences
in
patterns
of first
warning
source
between
TMI
and
the
natural
disasters
may
be
understood
in
terms
of
the
forewarning
issue.
However,
the
differences
point
to
an
important
j
controi
distinction
in
the
pattern
of
the
emergency response
to
the two
types
if
disaster.
In
natural
disasters, control
and
communication
tend
to
remain with
local
authorities
and
the
mass
media
play
a
less
distinct
role
during
the
emergency
time
phase.
Even
in
the
case
of
brief
1
--
,.,
,_
__ I_ II.
'n n_ _
_

37
TABLE
3
FIRST
SOURCE
OF
INFORMATION:
NONNUCLEAR DISASTERS
Volcano
Flood
Source
N % N %
Neighbor/Friend
21
23.3
107
24.7
Relative
13
14.4
60
13.9
Local
Emergency Authorities
38
42.2
209
48.3
Mass
Media
9
10.0
49
11.3
State
or
Other
Authorities
3
3.3
0 0.0
Saw
Eruption
or
High
Water
6
6.7
8 1.8
41
--- V - -
.-
-, -- . |- n

38
forewarning when authorities
may
not
be
highly
visible
initially,
they
tend
to
assume
generally undisputed
control
of
communications
and
disaster operations early
in
the
emergency
period.
Two important
factors
in
this
control are
that
in
natural
disasters:
(1)
technical
status
reports
on
the
disaster
go
from
experts
to
emergency
response
authorities
who
incorporate
the
information
into
their
plan
and
interpret
the
data
for
the
public,
and
(2)
emergency
response authorities
are
traditionally
visible
to
and
recognized
by
the
public
as
being
responsible
for
protecting
the
citizenry.
The
consequences
of
having
visible emergency
response
authorities
in
control
is
that
it
enables
the
public
to
define
the
disaster
as
an
event
which
can
be
managed
to
an
acceptable
outcome.
This
is
a
function
of
the
fact
that
the
public
sees
familiar
authorities,
performing
their
expected
role
as
emergency
responders,
who
communicate
disaster
relevant information
via
traditional
emergency communication
channels.
In
contrast,
from
the
public's
point
of
view,
there
was
considerable
question
throughout
the
accident regarding exactly
which
agency
was
fully
in
control
at
THI.
Interestingly,
the
finding
that
most
people
first
heard
about
the
accident
from
the
mass
media
foreshadowed
the
subsequent
reliance
on
the
mass
media
as
a
comnminication
channel
to
the
public
by
virtually
all
parties.
As
Chenault
and
his
colleagues
(1979:124)
note:
There
is
little
to
suggest,
however,
.that
the
Public
Information
Office
position
was
an
especially
prominent
one
in the
activities
of
any
county
(Emergency
Management
Office].
The
media-contact
aspect
of
the
public
information
task
was
taken
up
by
the
Governor's
and
Lieutenant Governor's
offices, by
the
Public
Information Officer
of
PEMA,
and
by
County Coimissioners
and
the
County Coordinators.
.......

39
Of
course,
many
factors
influenced
public
perceptions
of
the
emergency response
efforts
at
TMI,
including
high
visibility
of
political
figures
coupled
with
lower
relative
visibility
of
traditional
emergency response
personnel,
and
real
conflict
among
responder
agencies. The
use
of
the
media,
however,
as
a
main
comunication
channel
to
the
public
probably exacerbated
(and
no
doubt
sometimes
exaggerated)
problems
of
control.
When
the
mass
media were chosen
by
political officials
as
a
comunications
channel,
they
were
in
effect
spotlighted
for
the
public
as
a
source
of
information.
The
problem
which
arises
here
is
that,
when emergency
response
information
is
involved, mass
media
are
a
communication
channel
with
a
considerable
amount
of
"built-in
noise". That
is,
in
the
context
of
conveying
the
official
message,
the
media
can
be
expected
to
coment
on
it
editorially.
That
is,
the
media
serve
both
a
"notification
function"
and
pass
on
information,
but
the
media
also serve
a
"journalistic function"
vis
a
vis
the
public.
When
officials attempt
to
coimanicate disaster
relevant
information
largely
through
press
conferences,
they
appear
(to the
public)
to be
officially sanctioning
the
mass
media.
It
mst
be
remembered
that
while
the
media
do
disseminate
the
official
message,
they are
also
likely
(even
obligated)
to
.run
a
variety
of
related
stories
at the same
time.
Such
related
stories
may
or
may
not
be
consistent
with
the
official
message
and
may
or
may
not
be
technically correct.
The
impact
of
these
circumstances
is
that
the
public
is
confronted with
many
messages,
possibly
conflicting,
all
presumably
from
knowledgeable
sources. The
public
does
not get,
in
77i

40
straightforward
form,
the
official
message
which
is
presumably
based
upon
the
authorities'
plan
for
an integrated
response
to
the
emergency.
In
effect,
the
public
is
confronted with
many
spokesmen
who
have
many
messages
and
it
is
difficult
for
citizens
to
determine
just
what response
is
desired
of
them.
Source
Credibility
Use
of
the
mass
media
as
a
primary
channel
for
communicating
emergency response information
to
the
public
also
has
an
effect upon
the
perceived
credibility
and
usefulness
of
all
information
sources.
Table
4
shows
citizen evaluations
of
the
usefulness
of
the
information disseminated
from
eleven
sources.
Respondents were
asked
to
classify
the
utility
of
each source
into
four
categories:
"extremely
useful
or
useful;"
"of
some
use;"
"totally
useless;"
or
"don't
know
about
the
source."
The ratings
given
these sources
group
them
into
four
categories.
Local
television
and
radio
were
rated
highest,
with
e
67.0
per
cent
of
the
respondents
rating
each
source
as
extremely
useful
or
useful.
The
second
highest utility
ratings went
to the
Governor's
office,
network
television, newspapers,
and
the
NRC.
The
Governor's
office
and
the
NRC
were
rated
as
extremely
useful
or
useful
by
57.0
percent
of
the
respondents,
network
television was
given
this
rating
by 55.0
percent
and
newspapers
by
50.0
percent.
It
should
be
noted
that
for
all
six
of
these
sources, most citizen
ratings
are
at
least
as
high
as
"of some
use;"
very
few
people
rated
any
of
these
sources
as
useless.
Citizen
judgements
were
predominately positive
for
each
of
these
sources.
-- - - -

41
TABLE
4
UTILITY
OF
INFORMATION
FROM
SOURCES:
THREE
MILE ISLAND*
Percent
of
Respondents
Answering:
Extremely
Of
Useful
or
Some
Totally
Don't
Source Useful
Use
Useless Know
Governor's
Office
57.0
27.0
13.0 4.0
Nuclear
Regulatory
Commission
57.0 25.0
11.0
8.0
State
Emergency
Agencies
40.0
27.0 22.0
11.0
Local
Government
36.0
27.0
27.0
11.0
Metropolitan
Edison
11.0 18.0
60.0
11.0
Newspapers
50.0 31.0
14.0
6.0
Local
Television
67.0 20.0
9.0 6.0
Radio
67.0 20.0
7.0
7.0
Friends
30.0
27.0
38.0
5.0
Relatives
30.0
21.0
40.0
8.0
4
Network Television
55.0
25.0
15.0
5.0
*Adapted
from
Flynn
(1979:23-26).
' ,
*1

42
There
is
considerably
more variance
in
ratings
for the
two
lowest rated
groups.
State
emergency
agencies
and
local
government
were
rated
as
extremely
useful
or
useful
by
40.0
and
36.0
percent
of
the
respondents,
respectively.
This
indicates
that
the
public
rated
these
sources
as
moderately
useful. One must
balance
this
positive
judgement,
however,
by
acknowledging
that
33.0
percent
(emergency
agencies)
and
38.0
percent
(local
government)
of
the
respondents
rated these
sources
as
either totally useless
or
"don't
know
the
agency".
Hence, roughly
equal
proportions
of
citizens
saw
these
sources
as
being
on
opposite
ends
of the
utility
scale.
The
lowest
rated
grouping
is
composed
of
friends,
neighbors
and
relatives.
An inspection
of the
row
percentages
in
Table
4
shows
that
the
modal
rating
for
these
sources
is
the
category
"totally
useless".
While
in
each
case
about
30.0
percent
of
the
respondents
saw these
sources
as
useful,
there
is a
definite skew
in
the
direction
of
being
perceived
as
of less
use
than
the
other
sources.
This
is
not
a
particularly
surprizing
finding
for
two
reasons:
(1)
the
highly
technical
nature
of
the
emergency
was
such
that
one
would
not
expect most citizens
to
have
special information;
and
(2)
friends
and
relatives were
not
useful
in
suggesting
new
interpretations
or
providing
new
information because
the
mass
media
was
already
doing
so
on
a
frequent
basis
and
very
thoroughly.
Finally,
the
lowest
rating
was
given
to
Metropolitan Edison.
This
source was rated
as
totally useless
by
60.0
percent
of
the
respondents. An additional
11.0
percent
of
the
respondents
claimed
not
to
have
enough
information
to
even
rate
the
utility.
7- Z

43
To
summarize
this
discussion,
the
mass
media
are
rated
as
the
most
useful sources
of
information.
Local
television
and
radio
received
the
highest
ratings,
followed by
network
television,
newspapers,
the
Governor
and
NRC.
Substantially below
this
high
grouping
of
six sources,
we
find
the
state
and local
authorities.
Social
network
contacts
(i.e.,
friends
and
relatives)
rated
lowest
among
the
four
groupings
of
sources,
and
Metropolitan
Edison
received
by
far the
lowest
utility rating.
Given
the
consistently
high
ratings
assigned
the
mass
media,
it
is
difficult
to
judge
the
relative perceived usefulness
of
the
other
sources.
Table
5
shows
citizens' selections
of
a
single most
reliable
source from
a
list
which
did
not
include
mass
media.
When
asked
to
chose
among
non-media
sources, 58.0
percent
of
the
respondents
selected
the
NRC
spokesman,
Mr.
Harold Denton,
as
the
most reliable
source
of
information.
This
rating
sets
the
NRC
clearly
apart from
all
other
sources
which
received
negligible
endorsements
except
for
Governor Thornburgh,
who was
seen
as
most
reliable
by 19.0
percent
of
the
respondents.
it
is
interesting
that
when
given
this list
of
sources, 9.0
percent
of
the
people
answered
4
that
there
was
no
source
of
reliable information.
Respondents
also
rated
local
emergency
response
authorities
very
low
as
reliable
sources,
putting them
in
essentially
the
.same
category
as
Metropolitan
Edison
and
friends/neighbors.
Table
6
shows
the
most reliable
source
chosen
by
citizens
involved
in
the
two
nonnuclear
disasters.
In
both
types
of
disaster,
the
source
most frequently
selected
as
having
greatest
73
772

44
TABLE
5
MOST
RELIABLE
SOURCE:
THREE
MILE
ISLAND*
Source
N Z
NRC
(Harold Denton)
207
58.0
Governor
Thornburgh
69
19.0
Friends/Neighbors
8 2.0
Local
Officials
7 2.0
Metropolitan Edison
6 2.0
No
Reliable
Information
31
9.0
Other
29
8.0
No
Answer
2 1.0 I
*Adapted
from
Barnes
et
al.
(1979:14).
Z-

45
TABLE
6
MOST
RELIABLE
SOURCE:
NONNUCLEAR DISASTERS
Volcano
Flood
Source
N %N %
Neighbor, Friend,
Relative
7
7.8
129
26.9
Local
Emergency Response
Authorities
33
36.7 266
55.4
State/County Emergency
Authorities
1
1.1
18
3.8
Federal
Authorities
11
12.2
0 0.0
Mass
Media
20
22.2
38
7.9
Personal Judgement
18
20.0
29
6.0

46
reliability
is
local
emergency
response
authorities.
Indeed,
more
than
one-third
of
those
in
the
volcanic eruption
and
more
than
one-half
of
the
flood
victims
placed
their
highest confidence
in
local
authorities.
For
the
lower
reliability
ratings
there
is a
slightly different
pattern
between
volcanoes
and
floods.
For
the
volcanic
eruption,
after
local
authorities,
people
listed
most
reliable
sources
(in
order
of
descending confidence)
4
as:
mass
media
(22.2
percent), personal
judgement
(20.0
percent),
Federal authorities
(12.2
percent),
and social
networks
(7.8
percent). This
particular
pattern
of
public
confidence
in
different
sources
is
probably
best
understood
in
terms of
citizen perception
of
who
controlled
current
and
accurate information
about
the
volcano.
In these
data,
mass
media
refers
largely
to
local
radio.
The
fact
that
this
source
received
the
second
highest
confidence
rating
is a
function
of
two
circumstances:
numerous volcano
status
bulletins
were
issued on
the
radio
daily,
and
the
emergency
plan
disseminated
to
the
public
by
the
County
Sheriff's office
urged
citizens
to
monitor
radio
broadcasts.
Under
these
circumstances,
radio
was
seen
by
the
public
as
having
a
defined
role
in
an
eruption
response
and
could
be
perceived
as
an
extension
of
local
authorities.
Personal
judgement
is
rated
as
the
third
most reliable
source.
This degree
of
confidence
in
one's
own
judgement
reflects
the
fact that
volcanic
eruptions were
a
very
unfamiliar
hazard;
Mt.
St.
Helena
had
been
dormant
for
123
years.
Respondents
argued
that
the
decision
to leave
their
homes
was
a
personal
one,
which
they

47
felt
had
to
be
based
somewhat
on
their own
interpretation
of
the
risk
information
given
them
by
authorities
(cf.
Perry
et
al.,
1980a:21).
In
the
case
of
flood
victims, most
of
whom
placed
highest
confidence
in
local
authorities,
the
second most reliable
source
cited
was
social
networks.
Friends,
neighbors,
or
relatives were
chosen
as
the
most reliable source
by 26.9
percent
of
the
respondents.
As
Table
6
shows,
for
floods social
networks
or
local
authorities account
for
virtually
all
of
the
respondents;
a
few
selected
state
or
county authorities, mass
media,
or
personal
judgement
as
most
reliable, but
these
proportions
are
very
small.
The
relatively
high
levels
of
confidence
in
social
network
contacts
is
related
to
citizens perceived importance
of
past
experience
as
a
basis
for
responding effectively
to
floods.
In
the
United
States,
floods
are
the
most
widespread
geophysical hazard
(White,
1975), and
consequently
many
citizens have been exposed
to
this
hazard
at one
time
or
another.
Hence,
many
private citizens,
particularly
those
who
have
lived
in
an
area
for
some
time,
can
claim
to
have
special
knowledge
of
flood
patterns.
Many
times,
this
type
of
information
is
passed
around
social
networks
in
the
form
of
advice about
the
threatening
flood
based
upon
a
person's
knowledge
of
previous
floods.
1
Frequently
flood
coping information
acquired
in
this
fashion
is
useful
and
information
recipients develop
confidence
in
the
source.
llnterestingly,
from
the
standpoint
of
coping with
a
given
flood
situation, knowledge
of
previous
floods
is
not always
a
technically
accurate predictor
of
what
is
to
come.
(See
for
example,
Perry
et al.,
1980b).

48
Summary
In
these
comparisons
of
public
confidence
in
information
sources,
the
important
finding
is
that
at
TMI
the
public perceived
the
mass
media
as
the
most reliable
source,
while
in
the
nonnuclear
disasters
the
public
placed
highest
confidence
in
local
emergency
response authorities.
Furthermore,
in
both
types
of
natural
disasters,
the
proportions
of
people who chose
local
authorities
as
most reliable
was
considerably higher
than
the
proportions choosing
any
other
source.
Local
authorites
were clearly
the
preferred
reliable
source.
With
respect
to
the
nuclear accident
at
TMI,
it is
possible
to
explain
the
observed pattern
of
public
confidence
in
different
sources
by
carefully examining
events
during
the
emergency
period.
Several
factors
are
important
in
the
high
levels
of
confidence
ascribed
to
the
mass
media.
First,
it
was
via
the
mass
media
that
most people
initially
heard
about
the
accident,
and
virtually
all
parties involved,
particularly
political
officials, continued
throughout
the
emergency
to
communicate with
the
public via mass
media. Hence,
with
this
official
sanctioning,
the
public attended
to
the
media
and
came
to
expect
emergency
information
from
this
source. Second,
TMI
presented
citizens
with
a
threatening event
that
they
had
not
previously experienced, which
was
complex
and
not
easy
to
understand,
and
about
which
there
was
not
a
great
deal
of
information available.
In
such
situations,
when
no
other
source
dominates
the
scene,
the
mass
media
are
attractive
to
the
public
because
they
make
available
a
variety
of
information
from
different
______
____________ ~ ~l

49
sources,
all
presumably
with
some special
expertise. Third,
the
mass
media, particularly
radio
and
television,
are
available
to
the
public
on
an
almost
continuous
basis
and
therefore
are
presumed
to
have
very
current information.
Finally,
after
the
emergency
when
citizens
try
retrospectively
to
decide which
source
provided what
turned
out
to
best
fit
what
happened,
mass
media
have
an
advantage.
This
is
because
the
media
run
many
stories
and
accounts
of
the
incident,
and
thereby
have
a
greater likelihood
of
being
correct
just
by
chance.
(If
enough predictions
are
made,
one
of
them
is
apt
to
be
right.)
Although
this
requires
some
selective
recall
on
the
part of
citizens,
it
is
not
an
unheard
of
phenomenon.
The
problem
of
Metropolitan
Edison's
very
low
public
confidence
rating
as
a
source
of
information
is
interesting,
especially
since
the
utility
was
probably
the
single
source with
the
most
technical
expertise
and
special
knowledge
of
the
continuing
status
of
the
reactor.
Probably
the
major
contributor
to
the
low
confidence
j
rating
was
the
press
conference
held
at
4:30
p.m.
on
the
first
day
of
the
accident
in
which
Lt.
Governor
William
Scranton disassociated
his
office
with
the
utility
and stated
that
the
utility
was
disseminating conflicting
and
misleading
infotiation
about
the
accident
(Kemeny,
1979:109). This public
rebuke
of
Metropolitan
Edison
was
undertaken
by
the
State
because
it
had
evidence
that
Metropolitan
Edison officials were
misrepresenting
the
condition
of
the
reactor
and
the
resultant
risks
to
the
general public
(Martin,
1980:107-108).
As
the
President's Commission
concludes: "Met
Ed's
handling
of
information during
the
first
three
days
of
the
accident
resulted
in
loss of
its
credibility
. . "
(Kemeny,
1979:57).
• ." ;6 : - - .. .T
'.."
-
-

50
Finally,
we
can
address
the
question
of
why
local
emergency
authorities
at
TMI
were perceived
as
highly
reliable
by
such
a
relatively
small
proportion
of
the
public.
Three general
circumstances
seem
to
contribute
to
this
perception. First,
apparently
due
to
the
active involvement
of so
many
agencies
and
particularly
political
officials,
local
emergency response personnel
played
a
relatively
less
visible
public
information
role
in
the
emergency. They
were
infrequently represented
at
press
conferences
and
apparently
appeared,
at
least
to
the
public,
to
be
performing
support
functions rather
than
a
primary
management
function.
Unfortunately, much
of
the
massive
planning
efforts
for
5,
10,
and
20
mile
evacuations
by
the
counties were
"invisible"
to
the
general
public.
Second,
there
were
very
few
direct
communications
between
local
authorities
and
the
public.
Counties
did
maintain
rumor
control centers
and
some
distributed
evacuation
information
to
risk
area residents; and
if
an
evacuation had
been
ordered,
provisions
were
made
by locals
for
dissemination
of the
order
and
monitoring
the
exodus.
As
it
was,
however,
there
were
no
provisions
by
political
leaders
of
federal
agencies
to
routinely
channel
accident
information
intended
for
the
public through
local
emergency
personnel.
Third,
and
largely
because
of
the
above
described
communication
patterns,
the
public
did nqt
see
local
authorities
as
possessing
any
special
access
to
technical
information about
the
TMI
event.
Finally, based upon experience
in
managing
natural
disasters,
it
is
likely
that
by not
assigning
local
emergency
7r
4:N-n

51
officials
a
more visible
role,
political
authorities
inadvertantly
limited their
credibility
and
contributed
to
the
public
perception
that
the
accident
was
being
poorly
handled.
LI
II

52
CHAPTER
FIVE
EVACUATION DECISION-MAKING
After examining
citizens'
sources
of
information
about
the
threats
and
their
beliefs
about
the
utility
of
these
sources,
it
is
important
to
consider
how
people
acted
upon
the
information available
to
them. Here
we
are
concerned with
one
type
of
action:
evacuation
or
relocation
to
an
ostensibly
safer
place.
This
chapter reviews citizen answers
to
some
general
inquiries about
why
they
did
or
did not
evacuate
in
response
to
each
of
the
three
threats.
In this
way,
one
can
gain
perspective
on
the
way citizens
evaluated
the
threat through
examining
their
beliefs
about
what made them
act.
This
chapter
is
structured around
three topics:
reasons
given
for
evacuating; reasons
given
for
not
leaving;
and
a
discussion
of
the
overall
evacuation response.
Reasons
for
Evacuating
Table
7
shows
reasons
given
for
evacuating
by people
who
left their
homes
in
response
to
the
reactor
accident
at
TMI.
In this
case,
respondents were
read
a
list
of
possible
reasons
for
leaving
and
asked,
for
each reason,
whether
it
was
important
in
their
decision
to
evacuate.
These
data
show
that
people's
perception
of
danger
by
far
dominated
the
reasons
given
for
leaving.
Situational danger
was
cited
by 91.0
percent
of
the
respondents
as
an
important
factor
in
the
evacuation
decision;
this
perception
of
danger
is
probably
also
a
concern
for the
61.0
percent
who
mentioned
a
need
to
protect
children
and
the
8.0
percent who
cited

53
TABLE
7
REASONS
FOR
EVACUATING:
THREE
MILE
ISLAND*
Reason
Percent
Situation
seemed
dangerous
91.0
Information
on
situation
was
confusing
83.0
To
protect
children
61.0
To
protect
pregnancy
8.0
To avoid
the
confusion
or
danger
of
a
forced
evacuation
76.0
Pressure
from
someone
outside
family
(friend/neighbor)
28.0
Trip
planned before
incident
5.0
*Adapted
from
Flynn
(1979:18).
p )

54
concerns
about
pregnancy.
Confusing
information
about
the
threat
was
cited
as
a
reason
for
leaving by 83.0
percent
of
the
respondents.
This
confusion
on
the
part
of
the
public
was
no doubt
related
to
the fact
that
different
groups
of
presumed experts
were
disagreeing
about
the
dangers
involved
and
even
the
basic condition
of
the
reactor.
When
the
public
lacks
the
technical
skills
to
evaluate
the
disaster
itself, and those
who
have
the
technical
skills
disagree,
it
tends
to
create
relatively
high
levels
of
anticipatory
fear,
causing
people
to
try
to
minimize
their
total
potential
losses.
In
this
case,
evacuation
was
seen
as
a
prime
path
to
minimization.
The
third
reason
for
evacuating
cited by
a
substantial
proportion
(76.0
percent)
of
people
was
to
avoid
the
confusion associated
with
a
forced
evacuation.
In
this
case,
people were
endorsing
the
belief
that
the
situation
was
getting
worse--it
was
only
a
matter
of
time till
everyone would
be
told
to
go--and
it
seemed
best
to
get
"a
jump
on
the
situation"
by
leaving
before
exit routes
became
congested.
Interestingly, relatively
fewer
peonle (28.0
percent)
said
that
pressure
from
social
network
contacts
contributed
to
their
decision
to
evacuate.
Overall,
the
major
concerns
cited
by
evacuees
focused
upon
the
danger
involved,
the
difficulty
associated
with
obtaining
clear,
accurate
information
about
the
threat,
and
the
likely
problems
of
forced
evacuation.
To
narrow
this
field
down
and
assess
the
relative
importance
of
different
reasons,
one can
examine
the
single
most important
reason
for
leaving.
Table
8
shows
respondents' choice
of
a
single,
critical
piece
of
information
used
in
deciding
to
evacuate. These
data show
that
concerns
about
situational danger
were
indeed
paramount
in
the
decision
___________________i

55
TABLE
8
CRITICAL INFORMATION
IN
DECISION
TO
EVACUATE:
THREE
MILE
ISLAND*
Information Percent
Hydrogen
Bubble
30.0
Conflicting
Reports
19.0
Governor's
Advice
to
Leave
14.0
Threat
of
Forced
Evacuation
14.0
News
Bulletins
9.0
Urging
of
Relative
6.0
No
Single
Reasons
25.0
*Adapted
from
Flynn
(1979:22).
-
~
_
_ _ _ _ _

56
to
evacuate:
the
largest
proportion
of
evacuees
(30.0
percent)
said
that
the
danger
implied
by
the
formation
of
a
hydrogen
bubble
in
the
reactor
was
critical
in
their
decision
to
leave.
This
factor
is
followed
in
relative
importance
by
conflicting
reports about
the
threat
(19.0
percent),
the
Governor's
evacuation advisory
(14.0
percent),
and
concerns
with
a
forced
evacuation
(14.0
percent). News
bulletins
from
the
mass
media
and
urging from
social
network
contacts were
least
frequently
chosen
as
a
most
important
reason
for
evacuating.
Finally,
it
should
be
noted
that
for
many citizens
the
presence
of
danger,
conflicting
information,
an
evacuation
advisory,
threat
of
forced
evacuation,
and
other events had
an
additive
effect:
25.0
percent
of
the
respondents
said
it
was
a
combination
of
factors
rather
than
a
single
piece
of
information which
was
critical.
When
the
question
of the
most
important
reason
for
evacuating
in
nonnuclear
disasters
is
considered,
one
sees
a
slightly
different
pattern. Table
9
shows
volcano
and
flood
victims'
choices
for
the
critical
factor
in
the
decision
to
leave.
For
both volcanoes
and
floods,
the
two
reasons
cited
by
the
largest
proportions
of
respondents
as
most
important
are
(1)
seeing
evidence
of
the
threat, and
(2)
being advised
by
officials
to
leave.
Being
able
to
see
physical
evidence
of
a
threat
in
effect clarifies
many
questions
a
citizen
may
have
about his
susceptability.
Indeed,
when
one can
experience
first
hand
such
environmental
cues,
part
of
the
problem
of
evaluating personal risk
is
transferred
from technical
experts
to
the
citizen.
He
feels
able
to
look
at the
situation
and
make
a
personal
judgement about
whether
the
threat
is
likely
to
affect
him
or
his
family
and
decide
what protective action

57
TABLE
9
MOST
IMPORTANT
REASON
FOR
EVACUATING:
NONNUCLEAR
DISASTERS
Volcano Flood
Reason
for
Evacuating
N % N X
Neighbors/Relatives
left
12
15.2
44
13.7
Media warnings
5
6.3
5 1.6
Officials
urged
departure
21
26.6
93
29.0
Relatives
urged
departure
16
20.3
28
8.7
Past
experience
2
2.5
21
6.5
Saw
eruption/high water
23
29.1
130
40.5
____ ___ ____
___ ____ ____ *.77-

58
seems
warranted.
Research
on
natural
disaster
has
shown
that
visibility
of
a
threat
is
positively correlated
with
undertaking
protective
actions
(Perry
et
al.,
1980a;
Perry
et
al.,
1980b).
Gruntfest
et
al.
(1978)
have
documented,
for
example,
that
flash
flood
warnings
issued
in
the
absence
of
any
visably
threatening
environmental
conditions sometimes
go
completely
unheeded.
In
effect,
seeing
environmental
cues
allows
people
to
quickly
arrive
at
a
definition
of
the
situation
as
dangerous
and
requiring
special
attention.
Thus,
if
we
group
(as
it
seems
reasonable
4
to
do)
"seeing
the
threat"
with belief
that the
situation
is
dangerous,
one
sees that
perceived danger
was
cited
as
the
most important
reason
for
evacuating
by those
involved
in
both
nuclear
and
nonnuclear
disaster
events.
The
second
most
frequently
cited
reason
for
evacuation, again
in
both
the
volcano
and
flood
data,
was
that
the
respondent
was
urged
by
officials
to
depart.
These
data
reflect
citizen confidence
in
officials
as
(1)
having
access
to
special
hazard-relevant
information,
and
(2)
assuming responsibility
for
managing
the
emergency response
efforts
which
involve
the
public.
Under
these
conditions,
citizens
can
define
emergency
officials
as
important sources
whose advice
constitutes
information
which
should be
acted
upon. Although
the
proportions
of
respondents citing official advice
as
a
reason
for
leaving
are
higher
in
the
natural disasters,
the
official
advisory
from
the
Governor
was
the
third
most prominent
reason
for
evacuating
given
in
the
THI
data.
For
the
natural disasters,
the
next most frequently cited reasons
for
evacuating
relate
to
social
network
contacts:
either
the
respondent
witnessed
neighbors
and
relatives
evacuating
or
he
was
urged
by
relatives

59
to
depart.
In
both
natural disasters,
media warnings
and
past
experience
were infrequently given
by
respondents
as
most
important reasons
for
leaving.
In
summary, situational
danger
and
advisories
from
officials
were
cited most frequently
as
critical
reasons
for
evacuating
in
both
the
nuclear
and
nonnuclear
incidents. Indeed,
these
two
reasons
alone
account
for
more
than
55
percent
of
the
volcano
evacuees,
69
percent
of
the
flood
evacuees, and
nearly
45
percent
of
the
TMI
evacuees.
Also,
media
warnings
were infrequently chosen
as
the
most
important reason
for
evacuating
in
all
three
hazards.
It
was
found,
however,
that
social
network
contacts
were relatively more
important
to
evacuation
decision-making
in
the
natural
disasters
than
at
Three Mile
Island.
Reasons
for
Not
Evacuating
Having
reviewed
reasons given
as
important
in
deciding
to
evacuate,
we
can
now
turn
to
another
perspective
and
examine
the
reasons
given
for
staying
by
people who
chose not
to
evacuate. Table
10
shows
the
proportion
of
respondents
at
TMI
endorsing
each of
twelve
reasons
for
not
leaving;
note
that
respondents were
allowed
to
select
more
than
one
reason
for
not
evacuating.
A
fairly
substantial
number
of
respondents
(62.0
percent)
said
that
one
reason
they
didn't
evacuate
was
that
they
were
not
ordered
to
do
so.
Presumably,
tfany
of those
who
did
not
evacuate
were
waiting
for
an
unambiguous
directive
from
an
authority.
This
is
no
doubt related
to
the
fact that
42.0
percent
of
the
respondents
also
said
chat
the
many
conflicting
reports about
the
threat
were
relevant
to
their
decision
to
stay.
--- . --
_ _ _memo_

60
TABLE
10
REASONS
FOR
NOT
EVACUATING:
THREE
MILE
ISLAND*
Percent
of
Reason
for
not
Evacuating
Respondents
Endorsing
Not
ordered
to
evacuate
62.0
Too
many conflicting
reports
42.0
No
real
danger existed
38.0
Home
safe
distance
away
31.0
Fear
of
looting
24.0
No
children
involved
23.0
Could
not
leave
job
21.0
Neighbors
did
not
evacuate
16.0
Must
care
for
farm
6.0
No
place
to
go
5.0
Too
old
to
leave
3.0
Handicapped
2.0
*Adapted
from
Zeigler
et
al.
(1981:6).

61
The
data
in
Table
10
suggest,
however,
that
the
most
pervasive
reason
for not
evacuating
was
the
belief
that no
real
danger
existed.
No fewer
than
five
of
the
most
frequently
endorsed
reasons make
a
reference
to
low
levels
of
perceived
danger.
These
are:
no
real
danger
existed,
38.0
percent;
my
home
is a
safe
distance
away,
31.0 percent;
no
children
were
involved
(implying
no
danger
to
adults),
23.0
percent;
and
my
neighbors
didn't
leave
(thereby also
believing
the
danger
to
be
low),
16.0
percent).i
Finally, although
not
prominent,
fear
of
looting
was
cited
by
24.0
percent
of
the
respondents
as
a
factor
in
choosing
not
to
evacuate.
Various
logistical
difficulties--job
responsibilities,
farm
care
responsibilities,
no
place
to
go,
age,
handicap--were
also
endorsed
by
a
few
respondents.
Table
11
shows
the
most
important single
reason
given
for
not
evacuating
by
volcano
and
flood
victims.
It
will
be
noticed
immediately
that
virtually
all
respondents
in
the
volcanic eruption
evacuated:
only
10
people chose
to
stay
in
their
homes. Thus,
some
care
is
required
in
interpreting
these data.
Confidence
in
the
volcano
data
is
enhanced,
however,
by
the
fact
that
the
relative
ranking of
reasons
for not
evacuating
matches
the
rankings
in
the
flood
data.
As
we found
in
the
TMI data,
the
most prominent reason
for
not
evacuating
was
the
belief
that no
real
danger
existed.
This
reason
accounts
for
70.0
percent
of
the Mt.
St.
Helens
nonevacuees
and
67.5
percent
of those
who didn't
evacuate
in
response
to
floods.
The second
most frequently
cited
reason
for
staying
which
together
with
no
danger accounts
for
all
of
the
volcano
nonevacuees
and
nearly
80
percent
of
the
flood
nonevacuees
is
"stayed
to
--w .-
-
~--~--~
-
_ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _
_ _ _ _ _ _
_ _ _

62
TABLE
11
MOST
IMPORTANT
REASON
FOR
NOT
EVACUATING:
NONNUCLEAR
DISASTERS
Volcano
Flood
Reason
for
Not
Evacuating
N % N Z
No
evacuation
order
0
0.0
3 1.8
Did
not
believe
real
danger
existed
7
70.0
114
67.5
Feared looting
0
0.0
3 1.8
Stayed
to
help
others
0
0.0
13
7.7
Family
not together
0
0.0
3 1.8
Stayed
to
protect
house
3
30.0
18
10.7
High
water blocked
exit
0
0.0
1 .6
Survived
other
floods
unharmed
0
0.0
14
8.3 j
I

63
protect
house".
In
this
case,
reference
is
made
to
protecting
the
house
from
the
environmental
threat, not
from
a
threat
due
to
looters. The
problem
of
looting
is
generally
rare
in
natural disasters
(Quarantelli
and Dynes, 1970:168;
Dynes
et
al.,
1972:33),
a d
was not
perceived
as
a
reason
not
to
evacuate
in
the
data
at
hand.
In
summary,
one
should
remember
that
the
reasons
discussed
above
are
those
given only
by
people who
chose
not
to
evacuate.
In
their
decision-making
calculus,
these factors
were sufficient
to
make
them
believe
that
leaving
was
unnecessary.
For
both
TMI
and
the
natural
disasters,
most
of
those
who
didn't
evacuate
chose
not
to
because
they
did
not
believe
that real
danger
existed. Among nonevacuees
at
TMI,
the
presence
of
conflicting messages
and
the
absence
of an
official
evacuation
order
were frequently
cited
reasons
for
staying.
In
the
natural disasters
people
also
reported
that
they
chose
to
stay
so
that
they
could
protect
their
homes
from
the
environmental
threat.
Unlike
the
natural disasters,
fear
of
looting
was
given
as
a
reason
for
not
evacuating
at
TMI.
The
Overall
Evacuation
Response
After
reviewing
reasons
given
by evacuees
for
leaving
and
by'
nonevacuees
foc
staying,
to
gain
perspective
on
the
process
of
evacuation
one
can
consider
the
overall public
response.
In
general,
particularly
in
natural
disasters, getting
people
to
evacuate
is a
difficult problem.
Many
people refuse
to
leave
even
when
ordered
to
do
so
(Quarantelli
and
Dynes,
1972;
Quarantelli,
1981:15-20;
Quarantelli
and
Taylor,
1977;
Quarantelli
and
Dynes,
1977). In
the
volcanic
eruption studied here,

-r
64
11.1
percent
of
the
citizens
at
risk
failed
to
evacuate.
For
natural
disasters
this
is a
low
proportion
of
nonevacuees
and
has
been
explained
in
terms
of
the
uniqueness
of
the
disasters
and
the
high
levels
of
coumunity emergency
preparedness
in
the
affected communities.
The
more
commonly
seen
figure
is
that
for the
flood
communities where
48.6
percent
of those
who
received
a
warning
failed
to
evacuate.
At
TMI,
where
only
an
evacuation
advisory
for
pregnant
women
and
young children
was
issued,
it is
estimated
that
144,000
people,.
39.0
percent
of
the
total
population
within
15
miles
of
the
reactor, evacuated. This
relatively
high
proportion
of
evacuees contrasts
with
the
general
situation
in
natural disasters
and
requires
that
one assess
the
probable
reasons
for
this
response
at
TMI.
The
answer
to
the
question
of
why
so
many
people
evacuated
at
TMI
lies
in
an
examination
of
two
general
categories
of
reasons:
(1)
largely
circumstantial
factors
related
to
the
way
in
which
the
emergency
was
managed;
and
(2)
factors
related
to
the
public's
perception
of
the
risks
involved
in
nuclear
accidents.
With regard
to
managing
the
emergency,
the
situation
at
TMI
was
characterized
by three
elements. First,
the
public was
faced
with
an
unfamiliar
risk
which
was
difficult
to
understand.
In
the
entire
history
li
of
the
United
States
nuclear program,
prior
to
TMI
only
three times
have
there
been
equally
serious
reactor
accidents
and
none
of
these involved
radiation
releases
off-site (Donnelly
and
Kramer,
1979:3).
Thus not
only
the
public, but
emergency management
officials
too,
were
not
attuned
to
the
problems of response
to
this
kind
of risk.
Second,
especially
for
,he
first
three
days
of
the
emergency
period,
there
appeared
to
the
4

T
65
public
to
be
confusion
among officials
and
many
contradictory messages
about
the
accident--its
seriousness
and
the
risks
to
the
public--were
disseminated.
Therefore,
the
public
was
facing
an
unfamiliar hazard
regarding
which
there
were many
conflicting assessments
of
danger.
Third, however,
there
was one thing
upon which
most
experts
did
seem
to
agree
(which
also made
intuitive
sense
to
the
public):
safety
was
correlated
with
increasing distance
from
TMI.
Furthermore,
this
distance
idea
came
up
in
the
form
of
public
discussions
of
evacuation
by
officials
and
experts
a
number
of
times
during
the
emergency
period.
On
the
morning
of
the
second
day
(Thursday),
a
physician
being
interviewed
on
Harrisburg
radio
recommended evacuation
(Martin,
1980:125).
Friday
morning
the
Emergency Management Director
of
Dauphin
County
warned
that
an
evacuation
may
be
needed
very
soon;
he
also
described
things people
should
take
with
them
and
where
they
should
go
(Martin,
1980:144).
Although
this
evacuation "advisory"
was
not
made
"official",
that
afternoon
Governor
Thornburgh
did
advise
that
pregnant
women
and
small
children
evacuate. Also,
in
Dauphin, York, Lebanon,
and
Perry
counties
information
packets
(or
instruction
sheets) on
evacuating--what
to
take,
how
to
leave,
where
to
go--were
prepared
and
distributed directly
to
the
general
population (Chenault,
1979:124-129).
To
summarize,
people
were
confronted
with
an
unfamiliar
risk,
regarding
which
it
was
difficult
to
get
information, but
were
told
that
evacuation
was
a
definite
path
to
safety.
Put
this
way,
it
is
less
difficult
to
understand
why
a
person
seeking
to
minimize
potential
negative
consequences
would
evacuate.
While
it
is
correct
that
an
evacuation
was
never
officially
ordered,
evacuation
was
sanctioned
by experts
as
a
protective action.
Indeed,
-, - - -
.4
* .. ....... .. *o- -.. t,- .. - . -.4 - L -
•
I
"
I
I-I

66
while
experts
argued about
whether
the
situation
was
so
serious
that
people
should
evacuate,
they
agreed
that
evacuation would
substantially
reduce
the
danger.
One
would
expect
that
these
circumstances
would
encourage evacuations
independent
of
the
nature
of
the
hazard
involved,
whether
nuclear
or
nonnuclear.
The second
category
of
reasons
for
evacuation
at
TMI,
the
public
perception
of
the
threat,
depends largely
upon
what have
previously
been
described
as
"unique"
aspects
of
nuclear
threats.
Table
12
shows
citizen perception
of
the
threat
posed
by
TMI
to
the
family
for
three
different
distances
from
the
plant.
These
data
show
the
proportions
of
people who
rated
the
threat
from
TMI
as
very
serious
or,
at
the
opposite
end
of
the
continuum,
as
nonexistant. First,
the
data
indicate
that
distance
did
not
seem
to
have
an
effect
on
perceived
level
of
threat
in
the
case
of TMI:
for
all
three
distances
about
half
of
the
respondents defined
threat
as
very
serious
and
just over
ten
percent
saw
no
threat.
Table
13
shows
level of
perceived
threat
for
the
two
natural
disasters
classified
into
four
categories
of
perceived danger ranging
from
"none"
to
"severe".
A
critical
aspect
of
these
data,
in
comparison
to
the
TMI
data,
is
that
smaller proportions
of
people
saw
the
natural
disasters
as
posing
severe danger
and
larger
proportions endorsed
the
belief
that
the
threat
posed
no
danger.
For
both
the
volcano
and
floods,
nearly
half
of
the
respondents
said
that
the
threat posed
either
no
danger
or
slight
danger.
These
data
document
that
for
some
reason,
citizens
perceived
the
nonnuclear
risks
to
be
relatively
less
threatening
than
the
nuclear
risk.
A __- ..

67
TABLE
12
PERCEIVED
THREAT
TO
FAMILY
DURING
THI
ACCIDENT*
Percent Percent
Very
Serious
No
Distance
from
TMI
Threat Threat
0
to
five
miles
50.0
14.0
five
to
ten
miles
50.0
11.0
ten
to
fifteen miles
47.0
11.0
*Adapted
from
Flynn
(1979:30).
P -- -~ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

68
TABLE
13
PERCEIVED
THREAT
FROM NONNUCLEAR
DISASTERS
Volcano
Flood
Perceived Threat
N % N
Believed
hazard
posed
no danger
28
31.1
99
20.7
Slight
danger
16
17.8
146
30.5
Moderate
danger
10
11.1 143
29.9
Severe
danger
36
40.0
90
18.8
IL
, iV

69
Although
empirical assessments
of
perceived
threat
from
nuclear
disasters
are
virtually
nonexistent,
social
scientists
have
argued
that
citizens
have
a
distinct
view
of
nuclear
hazards
as
constituting
a
special
threat
different
from
other
man-made
and
natural hazards
(Perry
et
al,
19
8
0:c).
This
view
stems
from
public
beliefs
about
the
characteristics
of
radiation
as
hazard.
Of
interest
here
are
two
general
types of
belief
patterns
that
relate
to
the
problem
of
detection
and
the
concept
of
dose.
In
the
case
of
natural
hazards,
such
as
tornadoes,
floods,
or
volcanoes,
people
have
a
sense
of
what
constitutes
danger--wind, water,
mud
flows,
ash, etc.
These
agents
may
not
exactly
be familiar,
but
neither
are
they
completely
outside
the
citizens
realm
of
experience
or
imagination.
Also,
these risks are
spatially
defined
in
the
sense
that
they
are
"visible"
and
finite;
one
can
feel the
wind
or
see
the
water
or
mud.
A
citizen, relying
upon
his
senses--sight,
touch,
hearing,
etc.--can
reliably
detect
the
presence
or
absence
of such
risks
in
the
environment
and,
if
need
be,
generate
some
protective strategy
on his
own,
perhaps by
seeking
high
ground
or
some
special shelter.
Hence,
these
types
of
risks
can
be
perceived by
citizens
as
identifiable,
understandable,
and
as
threats
from
which
it
is
possible
to
protect
oneself. Interestingly,
this
view
of
natural hazards
has
been
cited
as
one
of
the
reasons
that
citizens
are
slower
to
respond
to
disaster
warnings
than
authorities
deem appropriate.
On
the
other hand,
studies show
that
the
public
views
radiation
risks
as
"involuntary,
unknown
to
those
exposed
and
to
science,
uncontrollable,
unfamiliar, potentially
catastrophic, likely
to
be
fatal
rather
than

70
injurous,
and
dreaded"
(Slovic
et
al.,
1980:5).
As
a
hazard,
then,
almost opposite
qualities
or
characteristics
are
attributed
to
radiation
than
are
attributed
to
natural disasters.
Radiation
tends
to
be
viewed
as
an
invisible,
lethal threat
that
radiates
in
all
directions
from
a
source,
against
which
protection
is
difficult
or
impossible
to
achieve.
Hence, radiation
is
an
unfamiliar
danger which
the
citizen
cannot
see,
hear,
smell,
feel,
or
taste
(Grinspoon,
1964:120)
without
special
equipment.
The
idea
that
a
hazard,
perceived
to
be
very
lethal,
is
for
the
most
part
undetectable
distinctly
sets
it
apart
from
other hazards.
With
regard
to
dangers
from
the
not
easily
detected
hazard
of
nuclear
radiation,
most citizens
are
familiar
with
the
rather dire nature
of
the
consequences
of
exposure.
The public
has seen
many
discussions
of
death
from
radiation exposure,
and
studies indicate
that
people
tend
to
associate death--either immediately,
or
within
a
few
weeks
due
to
radiation
sickness,
or
in
years
due
to
cancer--as
a
consequence
of
such
exposure
(cf.
Lifton,
1967:48-52;
Kiyoshi,
1967:93-98;
Slovic
et al.,
1980:8-12).
There
are,
of
course,
many
hazards
in
which
exposure
appears
to
result
in
death.
With
respect
to
radiation, however,
the
concept
of
dose
or
the
extent
of
exposure
is
very
important
in
determining
the
extent
of
negative consequences.
In
fact,
radiation
is
present
in
much
of
the
human
environment;
sensitive detection
instruments
must
even
be
calibrated
so that
background
radiation
levels
are
accounted
for
in
measurevients.
Humans
are
constantly
bombarded
by radiation--it
is
only
when
these
levels
of
exposure become high
that
health
consequences
seem
to
accrue.
The
idea
of dose
does
not
appear
to
be fully
appreciated
by
the
public
in
that
many
people
seem
to
equate
any
level
of
exposure
with
death,
the
most
serious
consequence.
A-1

71
It
is
very
likely
that
people
do
not
appropriately
distinguish
radiation
from
nuclear
power plants
from
radiation associated
with
nuclear
bombs.
Indeed,
when
asked
to
describe
health
consequences
of
radiation exposure,
people
tend
to
mention
symptoms
common
in
exposure
only
to
very
high
doses,
such
as
one
would experience
if
exposed
to
a
nuclear
bomb
explosion.
Parenthetically,
dose
levels
as
well as types
of
radiation
are
considerably
different
for
nuclear
power
plant
accidents
than
for
weapons. The
point
of
this
discussion,
however,
is
that
the
public
in
general
sees
radiation
as
a
difficult
to
detect
threat
which
produces very negative consequences
in
those
exposed. This
sets
it
apart
from
other
disasters,
particularly
natural disasters, both
in
the
way
people
think about
it
and
in
the
way
they
react
to
possible
exposure.
As
our
data
show
the level
of
threat
attributed
to
nuclear
disasters
is
much
higher
than
for
the
nonnuclear disasters.
This
heightened
threat
associated
with
the
nuclear disaster
is
also
no
doubt
related
to
the
frequently
cited "fear
reactions"
to
nuclear disaster (Glass, 1956:630;
Lifton,
1964:152);
that such
fear
characterized
citizens
during
the
TMI
incident
was
documented
(as
"demoralization")
by
the
Report
of
the
Public
Health
and
Safety
Task
Force (Fabrikant,
1979:275)
of
the
Presidents'
Comuission.

72
CHAPTER
SIX
SUMMARY:
IMPLICATIONS
FOR
EVACUATION
PLANNING
This
report
represents
a
first,
tentative
step
toward
developing
a
reliable
body
of
knowledge regarding
the
comparability
of
human
response
to
nuclear
and
nonnuclear
threats.
At
present,
it
is
the
only
study
available
in
the
open
literature
which
reports data
base4 comparisons.
The
analyses presented
in
Chapters Four
and
Five
exclusively
focus
upon
two
issues:
warning
source
credibility
and
evacuation decision-making
as
seen by
the
public.
These
particular
foci
were
chosen
largely because
analyses
were confined
to
published
data
on
the
Three
Mile
island
accident.
Regretably,
no
TMI
data bases
were
available
for
a
thorough
secondary
analysis.
The empirical
comparisons which were possible,
however, have
important implications
for
evacuation
planning procedures
for
nuclear
threats.
The importance
of
these
implications
lies
not
so
much
in
the
finding
itself, but
in
the
extent
to
which
findings
about special
issues
in
natural
disaster evacuations
are
applicable
to
the
nuclear
case. The
following
sections
sunnarize
a
number
of
conclusions which
may
be
drawn
from
the
comparative analyses
in
Chapters Four
and
Five. These
conclusions
are
discussed
in
terms of
two
general
categories:
implications
of the
comparisons
of
source
credibility
and
implications
of
the
study
of
evacuation decision-making.
Finally,
implications
for
further
research
are
addressed.
tI
--
-____________________ -.---.----- •--
-

73
Implications
Arising
from
Source
Credibility
o
During
the
course
of
a
nuclear reactor emergency,
local
emergency
response officials
should
be
integrated
into
the
public
information
system
and
should
constitute
the
public's
primary
source
of
official
accident-relevant
information.
In
the
eyes
of
the
public
this
enhances
the
authority
and
credibility
of
the
local
emergency
response
officials
who will
ultimately
be
responsible
for
the
operations
involved
in
getting
the
public-at-risk
to
undertake
some
protective
action--whether
it
is
evacuation
or
some
other measure.
By
highlighting
the
role
of local
authorities,
confidence
in
them
is
increased
among
the
public-at-risk, which
in
turn
promotes
public
compliance
with
emergency measues.
Of
course
this
does
not
mean
that
local
officials
should
be
the
only
information
disseminators;
it
does
require, however,
careful
coordination
and
cooperation among
emergency
response personnel
at
all
levels--city,
county,
state,
federal--and
between
emergency response
officials
and
political
officials
at
all
levels.
At
TMI political
officials
initially
assumed
and
retained
the
majority
of
the
public
in
formsation
task.d
e
When
an
emergency--either
nuclear
or
nonnuclear--is
in
progress,
the
mass
media
should not
be
relied
upon
as
a
primary
communication
channel
to
the
public.
The
mass
media constitute
a
communication
channel
characterized
by
considerable
"noise";
juxtaposing
"official"
messages
with
other
related
messages (sometimes
conflicting)
promotes
confusion
in
the
mind
of
the
public
regarding
exactly what
response
is
required
of
them.
In
natural disasters,
... -,.--,,__
___
. . ...__J_= -,,. -, '-','J ..,. --
L_
_ _ _ _ _ _

74
the
media have been effectively
used
as
a
supplementary
source
of
information
particularly
when,
as
part
of an
established emergency
plan,
officials
instruct citizens
to
monitor
radio
or
television
broadcasts
for
status
reports
regarding
a
hazard.
In
this
case
the
role of
the
mass
media
is
to
provide
the
public
with information
about
the
immediate
status
of the
hazard
which
allows
the
public
to
determine whether
specific
provisions
of
a
commnunity
evacuation
plan
should
be
implemented.
As
part
of
the
emergency
plan
for
responding
to
the
volcanic
activity
at
Mt.
St.
Helens,
for
example,
the
public
was
instructed
to
monitor
radio
bulletins
on
the
volcano's
status
(Perry
et
al.,
1980a), and
to
evacuate
specified
areas
in
the
event
of
an
eruption
alert.
SWhen
an
emergency
is in
progress,
officials
should
distinguish
the
function
of
providing
public
information about
the
emergency
from
the
function
of
sending messages
which
direct
some
emergency
response.
This
helps
the
public,
to
understand
when
they
are
expected
to
take
an
action
and
when
they are
not.
As
part
of
a
public
information
function,
officials
can
provide
updates
regarding changing conditions
regarding
the
event,
or
describe
a £
range
of
potentially
useful
measures where
the
decision
to
implement
is
left
to
the
public.
Although
ideally
such
matters
are
addressed
before
a
given
disaster
as
part
of
a
general community
preparedness
plan, public
information
during
an
incident
might
also
include
a
description
of
what
constitutes
a
warning
signal and
what
should
be
done
when
such
a
signal
is
received.
It
is
important,
however,
to
separate
such
public
information
clearly
from
an
7 -

75
emergency
response
directive.
This
latter
message
is
one which
instructs
the
population-at-risk
to
begin
a
planned (and
presumably
coordinated)
protective
response;
it
is
intended
to
evoke
full
participation
rather
than
being
an
option
for
which
the
public
is
left
to
make
a
decision
regarding
implementation.
When
these
two
types of
message
are
not
carefully
distinguished, particularly
in
the
case of
evacuations,
the
public can
be
expected
to
undertake
a
range
of
protective
actions
'ome
possibly substantially differing
from
the
actions desired
by
officials)
according
to
widely
differing
time
schedules.
*
In
all
disasters,
particularly
nuclear
disasters,
rumor
control
is
a
critically
important function.
In
general
problems
associated
with
rumor
control
will
increase
to
the
extent
that
the
disaster
or
hazard
is
less
familiar
to
the
public.
In
the
case
of
natural
disasters,
officials
are
usually concerned
with
dispelling
popular
myths
or
technically
inaccurate
conventional
wisdom
regarding
the
event.
In
dealing
with nuclear disasters
the
problem
is
even more
pronounced,
due
to
the
aforementioned
complexity
of
technology
associated
with
the
event,
general
unfamiliarity
of
the
public
with
nuclear
disasters,
and
the
tendency
of
the
public
to
define
most
radiation
hazards relative
to
the
health
dangers
associated
with
nuclear weapons.
The
importance
of
rumor
control
is
underscored
by
the
reported
high
utilization
of
public
information
telephone
lines
during
the
Three
Mile
Island
incident.

76
*
The
public
education
function
is a
particularly
important
component
nf
emergency response
plans
for
dealing
with
nuclear
power plant
accidents.
The
time
to
explain
the
nature
and
specific
dangers
involved with
a
given
hazard
is
before
a
crisis
occurs.
During
the
crisis
official attentions
should
be
devoted
to
achieving
protection
for
the
public;
this
function
is
unnecessarily
complicated
if
the
nature
of
the
risks
in
general
as
well
as
those
specifically
involved
in
the
immediate incident
must
be
described.
By
attempting
to
run
a
"mini" hazard
awareness
campaign during
an
incident,
authorities
force
the
public
into
a
general information
gathering posture
rather
than
allowing
the
public
to
assimilate
and
prepare
to
respond
to
a
specific
emergency
management
strategy.
It
is
likely
that
much
of
what
was
described
as
confusion
on
the
part
of
the
public
during
the
ThI
accident
was
related
to
the
fact
that
public
education
was
being conducted
simultaneously
with
crisis
management.
Numerous
examples
of
public
education
strategies
are
available,
both
from
natural disasters
(Davenport
and
Waterstone,
1979)
and
Civil Defense
(Perry
et
al.,
1980c).
Evacuation Decision-Making
e
Citizen
evacuation response during nuclear disasters
may
be
understood
in
terms
of
the
same
variables which explain evacuation
in
nonnuclear disasters.
Several
researchers
have
pointed
to
the
relatively
high
levels of
spontaneous
evacuations
at
ThI
and
argued
tliat
because
of
this
there
must
be
something
about
nuclear
disasters
that
makes
people
more responsive.
The
implication
heve
-.~ I

77
is
that
there
is
some
unspecified
basic
difference between
nuclear
and
nonnuclear disasters.
The
evidence marshalled
in
the
present
study
suggests
that
the
difference
is
the
fear
or
dread
characteristics (described
as
unique
in
Chapter
Two)
associated
with
nuclear
disasters.
The evacuation
response
at
TMI can
be
explained
using
the
same
variables developed
to
understand
evacuation behavior
in
other natural
and
man-made
disasters. Thus,
citizens
evacuate
when
four
conditions
are
met;
(1)
they
have
accounted
for the
safety
of
their
immediate
household,
(2)
they
have
been
given--by
authorities--or
have
personally developed
a
plan
for
protective action,
(3)
they
believe
that
a
threat does
exist
in
the
environment,
and
(4)
they
perceive
that
upon
impact
this
threat
could
result
in
some
level
of
damage
to
their
person
and
property
(see
Perry,
1979;
Perry
et
al.,
1980a;
Perry
et
al.,
1980b).
At
Three Mile
Island
the
nuclear
nature
of
the
threat
meant
that
people perceived
personal risk
to
be
very
high
(conditior
four),
but
in
general
those
who
evacuated were
people
for
whom
all
four
conditions were
met
(cf.
Zeigler,
1981).
*
The
high
level
of
spontaneous
evacuations
around
TMI
appears
to
be
related
to
the
above
described elevated perception
of
personal
risk
(threat)
by
the
public.
Although
the
same
evacuation
decision-making
variables
seem
to fit
both
nuclear
and
nonnuclear
disasters,
in
the
nuclear
case
citizens
apparently believed
themselves
to
be
at
risk
to
a
considerably
greater
extent. Thus,
the
perceived
negative consequences associated
with
failing
to
undertake
some
protective
action
or
doing
so
too
late
were
r---
----- .-
- --
V
- -- -
- - -
--------
-- --
_____

78
extremely
high.
In
planning
to
manage
nuclear disasters,
one
must
be
sensitive
to
the
effects
of
elevated perceptions
of
risk,
particularly
since
high
levels
of
sponcaneous
or
unsupervised
evacuation
are
not
necessarily
desirable.
Furthermore
the
presence
of
citizen
beliefs
that
nuclear--radiation--disasters
pose
very
high
risks
introduces
a
number
of
logistical
and
procedural
implications
for
managing
an
evacuation:
(I)
If
authorities
issue
evacuation
route
and
destination
to
the
public
early
in
an
incident
with
instructions
to
wait
until
officially
advised
before
leaving,
citizens
are
likely
to
ignore
the
instructions
and
depart
before being
told
to
do
so;
(2)
Evacuation
shadow
effects
will be
multiplied. That
is,
when
an
evacuation
is
announced
for
a
specific
geographic
area,
it
should
be
expected
that
residents
who
are
nearby
but
still
outside
this
area
will
also
evacuate;
(3)
Graded
or
group-specific evacuation orders--for
example,
for
pregnant
women
and
children under
five
years---will
generate
evacuations by others
as
well. In
general,
such
orders
that
would otherwise
divide
families
will
be
heeded
at
least
by
all
members
of
a
given
family;
(4)
Planning
attention
needs
to
be
devoted
to
the
problems
associated
with
getting
evacuees back
to
an
area
after
they
have
been evacuated.
Although
not
the
case
at
TMI,
if
a
significant radioactive
release
had
caused
citizens
to
evacuate
it
is
not
obvious
that
they
would
respond
readily
to
an
"all
P.t

79
clear"
signal.
A
study should
be
made
of
how
one
should
structure
and
disseminate
a
message
that
an
area once
threatened
by
radiation
is
nov
safe.
Although
they
are
not
directly derivable
from
the
specific
data
presented
here,
two
additional general conclusions
may
be
inferred
based
upon
the
overall analysis.
These
conclusions
have
implications
for
the
"dual-use" philosophy
adopted
by
FEMA,
and
for the
kind
of
organizational
coordination
required
to
effectively manage
large-scale evacuations.
•
The
"dual
use"
philosophy
appears
to be
founded
upon
reasonable
assumptions
in
that
the
basic principles
of
human
response
to
natural
hazards
also
describe
human
response
to
nuclear
threats.
The
data
analyzed
here
indicate
that
with respect
to
warning
source
credibility
and
evacuation decision-making,
the
same
theoretical
propositions
may
be
used
to
explain human
response
to
threats
associated with evacuation
in
two
natural
hazards
and
a
nuclear
power
plant
accident.
It
was
established, however,
that
the
analyst
must document
and
allow
for
certain unique
aspects of
all
disasters;
it
was
found
for
example
that
public
perception
of
threat tends
to
be
elevated
in
nuclear
threats.
*
Inter-organizational
and
inter-agency
coordination
and
preparedness
for
ordering
and
overseeing
a
mass
evacuation
are
crucial
problems
in
both
nuclear
and
nonnuclear disasters.
As
Quarantelli
(1980:149)
has pointed
out,
at
almost
all
jurisdictional
levels
such
preparedness
is
poor, and
the
data
examined
here
on ThI serve
to
further
verify
this
hypothesis.
It
is
particularly
interesting
to
note
that
in
the
face
of poor
coordination
among
and
guidance

80
from
authorities,
the
public
seems,
to
some
extent,
to
be
able
to
take
care
of
itself.
Remember
that
39
percent
of
the
population
within
fifteen
miles
of
TMI
managed
to
evacuate successfully
with
a
minimum
of
guidance
from
authorities.
This
should
not
be
interpreted,
however,
as
evidence
that
in
all
disasters
the
public
will survive
with
a
minimum
of
help
from
authorities. Instead,
it
indicates
that
even
in
the
absence
of
official
coordination,
the
public
is
not
reduced
to
panic
flight
or
a
total
breakdown
of
reason
(cf.
Quarantelli,
1960).
There
is
no guarantee
that
the
remaining
61
percent
of
the
population
around ThI
could
have
evacuated
without
some
coordinated
official intervention.
To describe
the
event
accurately,
it
must
be
acknowledged
that
inter-organizational
coordination
did
improve over
time
during
the
THI
4
incident.
It
seems
apparent
that
some
evacuations
can
proceed
with
a
minimum
of
coordination
by
officials;
this
has been
true
for
years
in the
relatively
small
and
short-term
evacuations
associated with
natural
disasters
in
the
United States.
As
the
number
of
people
involved and
the
time
outside
the
risk
area increases,
however,
it is
less
likely
that
mass evacuations
can
be smoothly
executed without
high
levels
of
inter-organizational preparedness
and
coordination.
It is
likely
that
the
number
of
spontaneous
evacuees
at
TKI
approaches
the
upper
limit
of
the
size
of
mass
evacuation
which
can be
.accomplished
in
the face
of
relatively
low
levels
of
inter-organizational
coordination.
If
we
are
to
accomplish
mass
evacuations
on
the
scale
necessary
to
implement
such
programs
as
Crisis
Relocation
or
as
might
be
required
in
response
to
a
I!

81
nuclear
power
plant
accident involving
a
breach
of
containment,
inter-organizational
coordination
among
emergency
management
organizations
must
be
substantially
improved.
Implications
for
Further
Study
It
is
interesting
to
note
that
most
of
the
conclusions
and
implications
discussed above
are
neither
striking
nor
absolutely unique
in
the
social
science
literature.
In
many
cases,
they
reflect
suggestions
made
in
connection
with
research
on
a
variety
of
disaster
agents
over
the
years.
It
is
of
course
important
to
document empirically
cross-disaster
agent
applicability
of
planning
and
citizen
response
principles,
as
this
report
did.
However,
given
that
much
information
is
available
from
research
on
how
people
can
be
expected
to
respond
to
evacuation
orders,
how
public
information
should
be
handled
during
disasters,
how
to
issue
and
structure
evacuation
advisories
and
warnings,
etc.,
it is
interesting
that
apparently
so
little
of
this
available
information
was
utilized
in
the
management
of
the
TMI
accident. Often
planners
and
policy-makers
read
the
conclusions
of
research
reports
and
respond
by
saying,
"I
already
knew
all
that".
The
experience
at
TMI,
which also occurs
in
natural
disasters, where various emergency
management
problems arose
in
spite of
the
availability
of
research-based
planning
and
response principles which
bear
upon
the
issues
causes
one
to
look
askance
at
such
claims.
More
importantly,
though,
it
raises
questions
about
how
research
results
are
disseminated
from
researchers
to
planners
and
policy-makers,
as
well
as
how
these
latter
actors
evaluate
and
incorporate
research
information
into
the
emergency
management

82
process.
Clearly, much information regarding emergency
response
performance
of
citizens
that
was
available
before
THI
was
not
used
in
managing
that
incident.
It
my
be that
the
research
was
not
used
because
relevant officials
were
not
aware
of
it,
because
it
was
not
in
a
form
which
could
be
implemented,
because
it
was
not
in
a
form
which
could
be
understood,
or
because
the
research
was
perceived
to
be
irrelevant
to
the
problems.
The
point
is
that
to
date
little
attention has been
devoted
to
examining
the
channels through
which research
findings
reach
those people
at
different jurisdictions
(local,
state
or
federal)
who
would
use
them,
and
how
feedback
travels
from users
back
to
the
research
community.
Understanding
and
perhaps
formalizing
such
communication
channels would
seem
to
be one
way
to
insure
that
research
is
relevant
to
planning
and
operational concerns
and
is
presented
in
a
fashion
which
lends
itself
to
evaluation
and
utilization
by
policy-makers
and
implementers.
Therefore,
an
important
issue
for
further study
is
the
process through
which
research
results
are
disseminated
to
planners
and
policy-makers,
and
the
factors
which
influence
utilization
patterns.
The data
reported
in
Chapter
Five
indicated
that
levels
of
perceived
threat
were very
high
in the
nuclear disaster relative
to
the
two
natural
disasters.
This
finding
is
also supported
by
several
attitudinal
studies
in
the
psychological
literature
which
indicate
that
people view
radiological
hazards
as
particularly
threatening
and
potentially very
dangerous.
Documenting
that
this
differential exists,
however,
is
only
a
first
step
toward
understanding
!±Z
people
define
radiaiton
dangers
as
so
threatening
and
the
implications of
this
for the
management
of
emergencies involving
a
nuclear component.
A
study
of
the
calculus used

83
by
citizens
in
assessing
risks
associated
with
radiation
relative
to
other
hazards--i.e.,
delineating
factors
which
influence
citizens
i '
definitions of risk--would
provide
a
basis
for
making decisions
about
how
best
to
commnicate
technical
information
on
the
forms
and
consequences
of
radiation
such that
citizens
can
make
informed
evaluations
of
risk.
Furthermore,
this type
of
study could
lay
the
groundwork
for
addressing
three
related research questions.
"
how
to
reduce
the
anxiety
reportedly
associated
with nuclear
disasters which
is
in
itself
harmful
to
the
public
and
potentially
could
serve
as
a
limiting
factor
on
citizens'
ability
to
comply
with emergency
instructions;
*
how
to
address
the
problem (documented
by
research) of
the
public's
tendency
to
equate
radiation
risks
generated
by
nuclear
weapons
with
the
radiation
risk
associated with nuclear
power
plants;
and
"
what
is
the
relationship
among
peoples'
cognitive
frame
of
reference,
their
verbally expressed
attitudes,
and
their
emergency
response
behavior,
and
what implications
does
this
have
for
the
design
and
implementation
of
emergency
plans.
Finally,
the
data
examined
in
this
report
indicate
that
there
is
a
need
for
research
on
the
design
and
implementation
of
both public
information
program
regarding
nuclear
disasters
and
dissemination
programs
for
specific
emergency response
.plans.
Such
research might
appropriately
address
four
general
issues.
First,
one
should
begin
to
identify
those
channels through
which risk
and
response information
can
be
efficiently
and
effectively disseminated
to
the
public
as
part
of
a
long-term
comprehensive
program
to
enhance
awareness
of
nuclear hazards.
1 !

84
Second,
an
attempt
should
be
made
to
specify
dissemination
roles
for
different
levels
of
cadre
(e.g.,
federal, state
and local
authorities)
in
the
United
States
emergency
management
system.
Third,
because
the
public
tends
not
to
distinguish
among
risks
associated
with
different
types
of
nuclear disasters--e.g.,
nuclear war,
power
plant
accidents,
radiological
transportation accidents--it
is
necessary
to
devise
a
strategy
to
sensitize citizens
to
distinctions among
them
which
have
implications
for
the
types
of
mitigation
or
emergency response
behaviors
which
should
be
undertaken.
Fourth,
there
is
a
need
to
devise
a
strategy
for
selecting
which emergency response instructions,
as
well
as
what
parts
of
existing
J
emergency management
plans,
should
be
disseminated
to
the
public.
This
would
include
decisions
about
what
information
should
be
part
of
a
general
public
education
program
and
what
should
be
disseminated
during
the
warning
phase
when
disaster
impact
is
believed
to
be
iminent.
Q~ -

85
BIBLIOGRAPHY
AND
REFEENCES
American
Nuclear Society,
1979.
"Special
Report:
The
Ordeal
at
Three
Mile
Island".
Nuclear
Neds.
Special
Report
April
6,
1979:1-6.
Anderson,
Jon, 1968.
Cultural
Adaptation
to
Threatened
Disaster.
Human
Organization
27
(Winter):298-307.
Anderson,
William
A.,
1969.
Local
Civil
Defense
in
Natural Disaster.
Colusbus,
Ohio:
Ohio
State
University
Disaster
Research
Center.
Barnes, Kent,
Brosius,
James,
Cutter,
Susan and
Mitchell,
J.K.,
1979.
Responses
of
Impacted
Populations
to
the
Three
Mile
Island
Nuclear
Reactor
Accident.
New
Brunswick,
New
Jersey:
Department
of
Geography, Rutgers
University.
Barton,
Allen,
1970.
Commnities
in
Disaster.
New
York:
Anchor
Books.
Chenault,
William,
Hilbert,
Gary
and
Reichlin,
Seth,
1979.
Evacuation
Planning
in
the
TMI
Accident.
McLean,
Virginia:
Human
Sciences
Research.
Christiansen, Robert,
1980.
Eruption
of
Mt.
St.
Helens:
Volcanology,
Nature
285
(June):531-533.
The
Columbian,
1980.
Mt.
St.
Helens
Holocaust.
Vancouver,
Washington:
The
Columbian,
Incorporated.
Cutter,
Susan,
Brosius, James, Barnes,
Kent and
Mitchell,
J.K.,
1979.
Special
Session
on
Three
Mile
Island:
Risk
Evaluation
and
Evacuation
Responses.
Proceedings,
Middle
States
Division, American Association
of
Geographers, Volume
12:80-88.
Davenport,
Sally
and
Waterstone,
Penny,
1979.
Hazard
Awareness
Guidebook:
Planning
for
What
Comes
Naturally. Austin,
Texas:
Texas Coastal
and
Marine
Council.
Donnelly,
Warren
and
Kramer,
Donna,
1979.
Nuclear
Power:
The
Three
Mile
Island
Accident
and Its
Investigation. Washington,
D.C.:
Library
of
Congress, Congressional Research
Service.
Dynes, Russell,
Quarantelli, Enrico
and
Kreps,
Gary,
1972.
A
Perspective
on
Disaster
Planning.
Columbus,
Ohio,:
Ohio
State
University
Disaster
Research
Center.
Earle,
Timothy
C.,
1981.
Public
Perception
of
Industrial Risks. Seattle,
Washington:
Battelle
Human
Affairs
Research
Centers.
Fabrikant,
Jacob,
1979.
Reports
of
the
Public
Health
and
Safety
Task
Force.
Washington,
D.C.:
Government Printing
Office.
T i

IN
I
86
Flynn,
C.
B.,
1979.
Three
Mile Island
Telephone
Survey
(NUREG/CR-1093).
Washington,
D.C.:
Nuclear
Regulatory
Commission.
Friedman,
S.M.,
1981.
Blueprint
for
Breakdown
at
TMI.
Journal
of
Communication
(Spring): 116-128.
Fritz,
Charles
E.,
1961.
Disasters.
In R.
Merton
and
R.
Nisbet
(eds),
Contemporary
Social
Problems.
New
York:
Harcourt,
Brace
and
World.
Fritz, Charles
E.,
1968.
Disasters.
In
International
Encyclopedia
of
the
Social
Sciences.
New
York:
MacMillan
and
The
Free
Press.
Gark,
S.,
1963.
Thermonuclear
Survival.
Missouri Medicine,
60:860-862.
Geophysics
Program,
1980.
Eruption
of
Mt.
St.
Helens:
Seismology, Nature
285
(June):529-531.
Gillespie,
David
and
Perry,
Ronald
W.,
1976.
An
Integrated
Systems
and
Emergent
Norm Approach
to
Mass
Emergencies.
Mass
Emergencies,
I
(number 3):303-312.
Glass,
Albert,
1956.
Psychologic Considerations
in
Atomic Warfare.
United
States
Armed
Forces
Medical
Journal,
7
(May):625-639.
Comittee
on
Government
Operations,
1979.
Emergency
Planning
Around
U.S.
Nuclear Power
Plants:
Nuclear
Regulatory Commission
Oversight.
Washington,
D.C.:
U.S.
House
of
Representatives.
Greene,
M.
R.,
Perry,
Ronald
W.,
and
Lindell,
Michael,
1980.
The
March
1980
Eruptions
of
Mt.
St.
Helens. Seattle,
Washington: Battelle
Human
Affairs
Research
Centers.
Grinspoon,
Lester,
1964.
Fallout
Shelters
and
the
Unacceptability
of
Disquieting
Facts.
In
G.
Grosser,
H.
Wechler,
and
M.
Greenblatt
(eds),
The
Threat
of
Impending
Disaster.
Cambridge:
MIT
Press.
Gruntfest,
E.,
Downing,
T. E.
and
White, Gilbert,
1978.
Big
Thompson
Flood
Exposes Need
for
Better
Flood Reaction
System
to
Save
Lives.
Civil
Engineering,
78
(February):
72-73.
Haglund,
K.
A.,
1979.
"At
Hershey:
Medical
System
Near
Failure
During
Three
Mile
Island".
New
Physician,
28:31-32.
Hans,
J.
and
Sell,
T.,
1974.
Evacuation
Risks. Washington,
D.C.:
U.S.
Environmental Protection
Agency, Office
of
Radiation
Programs.
Hoheneuser,
Chris,
Kasperson,
Roger, and
Kates,
Robert,
1977.
The
Distrust
of
Nuclear
Power.
Science,
196
(April):25-34.
Janis,
Irving,
1962.
Psychological
Effects
of
Warnings. In
G.
Baker
and
D.
Chapman,
Man
and
Society
in
Disaster.
New
York:
Basic
Books.
Il
, , ,, ,b ,

87
Janis, Irving
and
Mann,
Leon,
1977.
Emergency
Decision
Making.
Journal
of
Human
Stress,
3:35-45.
Kaplan,
F.,
1978.
Enhanced-Radiation
Weapons. Scientific American,
238
(January)
:44-51.
Kasperson,
J.
X.,
Kasperson,
R.
E.,
Hohenemser,
C.
and
Kates,
R.
W.,
1979.
Institutional Responses
to
Three
Mile
Island.
Bulletin
of
the
Atomic
Scientists,
35
(December):
20-24.
Kemeny,
John
G.,
1979.
Report
of
the
President's
Commission
On
the
Accident
At
Three Mile
Island.
Washington,
D.C.:
Government
Printing
Office.
Kiyoshi,
Shimizu,
1967.
Little-known
Effects
of
the
Bomb. Japan
Quarterly,
14:93-98.
Korosec,
Michael,
Rigby,
James, and Stoffell,
Keith,
1980.
The
1980
Eruption
of
Mount
St.
Helens
Washington,
Part
I:
March
20-
May
19,
1980.
Olympia,
Washington:
Washington
State
Department
of
Natural
Resources,
Division
of
Geology
and
Earth
Sciences,
Information
Circular
Number
71.
Kreps, Gary,
1979.
A
framework
for
comparing
nuclear
and
nonnuclear
disasters
in
terms
of
key
defining
properties
of
disaster
events
and
generic
functions
of
disaster
response.
Williamsburg,
Virginia:
Department
of
Sociology,
College
of
William
and
Mary.
The
Lancet,
1979.
"The
Worst
Nuclear Power
Plant
Accident
Yet".
The
Lancet
(April
28):909-910.
Lewis,
Kevin,
1979.
The
Prompt
and
Delayed
Effects
of
Nuclear
War.
Scientific
American,
241:35-47.
Lifton, Robert,
1964.
Psychological Effects
of the
Atomic
Bomb
in
Hiroshima.
In
G.
Grosser,
H.
Wechler,
and
M.
Greenblatt
(eds),
The
Threat
of
Impending
Disaster.
Cambridge:
MIT
Press.
Lifton, Robert,
1967.
Death
In Life.
New
York:
Simon
and
Schuster.
Lindell,
Michael,
Earle,
Timothy,
Hebert,
Don and
Perry,
Ronald
W.,
1978.
Radioactive
Wastes:
Public
Attitudes. Seattle,
Washington:
Battelle
Human
Affairs Research Centers.
MacDonald,
Gordon,
1972.
Volcanoes.
Englevood Cliffs,
New
Jersey:
Prentice-Hall.
Marshall,
Eliot,
1981.
New
A-Bomb
Data
Shown
to
Radiation
Experts.
Science,
212
(June
19):1364-1365.
Martin,
Daniel,
1980.
Three Mite
Island.
Cambridge:
Ballinger.
McLuckie,
Benjamin,
1970.
Warning
Systems
in
Disaster. Columbus,
Ohio:
Ohio
State
University
Disaster
Research Center.
4- | -.
-_4
-- - -

88
Mileti, Dennis,
1974.
A
Normative
Causal
Model
Analysis
of
Disaster
Warning
Response.
Boulder,
Colorado:
Ph.D. Dissertation, Department
of
Sociology,
University
of
Colorado.
Mileti,
Dennis
and
Beck,
Woody,
1975.
Communication
in
Crisis,
Communication
Research,
2
(January)
:24-49.
Mileti,
Dennis
and
Harvey,
Pat,
1977.
Correcting
for
the
Human
Factor
in
Tornado Warnings. Paper presented
at
the
10th
Annual
Conference
on
Severe
Local
Storms,
Omaha,
Nebraska.
Moore, Harry
E.,
1958.
Tornadoes
Over
Texas.
Austin,
Texas:
University
of
Texas
Press.
Morland, Howard,
1979.
"The
Meltdown
That
Didn't
Happen". Harper's,
259
(October):
16-23.
Owen,
H.
James,
1977.
Guide
for
Flood
and
Flash
Flood Preparedness
Planning.
Washington,
D.C.:
National
Weather
Service,
Disaster
Preparedness
Staff.
Parr,
Arnold,
1969.
A
Brief
on
Disaster Plans. EMO
National
Digest,
9:13-15.
Perkins,
Carol,
1980.
Fiddling While
Volcano
Burns.
Seattle
Post-
Intelligencer,
May
10,
Page
Al.
Perry,
Ronald
W.,
1979.
Evacuation Decision-Making
in
Natural
Disasters.
Mass
Emergencies,
4
(Number 1):25-38.
Perry,
Ronald
W.,
1980.
Population
Evacuation
in
Nuclear versus
Nonnuclear Disasters. Paper
presented
for
the
Committee
on
United
States
Emergency Preparedness.
Washington,
D.C.:
National
Academy
of
Sciences-National
Research
Council.
Perry, Ronald
W.,
Gillespie,
Dennis,
and
Gillespie,
David,
1974.
System
Stress
and
the
Persistence
of
Emergent Organizations. Sociological
Inquiry,
42
(February):113-121.
Perry, Ronald
W.,
Greene,
Marjorie,
and Lindell, Michael, 1980a.
Human
Response
to
Volcanic
Eruptions:
Mt.
St.
Helens,
May
18,
1980.
Seattle,
Washington:
Battelle Human
Affairs Research Centers.
Perry,
Ronald
W.,
Lindell, Michael,
and
Greene,
Marjorie,
1980b.
Evacuation Decision-Making
and
Emergency
Planning.
Seattle,
Washington:
Battelle
Human
Affairs
Research
Centers.
Perry,
Ronald
W.,
Lindell, Michael, and
Greene,
Marjorie,
1980c.
Evacuation
Warning
and
Crisis
Relocation.
Seattle,
Washington:
Battelle
Human
Affairs
Research
Centers.

_______________II____ ____. . . .
89
Perry, Ronald
W.,
Lindell,
Michael,
and
Greene,
Marjorie,
1981.
Evacuation
Planning
in
Emergency
Management.
Lexington,
Massachusetts:
Heath-Lexington.
Quarantelli,
E.
L.,
1960.
Images
of
Withdrawal
Behavior
in
Disasters.
Social
Problems,
9
(Number
1):68-79.
Quarantelli,
E.
L.,
1980.
Evacuation
Behavior
and
Problems.
Columbus,
Ohio:
Ohio
State
University Disaster
Research Center.
Quarantelli,
Enrico, and
Dynes,
Russell,
1970.
Property
Norms
and
Looting:
Their
Patterns
in
Community
Crises.
Phylon,
31
(Summer):
168-182.
Quarantelli,
Enrico,
and Dynes,
Russell,
1972.
When Disaster
Strikes.
Psychology
Today,
5
(February)
:67-70.
Quarantelli,
Enrico,
and
Dynes,
Russell,
1977.
Response
to
Social
Crisis
and
Disaster,
Annual
Review
of
Sociology,
3:23-49.
Quarantelli,
Enrico,
and
Taylor,
Verta,
1977.
Some
Views
on
the
Warning
Problem
in
Disasters
as
Suggested
by
Sociological
Research. Paper
read
at
the
American
Meteorological
Society
Conference
on
Severe
Local
Storms,
Omaha, Nebraska.
Rosenfeld,
Charles,
1980.
Observations
on
the
Mount
St.
Helens Eruption,
American
Scientist,
68
(September-October):494-509.
Ross, Donald,
1952.
The
Emotional
Effects
of
an
Atomic
Incident.
Cincinnati
Journal
of
Medicine,
33
(February):38-41.
Rubin,
Claire,
1979.
Disaster
Mitigation:
Challenge
to
Managers.
Public
Administration
Times,
2:1-2.
Sandman,
Peter
and
Paden, Mary,
1979.
"At
Three
Mile Island".
Columbia
Journalism
Review,
July/August,
pages
43-58.
Schorr,
John
and
Goldsteen, Raymond,
1980.
Public Response
to
a
Nuclear
Reactor
Accident.
Deland,
Florida:
Department
of
Sociology, Stetson
University.
Sillar,
William,
1975.
Planning
for
Disasters. Long Range
Planning,
8:2-7.
Slovic,
Paul,
Lichtenstein,
Sarah
and
Fischhoff,
Baruch,
1980.
Images
of
Disaster:
Perception
and
Acceptance
of
Risks
from
Nuclear
Power.
Eugene,
Oregon:
Decision
Research,
Inc.
Stallings,
Robert,
1971.
Communications
in
Natural
Disaster. Columbus,
Ohio:
Ohio
State
University Disaster Research
Center.
U.S.
Nuclear
Regulatory
Commission,
1981.
Report
to
Congress
on
Status
of
Emergency
Response
Planning
for
Nuclear
Power
Plants.
Washington,
D.C.:
Nuclear
Regulatory
Commission,
NUREG-0755.
____ ____ ___ ____ ____ ___

90
U.S.
Senate
Hearings,
1980.
Disaster
Assistance
Pacific
Northwest--Mount
Saint
Helens Eruption.
Washington,
D.C.:
Committee
on
Aopropriations.
White;
Gilbert,
1975.
Flood
Hazard
in
the
United
States.
Boulder,
Colorado:
University
of
Colorado.
Williams, Harry
B.,
1964.
Human Factors
in
Warning
and
Response
Systems.
In
H.
Grosser
(ed),
The
Threat
of
Impending
Disaster.
Cambridge:
MIT
Press.
Windham,
George
0.
and
others,
1977.
Reactions
to
Storm Threat
During
Hurricane
Eloise.
State
College,
Mississippi:
Mississippi
State
University.
Zeigler,
Don,
Brunn,
Stan
and
Johnson,
J.
H.,
1981.
Evacuation
from
a
Nuclear Technological Disaster.
The
Geographical
Review,
71
(January):
1-16.
4
I,
I
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_

MANDATORY
STANDARD
DISTRIBUTION
LIST
FOR
RESEARCH
REPORTS
(ALL
PROJECTS)
(Number
of
Copies
-
One
unless
otherwise
indicated)
Federal
Emergency
Management
Agency
Mitigation
and Research
ATTN:
Administrative
Officer
Washington,
D.C.
20472
(60)
Assistant
Secretary
of
the
Army
(R&D)
ATTN:
Assistant
for
Research
Washington,
D.C.
20301
Chief
of
Naval
Research
t
Washington, D.C.
20360
Defense
Technical
Information Center
Cameron
Station
Alexandria,
Virginia
22314
(12)
Oak Ridge
National
Laboratory
ATTN:
Librarian
P.O.
Box
X
Oak
Ridge,
Tennessee
37830
Los
Alamos
Scientific
Laboratory
ATTN:
Document Library
Los
Alamos,
N.M.
87544
The
RAND
Corporation
ATTN:
Document Library
1700
Main
Street
Santa
Monica,
CA
90401
$-
-- ' .. . . ,' .. ., - -- . .U. . . .

2300
DISTRIBUTION
LIST
Organization Organization
Dr.
William
W.
Chenault
The
Dikewood
Corporation
Human Sciences
Research,
Inc.
1613
University
Blvd.,
N.E.
Westgate Research
Park
Albuquerque, N.M.
87102
7710
Old
Springhouse
Road
McLean,
VA
22101
Ohio
State
University
Disaster
Research Center
Dr.
Jiri
Nehnevajsa
128
Derby
154
North
Oval
Mall
Professor
of
Sociology
Columbus,
OH
43210
University
of
Pittsburgh
Pittsburgh,
PA 15213
Dr.
Gerald Klonglan
Dept.
of
Sociology
and
Anthropology
Dr. Conrad
Chester
Iowa
State
University
ERDA,
Holifield National
Laboratory Ames,
IA
50010
7
P.O.
Box
X
Oak
Ridge,
TN
37830
Mr. Howard
McClennon, President
International
Association
of
Mr.
Walmer
E.
Strope
Fire Fighters
Center
for
Planning
and
Research
1750
New York
Avenue,
NW,
3rd
Fl.
5600
Columbia
Pike
Washington,
D.C.
20006
Bailey
Cross Roads,
VA
22041
General
Manager
Mr.
Don
Johnston International
Association
of
Research
Triangle
Institute
Fire Chiefs
P.O.
Box
12194 1329
-
18th
Street,
NW
Research
Triangle
Park,
NC
27709
Washington,
D.C.
20036
Mr.
Richard
K.
Laurino
Mr. Bjorn
Pedersen
Center
for
Planning
and
Research,
Inc.
International
Association
of
2483
East
Bayshore Road Chiefs
of
Police
Palo
Alto,
CA
94303
11
Firstfield Road
Gaithersburg,
MD
20760
Mr.
Bela
H.
Banathy
Far
West
Laboratory
Mr.
Ferris
Lucas
1855
Folsom
Street
National
Sheriff's
Association
San
Francisco,
CA
94103 1250
Connecticut
Ave.,
NW
#320
Washington,
D.C.
20036
Mr.
Ralph
L.
Garrett
2
Catspaw
Cape
Mr.
Gerald
W.
Collins,
Exec.
V.P.
Coronado,
CA
92118
National
Defense Transportation Assn.
1612
K
Street,
NW,
Suite
706
Mr.
Jesse Pugh,
III
Washington,
D.C.
20006
Dept.
of
Crime Control
and
Public
Safety Dr. Gilbert
F.
White
116
West
Jones
Street
University
of
Colorado
Raleigh,
NC
27611
IBS
#6
Campus
Box
482
Mr. Michael Kaltman
Boulder,
CO
80309
Nuclear
Regulatory
Commission
P-302
Washington, D.C.
20555
..-

Organization
Organization
National
Fire
Protection
Association Mr.
Robert
Harker
ATTN:
Library
SYSTAN,
Inc.
470
Atlantic
Avenue
28
Aliso
Way
Boston,
MA
02210
Menlo
Park,
CA
94025
National
Bureau
of
Standards Ms.
Marie
Hayman
Disaster Research
Coordinator
International
City
Management
Assn.
ATTN:
Mr.
C. G.
Culver
1140
Connecticut
Ave.,
NW
Office
of
Federal Building Technology
Washington,
D.C. 20036
Center
for
Building Technology
Washington,
D.C.
20234
Ms. Claria Rubin
Academy
of
Contemporary
Problems
Command
and
Control
Technical
Center
1501
Neil
Avenue
The
Pentagon
-
BE
685
Columbus,
OH
43201
Washington,
D.C.
20301
M
National
Academy
of
Sciences
(JH-312) System
Planning
Corporation
Commission
on
Sociotechnical
Systems
1500
Wilson
Boulevard
CUSEP
Suite
1500
2101
Constitution
Avenue,
NW
Arlington,
VA
22209
Washington,
D.C.
20418
Dr.
John
R.
Christiansen
The
Council
of
State
Governments
Department
of
Sociology
ATTN:
Mr.
Hubert
A.
Gallagher
183
Faculty
Office Bldg.
Disaster Assistance
Project
Brigham
Young University
1225
Connecticut
Avenue,
NW
#300
Provo,
UT
84601
Washington,
D.C.
20036
Dr.
Abner
Sachs
Dr.
Joseph
E.
Minor Science
Applications,
Inc.
Texas
Tech
University
1651
Old
Meadow
Rd.,
#620
Department
of
Civil
Engineering
McLean, VA
22101
P.O. Box
4089
Lubbock, TX
79409
Stanford
Research
Institute
ATTN:
Librarian
Dr.
John
W.
Billheimer
333
Ravenswood
Avenue
SYSTAN,
Inc.
Menlo Park,
CA
94025
343
Second
Street
P.O.
Box
U
Mrs.
Lori
O'Neill
Los
Altos,
CA
94022
DOE,
ERA/OUS
Emergency
Electric Power
Admin.
Ms.
Hillary
Whittaker (RG741)
National
Governors
Association
2000
M
Street,
NW
444
North
Capitol
Street,
NW
Washington,
D.C.
20461
Washington, D.C.
20001
Mr.
Duane
Baltz
Ms. Elizabeth
Eicherly
National
Association
of
Counties
Pennsylvania
Emergency Management
444
North
Capitol
Street,
NW
Agency
Washington, D.C.
20001
Transportation
and
Safety
Building
Harrisburg,
PA
17120
Emergency
Management
Project
Director
National
League
of
Cities
1620
I
Street,
NW
Washington,
D.C.
20006

0
10l14,0
1 0, Wrt 0 0t .0
m
0)
:3
0' O
0 0 4 a m m n
00
.4
00
6
0104
P P,1Oc,.-
0.0 0 01 rtt 0 0 < 1-0mm"I r
P1m0
M0 M1 m
0.
Go
~
4
M0 0r(Ar
(
0e) 1-4
OrtO
Go 0
m tI-. PI
0~
rt
M U )
00M
f
n 500000 4 4O'"En.. -P
t'4
3 .
mt
0
L
140
P~r Dm
m0000o. 0014
pi
:10
00 00 m m 0
a1
"- 0 10
W 0r.1
P "m % 0
00
m- l0b0 I-'
kA
00V
01'Ft~~l4
*
lrt
0FtP
4
* O
t
F .
M
m
0 1
00 0
r,
n
0P
0
00
m
t .
:s
0000 )I.'. 0 0-4
000
1-~a~
f 0
0-
0-
m0
0 4t P
n '
M1
0 4:
0
ft
n0r
0i
00
0 ?r
0 00 I.0 m
m
1-6 0
1m0.0
0
V
000
4
(A
0
D.
1
0 0.
t03
0 0
w
q
tq :
q
0 t
0c
m "
0 m.~
cm 0 m* -Qm
0 ot 0
0
=1
0 0 " 0
:
W 0
0000
0 c0 01r nO
m
00
m0 a 00h
01a'
0 0 n 0i. 1.&
0 "
CJM~
CL00
M-
r-' rfl W.
L
'I0~ M 00 0Mt
00.
W'O~.
000
IM
P0 0.
400 0 PI~' 1-4 0
(1
"m
1
P
0~0
o
m1
m
o
D
0I-
0
b
:mnm0 m
~
0 ~
t
0
(D
(
0
m m1
0.r
Wg
03
0rt
00 .P1
m v m
00
3 0
m
03
:r
t
0 bto
1
al
Op t
1- 0
01i
00
0
C 01-.
00m
0
m 0
S
00=
00
0 t M 0rt:31-
:
.
0.0
Ot
mt
0
0. m ON
rt
m
I. .
00
11 M IM
.M M f100 13.3
3
n0 0.P
o
00 )-
0M0
9.P..
n
0
M
0
(D -: b :t
0
mQQ
z M 0
:3
m 0 Ot' 0 .I ,
0
~
0)
0
t
0 C
oo.0
4
0
rm M
-m
CL
0
0
n
-
0 ~P
0
00. gm
tm00 .%
P1
:3 P1
:3c 4n 3
0 0
14m
vgo
r .