University of Southern California Dissertations and Theses

Determining acceptable seismic risk: A community participation-based approach

(USC Thesis Other)

Determining acceptable seismic risk: A community participation-based approach

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and continuing
from left to right in equal sections with small overlaps.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations appearing
in this copy for an additional charge. Contact UMI directly to order.
ProQuest Information and Learning
300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DETERMINING ACCEPTABLE SEISMIC RISK:
A COMMUNITY PARTICIPATION BASED APPROACH
by
Vilas Sitaram Mujumdar
A Dissertation Presented to the
FACULTY OF THE SCHOOL OF POLICY, PLANNING,
AND DEVELOPMENT
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PUBLIC ADMINISTRATION
August 2000
Copyright 2000 Vilas Sitaram Mujumdar
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3018111
Copyright 2000 by
Mujumdar, Vilas Sitaram
All rights reserved.
___ ®
UMI
UMI Microform 3018111
Copyright 2001 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
SCHOOL OF POLICY. PLANNING. AND DEVELOPMENT
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 90089
This dissertation, written by
Vilas M a j i m i b a
under the direction o f h.D issertation
Committee, and approved by all its
members, has been presented to and
accepted by the Faculty o f the School o f
Policy, Planning, and Development, in
partial fulfillm ent o f requirements fo r the
degree o f
DOCTOR OF PUBLIC ADMINISTRATION
Dean
DaU
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgments
The subject matter of broad seismic risk has been conceived over a period of
years. The focus on community participation in determining acceptable seismic
risk was arrived through discussion with a number of people. In particular, I
would like to thank the following individuals who helped me during studies
culminating in this dissertation.
Professor Ross Clayton, Chair of the dissertation committee, for his never
ending encouragement, thought provoking insights, thorough and timely review
of each draft and overall guidance. Professor William Petak, for providing critical
review of the material and guiding the overall quality by asking probing questions.
Professor Chet Newland for providing continual moral and intellectual support
and discussing the subject with warmth and enthusiasm; State of California,
specifically the Division of State Architect which allowed me to pursue my
academic goals; Nancy Holtz who was an invaluable help in word processing
and creating graphics.
Finally, a special note of appreciation to my wife Ingrid who not only helped in
word processing and proofreading numerous drafts but continued to provide
critical review of the manuscript. Without her understanding and continual
support throughout my academic studies such an accomplishment would not
have been possible.
ii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
Acknowledgem ents---------------------------------------------------- ii
List of Figures------------------------------------------------------------------------------- x
List of Tables -------- — --------------------------------- — xi
A bstract--------------------------------------------------------------------------------------- xii
INTRODUCTION---------------------------------------------------------- 1
Background----------------------------------------------------------------------------------6
Research Questions---------------------------------------------------------------- 9
Proposed Methodology------------------------------------------------------------- 9
Organization of the Study----------------------------------------------- 12
Notes---------------------------------------------------------------------------------------------- 14
CHAPTER 1: GENERAL RISK AND NATURE OF SEISMIC R IS K 15
Risk - General--------------------------------------------------------------------------- 15
Risk Definitions--------------------------------------------------------------------------- 17
Natural Hazard R isk-------------------------------------------------------------------- 20
Nature of Seismic Risk-----------------------------------------------------------------21
Seismic Regulations in California-------------------------------------------------- 23
Notes---------------------------------------------------------------------------------------------- 25
CHAPTER 2 : SOCIAL AND CULTURAL THEORIES OF R IS K 26
General-------------------------------------------------------------------------------------- 26
Risk Perception------------------------------------------------------------------ 27
Dramatic Cause---------------------------------------------------------------- 28
Probabilities--------------------------------------------------------— --------- 29
Risk Presentation-----------------------------------------------------------------30
Values-------------------------------------------------------------------------------- 31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Environment-------------------------------------------------------------------------------------31
Communities-------------------------------------------------------------------------- 31
Organizations------------------------------------------------------------------------------32
Institutions---------------------------------------------------------------------------------- 33
Ideology--------------------------------------------------------------------------------- 34
Tim e-------------------------------------------------------------------------------------------34
Global Interdependence---------------------------------------------------------------35
Seismic Risk Perception--------------------------------------------------------------------36
Measuring Risk-------------------------------------------------------------------------- 38
Critique of Social and Cultural Theories of R isk-------------------------------- 40
Applicability to Seismic Risk---------------------------------------------------------------41
Notes--------------------- — -------------------------------------------------------------------- 44
CHAPTER 3.TOTAL SEISMIC R IS K -------------------------------------------------- 45
General--------------------------------------------------------------------------------------------45
Components of Total Seismic R isk----------------------------------------------------- 46
Technological R isk---------------------------------------------------------------------- 49
Design Errors------------------------------------------------ 51
Construction Errors---------------------------------------------------------------52
Outdated Codes of Practice (Regulations)------------------------------ 53
Influence of Non-Structural Components-------------------------------- 55
Deterioration due to External Environment------------------------------56
Effects of Previous Seismic Events---------------------------------------- 57
Discovery of New Faults--------------------------------------------------------58
Poor Enforcement of Building Codes--------------------------------------59
Actual Earthquake versus Design Earthquake------------------------ 60
Economic R isk----------------------------------------------------------------------------63
Direct Economic Costs----------------------------------------------------------63
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Damage to Transportation Network---------------------------- 64
Damage to Utilities----------------------------------------------------- 65
Damage to Buildings and Structures----------------------------- 66
Loss of Building Contents and Household Goods 67
Indirect Economic Costs------------------------------------------------------ 68
Business Interruptions----------------------------------------------------69
Loss of W ages-------------------------------------------------------------- 71
Temporary Relocation Expenses-------------------------------------71
Loss of Market Share-------------------------------------------------- 72
Perception of Poor Economy in Future------------------------ 73
Future Stricter Regulatory Environment--------------------------- 74
Relocation of businesses to Other Areas--------------------- 75
Societal R isk-------------------------------------------------------------------------------76
Values and Culture-------------------------------------------------------------- 78
Geography-------------------------------------------------------------------------- 79
Risk Information------------------------------------------------------------------ 79
Interaction within Society---------------------------------------------------- 80
Societal Factors-------------------------------------------------------------------81
Damage to Valuable Physical Assets of the Community — 81
Medical Costs--------------------------------------------------------------- 82
Loss of Life and its Consequences----------------------------------83
Weakened Social Support Systems---------------------------------84
General Non-quantifiable Issues------------------------------------- 84
Notes---------------------------------------------------------------------------------------------- 86
CHAPTER 4:LOSS ESTIMATION METHODOLOGIES AND
ACCEPTABLE SEISMIC R IS K ---------------------------------------- 87
Loss Estimation Methodologies----------------------------------------------------------87
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Probable Maximum Loss (PML)------------------------------------------------ 87
Allstate Research and Planning Center---------------------------------------- 88
EW Blanch Co. Reinsurance Services------------------------------------------89
Applied Technology Council (ATC) Methodology------------------------- 90
Limitation of PML and ATC Methods-------------------------------------------- 91
EQE Engineering Model------------------------------------------------------------- 92
Stanford University Model------------------------------------------------------ 92
HAZUS Methodology-----------------------------------------------------------------93
Limitations of HAZUS---------------------------------------------------------------- 94
Reasons for Community Input----------------------------------------------------------- 96
Acceptable Seismic Risk to a Community---------------------------------------- 97
Definition of Acceptable Seismic R isk------------------------------------------98
Dependency of Acceptable Seismic R isk-------------------------------- 99
Facts and Values---------------------------------------------------------------------101
Variables of Acceptable Seismic R isk-----------------------------------------------102
Interpreting Scientific Data--------------------------------------------------------102
Assessment of Risk-Cost-Benefit Analysis--------------------------------- 106
Group Decision-making------------------------------------------------------------110
Institutionalizing the Process of Change---------------------------------113
Ethical/Moral Dimension-----------------------------------------------------------115
Notes-------------------------------------------------------------------------------------------- 119
CHAPTER 5: ATTRIBUTES OF ACCEPTABLE SEISMIC RISK,
STAKEHOLDERS, THEIR SELECTION AND RESOURCE
CAPACITY OF A COMMUNITY------------------------------------------------------- 121
Attributes of Acceptable Seismic Risk------------------------------------------------121
Life Safety Considerations------------------------------------------------------- 122
Business Interruptions-------------------------------------------------------------123
Infrastructure Damage------------------------------------------------------------- 125
Operability of Critical Facilities--------------------------------------------------126
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Economic Impact-------------------------------------------------------------------- 128
Community Assets------------------------------------------------------------------ 129
Community Preparedness--------------------------------------------------------130
Stakeholder Analysis----------------------------------------------------------------------- 131
Stakeholder Selection---------------------------------------------------------------------- 132
Stakeholder Groups-------------------------------------------------------------------------134
Controllers of Financial Resources--------------------------------------------135
Government-------------------------------------------------------------------- 135
Private Industry----------------------------------------------------------------136
Financial Institutions--------------------------------------------------------- 138
Political Decision -M akers-------------------------------------------------------- 140
Regulatory Agencies--------------------------------------------------------------- 142
Physical Property Owners-------------------------------------------------------- 144
Lifeline Support Systems Providers-------------------------------------------146
Engineering and Scientific Experts--------------------------------------------148
Rescue and Relief Agencies-----------------------------------------------------151
California Seismic Safety Commission-----------------------------------------------153
Federal Emergency Management Agency------------------------------------------154
Resource Capacity of a Community-------------------------------------------------- 156
Economic Capacity---------------------------------------------------------------- 157
Notes--------------------------------------------------------------------------------------------- 160
CHAPTER 6: PROPOSED METHODOLOGY FOR ANALYSIS
THE ANALYTICAL HIERARCHY PROCESS (AH P)-------------------------161
Identification of Experts-------------------------------------------------------------------- 161
Multi-attribute Decision Analysis M ethods------------------------------------------ 163
Elements of a MADA Problem---------------------------------------------------165
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Simplifying Assumptions---------------------------------------------------------- 166
Compensatory Methods----------------------------------------------------------- 167
Weighted Product------------------------------------------------------------ 168
TOPSIS-------------------------------------------------------------------------- 168
Distance from Target------------------------------------------------------ 168
Additive W eighting--------------------------------------------------------- 169
Non-traditional Capital Investment Criteria (NCIC) Method------------------ 169
Specific Requirements-------------------------------------------------------------170
Analytical Hierarchy Process (AHP)---------------------------------------------------171
Cardinal Numerical Scores------------------------------------------------------171
Cardinal Attribute W eights------------------------------------------------------- 171
Contributions to Desirability----------------------------------------------------- 172
Additivity----------------------------------------------------------------------------- 172
Problems with Weighting------------------------------------------------------- 173
Pair-wise Comparisons---------------------------------------------------------- 176
Principal Eigenvector Method-------------------------------------------------- 177
Hierarchy-------------------------------------------------------------------------------177
Numerical and Comparable Attribute Scores-----------------------------178
Normalizing Quantitative Data------------------------------------------- 178
Scoring Using Pair-wise C om parison--------------------------------- 179
Strengths and Limitations of A H P -------------------------------------------- 179
Notes--------------------------------------------------------------------------------------------- 182
CHAPTER 7 : EXPERT OPINION SURVEY RESULTS, ANALYSIS OF
SURVEY AND DISCUSSION OF RESULTS -------------- 183
General------------------------------------------------------------------------------------------ 183
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Findings from Questionnaire 1 ------------------------------------------------------- 185
Findings from Questionnaire 2 ------------------------------------------------------- 190
AHP Model of Acceptable Seismic Risk----------------------------------------- 193
Inconsistency Ratios----------------------------------------------------------- 200
Synthesis of Weig hts----------------------------------------------------------- 201
Acceptable Seismic Risk to Community------------------------------------------ 203
Explanation of Methodology Application-----------------------------------206
CHAPTER 8 : SUMMARY, FINDINGS, CONCLUSIONS
AND IM PLICA TIO NS---------------- 210
Introduction------------------------------------------------------------------------------------210
Discussion of Proposed Methodology----------------------------------------------- 211
Organization of Chapter-------------------------------------------------------------------213
Total and Acceptable Seismic R isk---------------------------------------------------214
Selection Criterion for Stakeholders------------------------------------------------- 216
Identification of Stakeholders----------------------------------------------------------- 216
Attributes of Acceptable Seismic R isk----------------------------------------------- 217
Ranking of Preferences by Stakeholders-------------------------------------------218
Findings-----------------------------------------------------------------------------------------220
Limitations of Methodology-------------------------------------------------------------- 221
Conclusions-------------------------------------------------------------------------- 222
Implications of Proposed Methodology for T he ory----------------------------- 223
Implications for Further Research-----------------------------------------------------224
Implications for Policy for Citizens---------------------------------------------------- 225
Application for Different Levels of Society----------------------------------------- 226
BIBLIOGRAPHY---------------------------------------------------------------------------- 230
APPENDIX A --------------------------------------------------------------------------------- 243
APPENDIX B --------------------------------------------------------------------------------- 247
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Figures
Figure 1-1--------Acceptable Risk Decision Process M odel---------------------- 8
Figure 3-1--------Components of Total Seismic Risk------------------------------ 48
Figure 3 -2------- Interdependence of Risk Components--------------------------48
Figure 3-3------- Technological Risk-----------------------------------------------------62
Figure 3 -4------- Economic R isk---------------------------------------------------------- 64
Figure 3-5------- Societal R isk------------------ 77
Figure 5-1-------- Resource Capacity of a Community----------------------------158
Figure 6-1-------- Hierarchical Structure----------------------------------------------- 174
Figure 6 -2------- Attributes o f Total Acceptable Seismic R isk---------------- 180
Figure 7 -1------- Mandatory Retrofit of Physical facilities-----------------------186
Figure 7 -2 Ranking of Critical Facilities to Remain Operational - 187
Figure 7 -3------- Ranking of Physical Facilities (Minimize Damage) — 188
Figure 7 -4------- Ranking of Community Assets-----------------------------------189
Figure 7 -5------- Acceptable Seismic Risk- Minimize Damage Cost-
Maximize Life Safety---------------------------------- —■ — 194
Figure 7 -6------- Physical Facilities Sub-attributes-------------------------------- 196
Figure 7-7------- Critical Facilities Sub-attributes---------------------------------- 197
Figure 7 -8------- Community Assets Sub-attributes------------------------------ 198
Figure 7-9------- Final Weights of Various Attributes-----------------------------200
Figure 7-10 Derived Priorities with respect to G oal----------------------- 200
Figure 7-11------ Synthesis of Leaf Nodes with respect to G oal-------------202
X
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List o f Tables
Table 1-1------ Deadliest Earthquakes of the Century ----------------------------- 2
Table 5 -1----- Influence of Stakeholder G roup------------------------------------ 134
Table 5 -2 Impact of Stakeholders on Risk Components---------------- 156
Table 6 -1 Verbal Scale for Pairwise Comparison o f----------------------176
Attributes in AHP
Table 7 -1 Importance of Attributes by Respondents--------------------- 190
Table 7 -2 Acceptable Seismic Risk- Minimize Damage C ost 191
-Maximize Life Safety
Table 7 -3----- Attribute Definitions----------------------------------------------------- 195
Table 7 -4----- Ranking Attributes of Physical Facilities------------------------ 197
Table 7 -5----- Ranking Community A ssets----------------------------------------- 198
Table 7 -6----- Ranking Attributes of Acceptable Seismic Risk----------- 199
Table 7 -7----- Synthesis of Leaf Nodes with respect to Goal--------------- 203
Table 8 -1----- Attributes of Seismic R isk---------------------------------------------229
xi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Abstract
Earthquake events are natural hazards that are beyond the control of human
beings. In that regard they are different from manmade hazards and other
natural hazards. Major earthquake events are low probability, high
consequence occurrences. Throughout the world these events have resulted
in death tolls numbering in thousands and have caused economic losses of
billions of dollars.
Significant recent earthquake events in California have demonstrated that
indirect economic losses far exceed direct economic losses. Although, the
federal government has aided communities in the past, the focus of funding for
disaster assistance has recently changed to promoting the concept of
partnership between local governments, residents and businesses.
Communities have not dealt with natural hazard mitigation measures
proactively in the past.
This study focuses on defining, determining and measuring acceptable
seismic risk to a community through community based input. The study has six
parts: social and cultural theories of risk, total seismic risk, acceptable seismic
risk, stakeholder identification, preferences of experts as proxies for
stakeholders and analysis of preferences through an Analytical Hierarchy
Process (AHP).
xii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Social and cultural theories o f risk are found to be applicable to seismic risk.
Total seismic risk is comprised of technological, economic and societal
components. Due to limited resource capacity acceptable seismic risk to a
community is less than the full impact of a seismic hazard. The impact can
also be modified by stakeholders’ actions.
Stakeholders are identified and their preferences on attributes of acceptable
seismic risk are sought through questionnaires. Expert opinions as proxies for
stakeholder preferences are utilized. Final weights for attributes are
determined using the Analytical Hierarchy Process (AHP) methodology.
Qualitative statements are converted into quantitative data using a conversion
scale used in AHP.
It is posited that resource allocations can be made based upon final weights of
attributes. The methodology developed is intended to be applicable to different
levels of government. It is believed that input from the community leads to
better decision-making in addressing public policy and resource allocations
and promotes personal voluntary actions to mitigate seismic risk.
xiii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
INTRODUCTION
Natural hazards are different than man-made hazards; their occurrences are
considered acts of nature beyond the control of human beings and therefore
find more societal acceptability than man-made hazards even though their
impacts may be modified by human actions. An earthquake event is one such
natural hazard.
Since the beginning of this century, major earthquake events throughout the
world have resulted in a death toll numbering in thousands. (Over 5,500 in the
1995 Kobe, Japan earthquake to over 650,000 in the 1976 earthquake in
Changshan, China). Fortunately, even the strongest U.S. earthquake in San
Francisco area in 1906 resulted in a death toll measured only in hundreds.
See Table 1-1.
Seismologists and Geo-scientists state that major earthquakes (magnitude 7
or greater on the Richter1 scale) are supposed to occur infrequently; e.g. in
California a major earthquake is defined as an event with a return period of
475 years, which translates into an annual probability of approximately 0.002.
Although major seismic events have a low probability of occurrence, they
result in high consequences.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Deadliest Earthquakes of the Century Table 1 -1
Thangshan, China 1976 665,000
Guatemala 1976 234,000
Tokyo and Yokohama, Japan 1923 200,000
Gansu Province, China 1920 100,000
Messina, Italy 1908 83,000
Nanchang, China 1927 80,000
Gansu Province, China 1932 77,000
Yungay, Peru 1970 68,800
Armenia 1971 60,000
Chimbote, Peru 1971 52,000
Northwestern Iran 1990 50,000
Armenia 1988 22,000
Ixmit, Turkey 1999 20,000+
Mexico City, Mexico 1985 18,000
Kobe, Japan 1995 5,500
Central Region of Taiwan 1999 2,400+
Maior California Earthauakes
San Francisco 1906 700 (approx.)
Loma Prieta 1989 63
Northridge
1994 57
Between 1900 and 1995, a total of 194 earthquakes of Richter magnitude 5.5
or greater occurred within or near California according to a study conducted by
the California Division of Mines and Geology. Out of these, only 27 were
above magnitude 6.6, but they covered a far greater area than that covered by
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
those below magnitude 6.6. Recent significant seismic events in California
have also occurred more frequently than seismologists predicted (San
Fernando, 1971; Coalinga,1983; Loma Prieta,1989; Northridge,1994). These
are not considered major seismic events because they did not reach
magnitude 7 on the Richter scale except for Loma Prieta which reached a
magnitude of 7.2. In these seismic events, the casualties measured in terms of
death, have been low; 57 people died in the Northridge event in Southern
California, and 63 people died in the Loma Prieta event.
It is important to note that these seismic events caused considerable
economic damage even though the death casualties were low and the
magnitude of these seismic events did not classify them as major events.
Estimated damages have ranged from $8.6 billion for the Loma Prieta seismic
event in Northern California to $25 billion for the Northridge seismic event in
Southern California. Studies conducted for the Insurance Services Office by
Friedman (1988) report that a scenario earthquake of magnitude 7.5 on the
modified Mercalli2 scale on the Newport-lnglewood fault in Southern
California would result in insurance payments of $14.6 billion (most likely
estimate). The fires following this earthquake would generate another $24
billion in losses (Scawthorn,1987, p.2). The Kobe, Japan earthquake resulted
in economic losses exceeding $200 billion. (City of Kobe Report, Feb. 2000)
3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Recent significant earthquake events, such as Loma Prieta and Northridge
have also demonstrated that indirect long-term economic losses far exceed
direct economic losses. Research studies suggest the ratio of indirect
economic losses to direct economic losses to be between 2 and 4. (Eguchi
and Seligson, 1993)
The Federal government, through its disaster assistance policy as
implemented by the Federal Emergency Management Agency (FEMA), has
provided funding for recovery efforts after damaging earthquake events in the
past. However, the costs of damages as a result of recent earthquakes and
other natural disasters have increased significantly.
A study by Petak and Atkisson (1982) on the impact of nine natural hazards
including earthquakes suggests that the exposure of people and property to
the nine natural hazards is estimated to be about $38 billion in the year 2000.
In the winter of 1811 & 1812 when three earthquakes figured to be 8.6, 8.4,
8.7 on the Richter scale struck near New Madrid, Mo. the trembler caused
damage in an area 15 times greater than that of the comparably sized 1906
San Francisco earthquake. A repeat of 1811-12 events would cause damage
estimated at more than $51 b. in 1990 dollars because the Eastern United
States is much more vulnerable to earthquake damage than the West Coast.
Buildings are not designed to resist earthquakes; it is more densely populated
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and businesses and individuals are not prepared for tremblors. (McCann,1990,
p.1-2)
Therefore, FEMA is promoting the public-private partnership concept as a way
to reduce its obligations and to shift a part of its financial burden to
stakeholders, i.e. local communities and private industry. “Project Impact is a
local partnership among government, residents and businesses. In one
hundred communities across America, businesses, citizens are taking
action to reduce the potential damage from disasters and create a safer
future.” (Witt, Director of FEMA, Aug.25, 1999)
It is clear that communities in the future will be forced to participate in paying
recovery costs along with the state and federal governments. Communities will
have to decide on the extent of seismic risk that is economically and socially
affordable and thus acceptable to them. Communities will also have to build
capacity to absorb higher levels of seismic risk while reducing the impact of
seismic hazards on the community.
In a community, private citizens and business entities may choose to take
actions to reduce seismic risk to protect their investments and well being.
Such actions fall in the category of reducing private risks. Community seismic
risk on the other hand is related to societal and economic aspects of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
community. Although private risks and public risks impact each other, this
study focuses on public aspect of seismic public risk.
The purpose of this study is to develop a methodology for measuring and
determining acceptable seismic risk to varied communities in the American
context of civic and governmental structure. It is hoped that such measurability
will lead to more informed public policy decisions related to seismic hazard
risk.
Background
Major seismic events have a low probability (as noted earlier, in California the
annual probability is approx. 0.002) of occurrence, but they result in high
consequences. Damages, resulting from major seismic events, are currently
estimated in economic terms by aggregating damages to physical facilities,
utilities and infrastructure. Current loss estimation methodologies do not take
into consideration indirect economic losses or long- term economic or societal
effects due to various uncertainties and lack of measurability. (Brookshire et
al. 1997)
Due to incomplete damage assessments total seismic risk, at present, is
underestimated. The true cost to the society is not known for several years
after an event. It is suggested that total seismic risk be comprised of three
major components; technological, economic and societal. Unless total seismic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
risk is considered, the information on risk is not complete; and therefore,
potential total damage to a community is not estimated. Such incomplete
information, compounded by the low probability of occurrence of major
earthquakes, results in poor seismic hazard mitigation public policy and leads
to misdirection and misallocation of limited public resources. Poor and
incomplete information also weakens the case for seeking adequate monetary
resources for seismic hazard risk mitigation as a preventative measure.
A community is faced with both the potential impact of the seismic hazard and
the capability of the community to deal with total seismic hazard. Although the
magnitude and frequency of occurrence of a major seismic event cannot be
controlled by a community, its effects can be modified by actions taken by the
community. Two communities facing the same seismic hazard may undertake
different actions resulting in different consequences from the seismic event. As
a result, total seismic risk may be different for the two communities.
It is proposed in this study that risk is a social and cultural construct, and the
assessment of total seismic risk needs to have community input. “As people
begin to assess and manage risks at the local level, they will be preparing
themselves to cope as citizens of a democratic society moving into a future
dominated by barely imaginable technologies and fraught with unfamiliar risk”
(Ruckelhaus, W.D., 1985). Subsequent to assessing total seismic risk,
acceptable seismic risk to a community needs to be determined by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
stakeholders in the community. Since a major seismic event causes
considerable damage economically and socially to a community as a whole, a
decision on acceptable seismic risk involves many players at different levels of
a community. Acceptable seismic risk is also a function of the total resource
capacity of a community, which may be defined as the summation of economic
and human resources of a community.
It is further proposed that to assess the technological risk component, input
from engineers and scientists should be sought. The “standing” of various
stakeholders in the community ought to be decided so that their opinions on
various components of seismic risk can be considered. Conceptually, a
decision process model for acceptable seismic risk to a community is shown in
Figure 1-1.
Stakeholders
Total Seismic Risk Acceptable Seismic
Seismic
— — to a
------------1 -------^
Risk to a
Community Community
Resource Capacity
of
Community
Acceptable Risk Decision Process
Figure I -1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Research Questions
How should acceptable seismic risk to a community be defined? Who are the
stakeholders in defining acceptable seismic risk, and how should they be
selected? What are the various attributes of seismic risk? How can the
quantitative and qualitative attributes of seismic risk be combined to obtain a
single measure of acceptable seismic risk? Can the proposed methodology be
generalized to determine acceptable seismic risk at different levels of society?
This study attempts to address these and related questions, and proposes that
the methodology developed can be utilized in making more informed public
policy decisions related to seismic hazard risk.
Proposed Methodology
Various attributes of seismic risk to a community are identified. These
attributes range from life safety considerations of building occupants to
preservation of the valuable assets of a community. Each attribute has either
a defined economic value or can be judged worthwhile to be preserved by the
community as a whole. These attributes fall into the three main components of
total seismic risk: technological, economic, and societal. Input from
stakeholders is sought on each of the attributes.
To identify stakeholders, it is necessary to consider the economic base and
social composition of a community. For this study, it is assumed that a
9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
representative community is comprised of businesses, financial institutions,
physical property owners, political decision-makers and community
organizations dealing with emergency preparedness. Individuals representing
these groups, in addition to technical experts and code regulators, are deemed
to be the stakeholders for the representative community.
Selection of stakeholders is based upon their expertise in various components
of risk. Selection criteria are explained further in Chapter 5. Stakeholders who
met the criteria were selected.
Stakeholder input was sought through carefully designed questionnaires.
Experts representing each of the areas are used as a proxy for stakeholders;
they were asked to rank various attributes of seismic risk in a pair-wise
comparison. Survey response scores to each question were averaged and
analyzed using Multi-Attribute Decision Analysis (MADA)3 methodology; i.e.
each attribute is compared to the other. Specifically, the Analytical Hierarchy
Process method (AHP) which is one of the MADA methods is used to
construct the matrix of pairwise comparisons. In the AHP, the scale used
ranges from 1 (equally important) to 9 (extremely more important). The matrix
of pairwise comparisons allows determination of which of the two attributes in
the pair is more important.
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To convert paired comparison data into attribute weights, a particular
technique known as the principal eigenvector method is used. A computer
program is used for this purpose. All non-quantitative attribute scores are
converted into quantitative scores using a scale of 1 to 9. If an attribute has
sub-attributes, similar methodology to analyze sub-attributes is used.
The final score is generated based on the derived attribute weights. This final
score, by itself, probably is not sufficient to divulge much information to public
policy officials and should not be considered a measure of quantity of
acceptable seismic risk. Better information for public policy resource allocation
decision-making is available by dis-aggregating the final score. The weights of
individual attributes can provide useful information for the most effective use of
resources. However, attributes can be classified as “cost” attributes and
“benefit” attributes. For “cost” attributes, the lowest possible score (close to
zero) is desirable. On the other hand, if the attributes are considered to be
“benefit” attributes, then the highest possible score (maximum 1.0) is
desirable.
This method allows a community to focus on various attributes contributing to
the final score. A community can then focus on those attributes that can
minimize its costs or maximize its benefits.
l i
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Organization o f the Study
Chapter 1 discusses concepts of risk in general, risk definitions, natural
hazards, and the nature of seismic risk, in particular.
Chapter 2 focuses on social and cultural theories of risk, risk perception,
probabilities, risk presentation, risk measurement, and the limits and
applicability of social and cultural theories to seismic risk.
Chapter 3 discusses total seismic risk, components of total seismic risk,
factors affecting the components of risk, models for each component of risk,
and a model for total seismic risk.
Chapter 4 proposes a definition and discusses the concept of acceptable risk
to a community, discusses fact and value dilemmas, group decision-making,
and cost-benefit analysis. This chapter discusses current loss estimation
methodologies and their limitations.
Chapter 5 discusses the attributes of acceptable seismic risk, the
stakeholders of the community, criteria for their selection, and resource
capacity of a community.
12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 6 discusses identification of experts, expert opinions, the proposed
multi attribute analysis methodology, the analytical hierarchy process, and
strengths and limitations of that process.
Chapter 7 discusses expert opinion, analysis of the surveys, final scores of
attributes for the community, data analysis, and summary of findings. This
chapter also proposes a theoretical base for assessing acceptable seismic risk
to a community.
Chapter 8 discusses the application of the proposed methodology for making
public policy and allocating resources for seismic hazard mitigation,
implications for theory and for further research. This chapter also discusses
how the methodology can be used at different levels of governments to make
informed public policy decisions.
In Chapter 1 which follows, the study discusses general concepts of risk, risk
definitions, natural hazard risk, and the nature of seismic risk.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Notes
1 Charles Richter developed this scale in 1935, which measures earthquake intensity quantitatively.
Richter defined the magnitude o f a local earthquake based upon seismic wave amplitude recorded on a
specific seismograph located at a distance of 100 km from the earthquake epicenter. Since the
measurement is on a logarithmic scale with a base of 10, each increase in scale magnitude represents
earthquake wave magnitude increase 10 times.
2 Italian seismologist Mercalli developed the scale in 1902 to measure earthquake intensity. The scale
has a range of twelve degrees, from I to XII. It is a descriptive scale and based upon the observed
damage, an earthquake event magnitude can be estimated. It was modified in 1931 by Wood and
Neumann to fit the conditions in California. Since 1931, in the U.S. it is known as the modified
Mercalli intensity scale.
3 Multi-Attribute Decision Analysis (MADA) methods consider non-financial attributes expressed
qualitatively and quantitatively. Analytical Hierarchy Process (AHP) is a specific MADA method.
AHP includes an efficient attribute weighting process of pairwise comparisons. The use of AHP is
facilitated by available computer software.
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1
GENERAL CONCEPTS OF RISK, RISK DEFINITIONS, NATURAL HAZARD
RISK, AND NATURE OF SEISMIC RISK
Risk - General
Nearly all activities undertaken by human beings since the beginning of
civilized existence have contained some form of risk or uncertainty. The only
certainty in life is death; people usually strive to live life to the fullest and to
improve their quality of life, till the onset of death. Threats to achieve these
objectives involve risks, whether natural or manmade. Due to extent and scale
of human activities in the ever-expanding industrial society since World War II,
the array of risks has increased sharply. Attention paid to technological
hazards and media coverage are changing the public’s perceptions of risks
(Rowe, 1977). The need to focus on those problems in a regular and
consistent manner started the formal analysis of risk as well as efforts to
address the more subjective problem of acceptable risk.
Decision theories, systems analysis and operations research have focused on
analytical techniques to estimate risk, but these methods have been relatively
unsuccessful in the subjective area of risk acceptance. Since risk involves
people, analysis of risk is as complex as individual, group and societal
behavior at any point in time. Personal risk is an acknowledged part of life.
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Everyone reacts differently to risks they take voluntarily and to risks that are
imposed by someone outside the group that produces involuntary exposure.
Group decisions by other people imposing involuntary risk on others are
common (Starr, 1980). The value system of a society and its behavior towards
risk is based upon changing information, therefore, risk assessment should not
be considered fixed in time.
Assessing risk is a subjective phenomenon. Even though scientific data is
considered to be objective, value free interpretation of the data and
hypotheses built on them include subjectivity. Rowe (1977) argues that
subjective perception is the basis for risk acceptance, regardless of objective
or quantified evaluation. Lowrance (1980, p.5) states that “risk is taken to
mean probability of harm (objectively ascertainable), and safety is a social and
legal judgment that probability is low.“ However, some others have argued that
assignment of probabilities reflects the assignees belief or confidence that the
event in question will occur. The subjective view thus blurs the difference
between risk and safety since probability and judgment are intertwined and not
separate. (Lee, 1976)
Risk may be expressed in many ways: number of lives lost per year, average
shortening of life span, cost of annual damages or frequency of heavy snow
storms per year. Once the size of risk is estimated, decisions have to be made
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
whether to bear that risk. Actions require personal and social decisions that
consider many normative factors. (Lowrance,1980)
In this study, it is expressly acknowledged that decisions on acceptable risk
are subjective; and, since the community as a whole makes decisions, they
reflect the value judgments of the community at a particular point in time. This
study does not attempt to discuss risks taken by individuals or private entities
in their business operations. The focus of the study is on public risk as it
relates to a community.
In addressing the question of an acceptable level of risk, a special task force
for the Office of Science and Technology came to the conclusion in 1978 that
acceptable risk is not value free.1
Risk Definitions
Defining risk is not simple because perception of risk and its meaning are
different for different people, organizations and institutions. In analysis of risk,
Fischoff, Watson and Hope (1984) state that “no definition is advanced as the
correct one, because there is no one definition that is suitable for all
problems”. Lowrance (1980,p.6) defines risk as “a compound measure o f the
probability and magnitude o f adverse effect.” The meaning of risk is also time
dependent and influenced by culture. The Webster dictionary definition of risk
is “the possibility o f loss, injury, disadvantage or destruction.” Other proposed
definitions are; “probability o f suffering potential harm o r undesirable
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
consequences from a hazard ” (Cohrssen and Covello.1989, p.7) and “as the
measure o f hazard, to which people and what they value are exposed.” (Hertz
and Thomas, 1984, p.11)
The insurance industry, on the other hand, defines risk in terms of probable
maximum loss, i.e. estimate of maximum loss that would occur if the event
happened in the period under policy coverage; and possible maximum loss;
i.e. largest loss that could occur, if the event happened. The insurance
industry considers risk only in terms of potential financial liability to settle
claims. Risk can also be defined as a process of combining present
knowledge about the future with public choices among desired alternatives.
(Douglas and Wildavsky,1982, p.5)
Risk definition is also dependent on one’s own utility function (Taversky and
Kahneman, 1974), e.g. if a businessman’s utility function includes bankruptcy,
he would tend to be risk averse. Psychological findings suggest that
individuals perceive risk in terms of probability and amount of loss, rather than
the probability of the event causing the loss (Slovic et al.,1980). Risk is both a
descriptive and a normative concept. It includes the analysis of cause-effect
relationships which may be scientific, anecdotal, religious or magical; but it
also carries the implicit message to reduce undesirable effects through
appropriate modifications of the causes, or through less desirable mitigation of
the consequences (Renn,1992). The differences in these definitions stem
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
from one’s position, perceptions and potential for economic or other losses. To
paraphrase Miles’ Law: You stand where you sit.
The risk universe of an individual goes beyond the financial aspect and
includes non-quantifiable effects, e.g. the loss of irreplaceable items and
disruptions to personal life. A business entity would assess risk not only in
financial terms of business losses, but also potential legal liabilities arising out
of delayed delivery of products and services.
Because varying definitions of risk result in different methods and levels of
protection and consequently different outcomes from mitigation measures, it is
important to agree on one definition o f risk. An agreed upon definition of risk
will result in clarity of direction for public policy and hazard risk mitigation
measures. (Fischoff etal., 1984)
It is recognized that there are different definitions of risk. In this study, it is
simply acknowledged that risk contains uncertainty and has the probability of
causing some degree of loss, either to property or life. The magnitude of risk
varies with each situation. Taking action to minimize the negative
consequences of risk is the goal.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Natural Hazard Risk
In the United States, natural hazards are chiefly earthquakes, floods,
hurricanes, and tornados. These are considered Acts of God. People may
understand their mechanisms and effects, but experts are still largely unable
to predict the frequency of their occurrence. Although some flood control
measures have been successful, in general, the best that societies can do is
to take measures to mitigate their effects. These events can also be classified
as those with a low probability of occurrence, but resulting in high
consequences. Lowrance includes failure of large technological systems such
as dams and power plants in this category, but these are not natural hazards.
If natural hazards have a statistically valid pattern of repetition, then
qualification of their effects could yield useful results and be utilized in
designing mitigation strategies. Few risks, though, fall into this category; and, if
the data on occurrences is insufficient, calculations of effects can only be
partly accurate as they may be event and site specific. Generalizations that
may be difficult to quantify include the long-term impacts on the health of
individuals, businesses and economies of regions. Traumatic events, such as
major earthquakes, affect people's lives on a long-term basis and can be
minimized or amplified depending upon the cultural context.
Major earthquakes can produce serious psychological reactions in people
experiencing the shaking and seeing the damage. The emotional reaction is
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
more severe in young people and children. Experience indicates that it is
worse in those who lose loved ones or whose property is damaged or in
rescue workers who see the destruction. These individuals may suffer post-
traumatic stress disorder in the months after the event. Following the 1988
Armenia earthquake, 50-60% of the citizens suffered memory loss, had much
shorter attention spans and were much less productive at work. Following the
October 17, 1989 Loma Prieta earthquake event in San Francisco Bay Area,
nearly 60% reported in a survey that they had emotional difficulties. While few
had property damage, nearly all said their emotional well being was affected.
(McCann, 1990, p.2)
Natural hazard events pose a different kind of risk than man made hazards.
Since natural hazards are thought of as acts o f nature beyond our control;
their frequency of occurrence and magnitude cannot be controlled; and
therefore, the public has no choice but to accept these events although their
impacts may be modified by actions taken by the public.
Nature of Seismic Risk
Earthquakes long have been feared as one of the most terrifying phenomena
of nature. According to plate tectonic theory, the earth is divided into a series
of plates, which continually interact with each other. Movement of tectonic
plates against each other creates forces at interfaces, causing earthquakes.
The magnitude of an earthquake event depends upon the rate of movement of
21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
tectonic plates and the degree of interlock between them. Major earthquakes,
which are defined as having a magnitude 7 or higher on the Richter scale, are
expected to have long time periods elapse between their occurrence.
However, the frequency of damaging earthquakes in California has increased
lately. (San Fernando 1971, Whittier 1985, Loma Prieta 1989, and Northridge
1994)
Major earthquakes are a threat to life and property. Public and private property
can disappear within a matter of seconds; and jobs, services, and business
revenues can suffer. For many, homelessness may become a reality. It also
takes a long time to recover from the economic and social losses of such
devastating events (FEMA 1983). Except for the Loma Prieta event, other
recent urban earthquake events in California did not register a magnitude 7 on
the Richter scale and, therefore, are not considered major earthquakes. Yet,
these recent earthquakes caused heavy damage, resulting in significant direct
economic losses as well as significant indirect economic losses.
Communities, on the other hand, face seismic risk as an added burden in
providing emergency social services, at significant cost to the community, and
potentially increased debt burden to repair damage to the transportation
network, communication infrastructure and utilities. Included in seismic risk to
the community are also indirect effects resulting from damage to the
transportation network, communication infrastructure, and utilities.
22
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The uncertainties in assessing seismic risk result from many factors:
earthquake fault location, patterns of shaking, site effects, building damage,
fault rupture, effects of collateral hazards (such as liquefaction and land
slides), and other loss conditions. As a result of uncertainties in loss
estimation, earthquake insurance coverage is faced with the intransigent
problem of assessing potential losses in order to reserve sufficient funds to
pay earthquake claims when a damaging event occurs.
Seismic Regulations in California
Seismic hazards and seismic safety considerations in California are governed
by various state laws and are generally implemented and enforced by local
governments. Implementation of state seismic safety policy is accomplished
within five major areas:
1. Identification of seismic hazards and mitigation elements of a community’s
general plan;
2. Identification of seismic hazards and mitigation pursuant to the California
Environmental Quality Act (CEQA);
3. The use of fault rupture hazard zones ( Alquist/Priolo zones) identified by
the state geologist;
4. Application of the Uniform Building Code (UBC) or a locally adopted
building code; and
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5. Identification of potentially hazardous buildings in specified areas of the
state and establishment of a mitigation program.
Since 1970, state law requires that seismic safety be included in a
community’s general plan. Safety and seismic safety were consolidated in
1984, by chapter 1009 of the government code. The Governor’s Office of
Planning and Research revealed that all 58 counties and 98% of cities have
completed a safety element as of 1988. However, it is not clear as to what
constitutes a safety element in their plans; thus the inclusion of the safety
element and its quality in a community’s general plan is questionable.
Public information on seismic hazards in California is provided by three
different agencies; the California Seismic Safety Commission, the Division of
Mines and Geology and the Governor’s Office of Emergency Services. There
is no question that risks from natural and technological hazards pose serious
difficulties for state, federal and local levels of government as well as for
individual businesses and households. (Petak, 1993, p.1)
In Chapter 2, which follows, the study discusses social and cultural theories of
risk, risk perceptions, understanding probabilities, risk presentation, and risk
measurement. The chapter also discusses limits and applicability of social
and cultural theories of risk to seismic risk.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Notes
1 All social goals incorporate values that must be weighed against costs of achieving various objectives.
Several factors must therefore be considered in defining “ acceptable risks ” or “residual risks ” to life
and property in relation to the costs and outlays required. There is no uniform level of acceptable risk.
Acceptable safety levels vary with time, place and circumstances; they must be related to costs; and
they are influenced by cultural and economic factors as well as the subjective feelings and emotional
reactions of policy makers.
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2
SOCIAL AND CULTURAL THEORIES OF RISK
General
As discussed in the previous chapter, the range of definitions of risk is broad,
and perceptions of risk by the public play a significant part in its definition.
Since the public’s perception is socially and culturally based, many
researchers have argued that risk is a social and cultural construct.
Renn (1992) notes that people determine the character and magnitude of risks
after having filtered information received from family, friends and coworkers.
Douglas and Wildavsky (1982, p.194), in their original work on selection of risk
by the public, point out that societal selection of risk is a function of social and
cultural values rather than scientific evidence. These authors go on to argue
that social and cultural values also determine risk agendas selected for
attention; e.g. certain risks are discussed soon after an event causing a
significant loss has occurred because pressures from social criticism demand
such action. The Field Act requiring school facilities to meet a higher structural
safety standard to withstand a major seismic event was enacted within one
month of the 1933 Long Beach earthquake in which many school facilities
were damaged beyond repair.
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Wildavsky and Thompson (1982,p.8) have demonstrated that social influences
shape the behavior of individuals and their response to risk. These authors
employ a technique called grid-group analysis and use two indices. The first
index measures the group-index (an individual’s identification with the group)
and the other measures the grid-index (to what degree an individual accepts a
system of rules and hierarchy). These authors have identified four types of
cultural behavioral patterns based on this grid-group analysis; those patterns
then can be evaluated in the context of organizations or institutions. These
patterns of behavior are: Competitive Individualist, Egalitarian, Bureaucrat,
and Stratified.
Risk Perception
According to Douglas and Wildavsky (1982), modern Americans perceive the
world to be an inhospitable and frightening place. Americans are afraid of
“Nothing much ... except the food they eat, the water they drink, the air they
breathe, the land they live on, and the energy they use” (1982,p.10). A survey
conducted by Louis Harris and Associates in 1980 indicated that nearly 80% of
all Americans agree that “people are subject to more risk today than they were
20 years ago.” (1982,p.11)
One of the most important aspects in risk perception is that people take into
account a large number of factors evaluating the seriousness of a risk (Slovic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
et al.1980; Covello, 1984). Some of the important factors people consider are:
catastrophic potential, familiarity, voluntariness and dread.
Douglas and Wildavsky further propose that people’s choices about attention
to risk are influenced by values, social institutions, nature and moral
behaviors. These authors state that it is essential for people to select risks for
attention as it is impossible for people to be aware of the thousands of risks
that are posed every day. Even its critics acknowledge the work by Douglas
and Wildavsky as a major contribution to the understanding of risk through the
role of social factors.
Another contemporary research effort has been in the area of selection of risk
through cognitive limitation of risk information processing (Slovic et al, 1980;
Tversky and Kahneman, 1984; Covello, 1984). A more comprehensive view of
risk selection by individuals can be presented by integrating social and cultural
aspects with cognitive psychology research. Psychological research studies
on risk point out four areas of risk perception: dramatic cause, probabilities,
risk presentation, and values.
Dramatic Cause
People overestimate the risks of dramatic causes of death, such as airplane
accidents; such overestimates are partly due to the greater memorability and
imaginability of such events (Fischoff, Slovic, et al.,1978; Covello and
28
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Johnson, 1987). Intense media coverage also distorts risk perception.
(Ravetz,1979)
Social networks and their ability to counteract or further reinforce
psychological factors significantly alter perceptions of risk (Fitchen et
al.,1987). in the absence of social networks or their support, the public
perceives risk higher than warranted.
Probabilities
Probabilistic thinking appears to be rather simple and logical. Understanding
probabilities requires knowledge of three basic principles: randomness,
statistical independence, and sampling variability.
Lay persons often have difficulties in understanding and interpreting
probabilistic information, especially when probabilities are small (Slovic et
al.,1980, Covello, 1984); e.g. an earthquake with a 500 year recurrence interval
period has an annual probability of 0.002 which is deemed insignificant by
laypersons. Small numbers carry with them the sense and feeling of
insignificance. In their daily work, sailors, fishermen and farmers use
probabilities informally. It is not the informal practice of probabilistic thinking
but the use of formal probability analysis that poses difficulties for people in
understanding their risk.
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Although, probabilistic thinking could be left to experts, its translation for the
public leaves a ‘gap’ between an expert’s judgment and a lay person’s
understanding. This gap, according to Shogren (1990), creates a dilemma for
policy makers. Explanation of seismic risk to laypersons ought to be on a
basis of comparing seismic risk with other routine risks with which they are
familiar. Shogren suggests that government can play an important role by
providing risk information to the public and reducing gaps in its perception.
Risk Presentation
Another influencing factor on risk perception is its presentation. Whether a
risk is presented as a cost or a benefit, influences risk selection, perception
and concerns (Tversky and Kahneman 1984). The very concept of cost and
benefit is value driven, whether economic or social, and is further impacted by
social and cultural factors. Powerful vested economic interests could increase
or decrease the extent of risk perception; e.g. in a community an owner of
several investment properties could decrease the perception of risk to his
properties by claiming that his properties do not need upgrades to meet the
current seismic standards. Imposition of requirements to upgrade the
properties would obviously result in significant costs. In fact, an inherent risk in
his properties could be present, compromising the safety of tenants.
Trustworthiness and credibility of institutions managing risk also play a
significant part in reducing or exacerbating concerns about risks. (Fitchen et
al.,1987)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Values
Research is still incomplete on determining the influence of value systems on
risk perception. It is not only the individual’s value system, but also the
interaction with the collective value system of a society which makes it even
more complicated. Stallen (1980) has developed four responses by society to
technological risks: the secure, the defensive, the vigilant and the adaptive,
but a general theory of collective risk perception is yet to emerge. Douglas
has argued that the problem lies in the assumption that a group preference
could be developed as an aggregate of individual preferences and suggests
that the study of the relationship between social processes and shared values
may be more useful. Another context in which psychological factors operate is
the environment.
Environment
Perceptions of risk are also contextual and are influenced by the environment
in which people operate. The contextual factors include communities,
organizations, institutions, ideology, time, and global interdependence.
Communities
Social science research indicates that risk perceptions are influenced by social
interactions between members of a community and the level of satisfaction
among its members (Kielcott and Nigg, 1982). The greater the satisfaction, the
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
lower the perception of risk. Closely-knit communities also perceive risks at a
lower level than those which have cultural, social, and religious tensions.
Organizations
Organizations view risks differently than individuals. Organizational cultures
and utility functions are commonly different than those of individuals. The
perception of risk also varies with different stages in the life cycle of an
organization. The goals of organizations change over time; organizations
select different risks for attention at varied stages in their life cycles.
Organizations also control the amount of information on risk to be shared with
employees. Internal factors such as the duration of risk; i.e. long term or short
term, and abilities of employees to comprehend and act upon risk related
information determine management’s attitudes towards sharing risk
information with their employees.
External factors which influence management decisions on sharing risk
information include activism by union officials, State and Federal laws and
regulations. The current system does not offer sufficient incentives to
managers to communicate risks fully and effectively with workers. Even when
the information is fully shared, the response to risk by employees is
substantially less than optimal. Employee responses vary from denial to
fatalistic acceptance or activism. (Brown, 1987)
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Institutions
Institutional responses to risk can be quite orthodox and bureaucratic. Public
institutions, particularly, operate on a routinized basis and have numerous
rules, laws, procedures and regulations. Public institutions perceive risk on an
operational basis and are slow to respond to hazards as lines of
communications are not clear. Generally, with several layers of hierarchy,
responsibilities are distributed vertically and laterally, leaving no one in charge.
In case of catastrophes, the emergency services have generally developed a
plan of action and are able to respond quickly. This is particularly true of
police services, fire services, and other rescue agencies.
Relationships between local, state and federal institutions also play a major
role. Particularly when national resources are shared, headquarters officials
may be in charge. Local officials, who deal with the risk directly, may be the
last ones to hear about risk management plans. In the1995 Kobe earthquake,
the local authorities waited as long as forty-eight hours before taking action as
they awaited directions from national authorities in Tokyo. Such a delayed
response could perhaps be attributed to cultural behavioral patterns in
Japanese society. However, it also demonstrates a rigid hierarchical structure
can delay a timely response. Local officials also have to deal with numerous
federal government departments and agencies and various personnel which
creates tensions. The perceptions of risk by local and state officials are also
different from federal officials, due to different frames of reference.
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ideology
Social movements based on ideology can change societal perceptions of risk.
Numerous examples can be cited: ecology movements, no growth
movements, baby infant formula movements and so on. (Kielcott and
Nigg,1982, p.131)
When individuals decide to organize themselves based on a perceived risk,
they usually change the perception of risk by bringing focus to the issue
through network processes, political savvy and economic might. Sometimes
the causes behind organized protests are purely ideological and are based
upon perceptions; at other times protests are based upon harmful effects of
which the general public may not be aware. Recent lawsuits against tobacco
companies and damages to settle claims in hundreds of billions of dollars are
examples of pursuit based on harmful effects of tobacco and deliberate
misuse of data by tobacco companies for profit.
Time
Risks change with time and therefore they need to be considered in a time
context. As the values in a society change, so do the acceptability of risk and
its perception. The choices or alternatives available to a society at a particular
time determine its risk perception.
Society also determines the costs of risks in a time context. Economic
development and pace of change, which are both time related, also determine
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
risk perception in a country; e.g. if the change in a society takes place slowly,
risks may be absorbed better than when a society has to respond to radical
changes. The time context in risk perception is extremely important to the
analytical process of risk assessment.
Global Interdependence
Today, most national economies in the world are mutually interdependent and
are expected to be even more interdependent in the future. This scenario
changes the distribution of the economic component of risk; e.g. many U.S.
multinational companies manufacture their products in Osaka, Japan. When
the Kobe earthquake occurred in 1995, its seaport facilities were damaged
and the port was closed for several months. U.S. multinational firms could not
get their products to U.S. markets for months; their businesses were put at a
financial loss and impacted consumers in the United States (Mujumdar,1996).
Similar situations occurred during Taiwan’s earthquake in 1999. Taiwan is a
major manufacturing base for computer chips. Many U.S. companies have
computer chip manufacturing operations in Taiwan. Due to damage to
facilities, the production suffered impacting these businesses in the U.S.
(Sacramento Bee, Sept. 30, 1999). Such mutual interdependence forces one
to consider the global economic and business aspects in overall risk
perception.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Seism ic Risk Perception
Several misconceptions of seismic risk exist in the United States, (FEMA
1983) such as:
1. Earthquakes occur only in few places in the United States, primarily in
Alaska and California. In fact, more than 40 of the 50 states as well as
many United States Territories are at risk from earthquakes.
2. It is unlikely that most Americans will experience a large, damaging
earthquake in their lifetime. While earthquakes occur in “geological time,”
which is far longer than people’s lifetimes, records show that some seismic
zones in the United States experience moderate to major earthquakes
about every 50 to 70 years and other areas have recurrence intervals of
300 to 400 years. These probabilities are, however, simply best estimates.
One or several earthquakes could occur in a much shorter period than the
normal recurrence interval, (e.g. San Fernando 1971,Loma Prieta 1989,
Landers 1992, Northridge 1994). All of these were significant earthquakes
in California. Some seismic experts state: “the farther you are from the last
one, the closer you are to the next one.”
3. Local building codes and regulations in areas o f seismic risk generally
include seismic safety provisions. There are three1 model building codes in
the United States and they reflect adopted seismic provisions; but local
36
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
building codes in many communities at risk from earthquakes still do not
include such provisions.
4. Buildings built in accordance with seismic provisions o r “to code” will not
suffer significant structural damage in an earthquake.
The principal purpose of seismic code provisions is to put forth a set of
minimum standards to ensure public safety, health and welfare. Total safety in
a building may be prohibitively expensive depending upon its function and
use. The code provisions, by procedural necessity, are an economic
compromise. Since collapse of a building is the primary cause of loss of life
and serious injury, the codes are written to provide safety against structural
collapse in a major earthquake. Significant damage may still occur in a
building designed to code during a major earthquake, and the potential for
such damage is acknowledged by codes.
Geological and geophysical science based upon known locations and
movement of faults can now identify with considerable accuracy where
earthquakes are likely to occur and what forces they will generate. In the U.S,
engineering technology permits the design and construction of structures that
can survive these forces, (FEMA.1983) even in areas of high seismicity,
without collapse but may still suffer significant damage resulting in economic
consequences.
37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Measuring Risk
Probabilities of risks to health, safety and the environment are difficult to
measure. To determine even the probability distribution of a risk requires
frequent occurrences of a hazard. When probabilities are not correctly
assessed by the society, inappropriate policy responses get formulated. While
some people believe that government interference to control risks is desirable
(Hadden,1984, p.8), there is also evidence that people demand less restrictive
measures. For example, demand for removal of helmets by motorcyclists,
resistance to requiring upgrades to older buildings to make them seismically
more safe, etc. Responses to severity of risk are also difficult to assess
because, while some people deliberately engage in risky activities like
skydiving, others are afraid of walking alone on deserted streets.
One way to assess the degree of risk-taking behavior is to compare different
risks. Hadden (1984), in her research, has found that people tend to prefer
high probability, low-consequence events to low probability, high-consequence
events. However, such preferences are less clear when high probability exists
only in a limited group; e.g. in some occupational hazards. When people
cannot even agree on the need for social response, it is extremely difficult to
formulate acceptable policies.
Rayner (1987), an anthropologist by profession, discusses the difficulty faced
by the scientific community in determining the acceptable level of risk by the
38
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
public; i.e. the level where individuals are unable to come to terms in a
satisfactory way. Rayner (p.8) elaborates four different situations, where such
a scenario could occur: isolation, transaction costs, irreconcilable interests,
and preemption.
In the case of isolation, individuals try to reduce risk by their own actions such
as avoidance, in the hope that others will act similarly. To avoid earthquake
induced risk, individuals may move to low seismic areas of the country from
high seismic areas such as California. These people tend to isolate
themselves from the effects of causative hazard. Such actions by individuals
have not led others to change their behavior.
There are situations, where transaction costs for individual solutions are too
high because numerous small risks are created which are not supported in the
aggregate; e.g. pollution created by private cars.
Irreconcilable interest situations occur when projects are planned on a large
scale for the social good, but social consensus about their safety is lacking;
e.g. large-scale dams where population resettlement is an issue or nuclear
power plant siting. Similar arguments can be made about retrofitting older
school buildings to meet earthquake safety standards. The necessity to
undertake such action statewide is perceived differently by school boards,
teachers, students, parents, and elected and state officials.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Preemption refers to decisions, which were made by one generation, affecting
future generations; e.g. disposal of toxic waste. In case of earthquake
mitigation the decision not to retrofit older buildings which may be vulnerable
in a major seismic event is simply postponing the risk to future generations.
Rayner posits that each of the four subcultures; isolation, transaction costs,
irreconcilable interests, and preemption, within a society prefers a different
basis for assessing risk. According to psychological studies, dread, familiarity
and exposure seem to be just as important as probability or magnitude (Slovic,
Fischoff and Lichtenstein, 1980). According to the sociology of perception, a
person determines a risk, its magnitude and level of acceptability based upon
his ideals. The ideals on the other hand are derived from cultural constructs,
which are based on an individual’s place in societal institutions.
In this study it is posited that, in cases of natural hazards, specifically in an
event such as a major earthquake, the cause and magnitude of hazard is
beyond the control of society, but the response to mitigate the risk arising out
of a natural hazardous event is within the capacity of the society.
Critique of Social and Cultural Theories of Risk
Social and cultural theories of risk have met with mixed receptions from the
scientific community and policy makers who depend on guidance from
technical experts. The critics reject cultural relativism on two points; scientific
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
knowledge is independent of cultural and social context, and cultural and
social aspects lack measurability. Such views even led the National Academy
of Science (NAS) to state in a 1983 report, that risk analysis should be divided
into two separate stages. The first stage is to establish facts and the second
stage is to evaluate their social implications.
Cultural relativism is also rejected by businesses, and particularly labor, where
compensation for harm is determined in a scientific and legal framework.
These institutions contend that considerations of cultural and social aspects
would leave the door wide open to numerous lawsuits over harm. Kaprow
(1985) argues that labor and community leaders fear that their legitimate
claims may be rejected on the basis of a lack of objective existence in nature
and on the basis that risk is a subjective notion.2
Kaprow further states that cultural relativists must generate more practical
ways to account for social and cultural aspects if these are to be useful in the
policymaking process. The group-grid analysis as proposed by Renn,
Wildavsky and Thompson is an attempt to attain that practicality.
Applicability to Seismic Risk
The quantification of societal and indirect economic components is
problematic, however, in seismic risk assessment, social and cultural theories
of risk must be incorporated into the knowledge base of scientific experts in
41
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
their decision process. Normative models of choice, such as expected utility
theory are inadequate descriptions of individual choices. The ambiguity of low
probabilities also affects decisions in ways that are not normative. (Camerer
and Kunreuther, 1989)
To deal specifically with low probability, high consequence events, other
theories have been proposed in the past; for example, prospect theory, and
generalized utility theory which incorporates regret as an attribute (Bell, 1982).
In a research study, Camerer and Kunreuther (1989) further explore the
implications of these theories. Their findings suggest that in low probability
events, surveys and experiments can help analysts better understand the
decision process.
For earthquake risk and most other situations of uncertainty in daily life, it is
believed that a more adequate approach combines a theory of reality
construction with a model of bounded rationality; this is then coupled with the
rejection of any purely cognitive model of decision making. More sophisticated
theories of rational decision making seem to be most applicable to situations in
which the components can be quantitatively assessed on fairly objective
grounds. (Turner et al.,1986)
Contingent weighting models (Tversky, Sattah and Slovic,1988) consider
tradeoffs between dimensions of alternatives, such as probability and utility.
42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The weights assigned to dimensions are contingent because they vary with
the form of their responses. Contingent weighting yields different results under
different tradeoffs, contrary to predictions of other theories. (Hershey and
Shoemaker, 1985; Johnson and Schkade, 1989)
Low probability risks, such as major earthquake events, seem to follow a
contingent weighting decision process. One of the critical dimensions is the
probability of an accident. A threshold level approach in the decision process
is evident; i.e. if the probability is below a certain level, no action to prevent the
risk is taken. Such an approach has been used by the Nuclear Regulatory
Commission in design considerations for nuclear power plants in the United
States (Fischoff, 1983; Bowman and Kunreuther, 1988). Individuals also seem
to utilize this approach in making their decisions whether or not to purchase
earthquake insurance. Decision-making must combine the requirements for
economic efficiency and an acceptable social decision process (Gutmanis and
Jaksch,1984, p.397). This study proposes a social decision process with
opinions of experts as proxies for stakeholders in the community.
In the next chapter total seismic risk is defined and discussed. The discussion
also pertains to components that comprise total seismic risk and factors that
affect the total risk components. Models for each risk component and for total
seismic risk are proposed.
43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Notes
1 The three model codes are being combined into one national document called the International
Building Code. The three code associations, International Conference of Building Officials (ICBO),
Building Code Officials and Administrators (BOCA) and Standard Building Code (SBC) have
established an International Code Council (ICC) that will publish its first national code known as IBC
2000 in July 2000.
2 Such diverse opinions result from the perception that empirical feedback from human interaction with
an objective universe would not be factored into social construction of risk.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3
TOTAL SEISMIC RISK
General
It is tempting to propose that total seismic risk be comprised of political,
technological, economic and societal components. However, it is argued that
the political component be considered an initial filter in the overall risk
assessment process rather than an integral component of total seismic risk.
The reasons for this argument are as follows:
Politicians, whose role is to initiate and formulate public policies and to enact
them into laws, have shied away from defining acceptable seismic risk.
Politicians, in their daily lives, have to compromise on various policy issues
depending upon the political climate in the legislative body they serve.
Politicians also do not act on policies detrimental to the interests of their
constituents, even though they themselves may subscribe to the merit of the
issue; e.g. those legislators representing tobacco-growing states will not
readily vote for legislation affecting the financial interests of tobacco growers.
Another aspect of political risk is the relationship between the legislature and
the executive branches. Policy proposed by the political party controlling the
Congress might be vetoed by the President, depending upon the political party
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
he belongs to. Similar scenarios could be repeated in state and local
governments. Seismic risk related policies might have to be severely
compromised in such a scenario. Finally, no legislator is prepared to take a
risk, which would impact re-election. Politicians thus behave in a politically
rational mode.
Components of Total Seismic Risk
Total seismic risk is therefore, proposed as being comprised of technological,
economic and societal components. These risk components are
interdependent, and together they impact the extent of damage and costs
associated with the damages. Although the components are interdependent,
the risk components are discussed separately for clarity. Total risk generated
by seismic hazard can be divided into two broad categories: fixed and variable
and can be written as equation (1).
T r = F r + V r ----------- Equation (1)
T r — Total seismic risk
F r — Fixed component of total seismic risk
VR — Variable component of total seismic risk
The fixed risk component can be defined as that arising out of damages to
physical facilities. This component can be further divided into technological
risk and economic risk and can be written as in equation (2).
46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Technological risk arises due to weaknesses in structural systems. Direct
economic risk can be defined as the cost of repairs of the damaged building
structures, utilities, and transportation network.
F r = Ft + Foe --------------Equation (2)
F t - Technological risk component
F de — Direct economic risk component
The variable risk component can be defined as consisting of those risk
components which are dependent upon the adequacy of emergency response
preparedness, composition of the community, future economic growth plan of
the community, size and number of critical facilities (hospitals, schools, fire
stations and police services) in the community and other community assets
considered important to be saved. The variable risk component can further be
divided into indirect economic risk (long and short term) and societal risk and
can be written as in equation (3).
The indirect economic risk component arises out of effects on the economy of
the community due to damage to physical facilities. Such effects could be a
loss of wages, loss of tax base, and cost of immediate and long-term health
care due to a seismic event.
47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Societal risk is comprised of the aspects listed above and also includes a
weakened economic base, and mental and psychological scars left on the
residents of the community. This component is extremely difficult to quantify.
The total seismic risk can be rewritten as shown in equation (4)
T r = F t + Fqe + Vie + V s Equation (4)
Diagrammatically, the three risk components and their interdependent nature
are shown in Figures 3-1 and 3-2. Factors, which determine each of the
components, are discussed under descriptions of each component.
V r - V ie + Vs Equation (3)
V ie — Indirect economic risk component
Vs — Societal risk component
Economic
Risk
Societal
Risk
Societal
Economic
Technological
Technological
Risk
Components of Total Seismic Risk
Interdependence of Risk Components
Figure 3*1 Figure 3 -2
48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In figure 3-1, the total area of the triangle can be considered as representing
the total seismic risk. Although there is no hierarchy to the components,
generally, seismic risk starts with the technological component resulting in
economic and societal consequences. The area under each component as
well as the total area of the triangle varies from community to community. The
interdependent nature of the components is shown in figure 3-2. If
technological risk is considered to be an initiator of seismic risk, then the
economic and societal risks ensue. However, considerations of societal risk
and economic risk may initiate investments in physical structures to reduce
technological risk. Such investment or lack of it impacts the extent of the
technological risk component. The magnitude of each of the risk components
is dependent on the others and actions taken to change one component
impacts other components.
It is better to evaluate the technological, economic and societal components;
and then, based upon the total assessment, formulate public policies which by
necessity will have to be enacted by legislators considering their political
rationality.
Technological Risk
In general, technological risks are posed when new technologies are
introduced and their long-term effects are not known. In the case of seismic
hazards, technological risk is defined as inadequate capacities of existing
49
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
physical structures, utilities and infrastructure to withstand the forces of a
major earthquake without significant damage. The damaged state of building
structures could pose a life safety risk to their occupants. Similarly, damages
to utilities and infrastructure could cause disruption in vital lifeline support
services to a community.
Buildings and structures are constructed using wood, brick, concrete and steel
or combinations of these materials, creating various types of structural
systems. Different structural systems behave differently when subjected to
major earthquakes. In general the factors that contribute to these different
behaviors are; the shape and size of the structure, materials used, and the
sites on which the structures are located. The type of soil determines the
extent of amplification of seismic forces transmitted to the structure resulting in
behavioral differences of two identical structures that are sited on two different
types of soil.
Research in laboratories on the behavior of various structural systems and
different building materials has significantly increased our understanding of
their behavior in actual earthquakes. Major seismic events also provide a
natural laboratory to assess the behavior of various structural systems.
Specific factors that contribute to technological risk are listed below:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 Design errors
2 Construction errors
3 Outdated codes of practice (Regulations)
4 Influence of non-structural elements
5 Deterioration due to external environment
6 Effects of previous seismic events
7 Discovery of new faults
8 Poor enforcement of building codes
9 Actual earthquake versus design earthquake
Each of the factors and its potential contribution to technological risk are
discussed:
1. Design Errors
Design of buildings and structures is governed by provisions of model building
codes. However, the basic responsibility for structural design rests with
licensed civil engineers who are qualified to perform such designs and are
licensed by the State of California to practice their profession. Structural
engineering is a specialized discipline within the general field of civil
engineering. “Structural engineer” is a specialized license given only to those
who pass the additional examination given by the State of California,
specifically for the purposes of testing the qualifications of the candidates in
seismic engineering.
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
As is true in many professional disciplines, the practice of structural
engineering is not free of error. Generally, the competitive fee structure in the
structural engineering profession does not allow for an independent check of
the design by another engineer or the undertaking of a peer review process of
the design. The design is checked by regulatory process of building
departments. Building department plan checks concentrate on compliance
with the code; they are not a detailed plan check of all aspects of a building
structure. It is not uncommon, that due to inadvertent oversight, design errors
occur. Design errors alter the expected behavior of buildings and structures in
an actual earthquake. Design errors weaken the ability of the structure to
withstand the effects of a major seismic event and can result in more risk than
anticipated.
2. Construction Errors
Buildings and structures are constructed by contractors who specialize in the
construction field. The construction industry is in general a self regulated
industry. Although building regulatory agencies provide inspection oversight at
specific stages in construction, the basic responsibility to complete a project
according to approved plans and specifications rests with the contractor.
The design professionals; i.e. architects and engineers, visit the project site
during construction on a periodic basis to assure themselves that the design
intent is earned out in construction of the project. However, such visits are
neither considered nor supposed to be inspection visits. Construction errors
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
occur due to poorly built construction joints, misreading of plans or poor
construction workmanship in general.
Construction errors are difficult to detect on a timely basis and are normally
discovered only after a catastrophe, such as a seismic event. The impact of
construction errors is not quantifiable a priori as the extent and location of
such errors are not known. However, construction errors may weaken a
structure; and as a result, the capacity of the structure as intended by design,
is compromised. Such weakened structures pose technological risk.
3. Outdated Codes of Practice (Regulations)
Buildings and structures are designed to conform to model building codes
such as the Uniform Building Code (UBC). The model building codes are
developed by professional architects and engineers through discussions and
deliberations in code committees. The proposed provisions are voted upon
and adopted by associations representing local jurisdictions, e.g. International
Conference of Building Codes and Officials (ICBO), an association of local
jurisdictions of the Western States in the United States. These local
jurisdictions adopt different versions of the model code for enforcement,
except in California, a specific version of the UBC is adopted by the California
Building Standards Commission (CBSC). Local jurisdictions in turn adopt the
code version adopted by CBSC. Three such associations have existed
nationwide; they recently merged into one association covering the entire
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
country. The purpose of the code committees is to continually update and
modify the existing code provisions. The modifications are worked out in a
political-technical climate.
As better information on the design and construction of buildings and
structures becomes available through research, experience and other sources,
this information is considered by the code committees for inclusion in new
code provisions. Code change proposals are developed and submitted to the
adopting association. The members of the association vote on the code
change proposals. This code change process repeats itself every three years
and codes are modified on a triennial cycle.
While new buildings and structures are designed to the latest code provisions,
a stock of existing buildings and structures, which were designed to older code
provisions exists. This stock of buildings does not meet current codes. Such a
situation may not necessarily be a cause of concern if the modified code
provisions do not differ from the previous code provisions significantly.
However, sweeping changes to code provisions have been made in the recent
past; e.g. the UBC was modified significantly in 1976 to correct deficiencies
that building systems exhibited in the 1971 San Fernando earthquake in
Southern California. Similarly, basic changes to the UBC were also made in
1988 and recently in 1997 to recognize and correct deficiencies exhibited in
the 1994 Northridge earthquake event in Southern California.
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
As more and more deficiencies in building systems are discovered through
their behavior in earthquakes, the stock of buildings and structures built to the
older codes poses technological risk unless the deficiencies in older buildings
are corrected, so that they can withstand major seismic events without
significant damage. Until that time, technological risk continues to exist and
perhaps increases as the public awareness and perception of risk changes
with increased knowledge.
4. Influence of Non-Structural Elements
Model building code provisions address design requirements for structural
systems and for non-structural elements generally. Non-structural elements
are defined as those which are not required to contribute to the structural
capacity of a building. These elements are considered appendages to the
structural system of a building. Examples of non-structural elements are:
interior partition walls; exterior window walls; and staircases. The attachment
of these elements to the primary building structural system may change
system behavior in an actual earthquake event by providing unintended
resistance. This behavior different than intended by design could cause
additional damage thus increasing the extent of technological risk.
Non-structural elements are also specialized systems supplied and installed
by specialists. Structural engineers do not design these systems as non-
structural elements are not part of the basic structural system providing the
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
resistance capacity in the building. Details of attachment of non-structural
elements to the basic structural system do not receive the attention by
structural engineers it deserves.
Adding to the complexity of determining professional responsibility, building
owners modify the layout and location of partition walls to suit the needs of the
tenants. Such changes are considered non-structural and do not receive the
scrutiny of structural engineers thus contributing further to technological risk.
There is another aspect of technological risk created by non-structural
elements. Damages to these elements pose a potential risk to the building
occupants; e.g. in major earthquake events, it is common for ceiling tiles, light
fixtures and partition walls to be severely damaged. These damaged
elements create hazards by not allowing occupants to exit safely. In recent
earthquake events, non-structural damage has exceeded structural damage
by a factor of two to four. (Eguchi and Seligson,1993, p.81)
5. Deterioration Due to External Environment
Generally, common building materials such as concrete, steel, masonry and
wood are considered to last beyond the economic life of buildings. However,
since the economic life changes when the ownership changes, building
materials can be exposed to adverse exterior environment for a long time.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Deterioration along coastal areas due to salinity in the sea air is critical.
Deterioration is severe in concrete buildings as concrete does not withstand
salty air well. Similarly, buildings constructed with wood are subject to
damage due to termites and dry rot. Deterioration increases the technological
component of risk.
Buried utilities can suffer damage due to chemicals in soils reacting with the
materials of construction. Similarly deterioration can also be noticed in bridge
superstructures subjected to continuous exposure to chemicals spilled by
vehicles. All these causes of deterioration add to the technological risk
component.
6. Effect of Previous Seismic Events
Buildings and structures are designed to withstand the effects of a major
seismic event. However, numerous seismic events of minor magnitudes occur
frequently. These events do not result in damages that can be easily detected
as they may not be readily visible. However, repeated seismic events result in
micro-cracking of the basic structural concrete frame and continue to weaken
the structure. Such weakness may result in damages greater than anticipated
for a particular building design in a major seismic event although it may not
affect the structural capacity to provide life safety. This type of increased
weakening of the structure is particularly significant in concrete structures,
masonry structures and wooden structures with bolted joints where nuts may
57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
be loosened over a period of time or bolts have crushed the material around
them making the joint loose.
The undetected reduction in the capacity of a structure to withstand a major
seismic event is most significant in buried utilities, where movement of soil
could weaken pipe joints. Transportation infrastructure, particularly bridges
and tunnels constructed with concrete, is also a prime candidate for such
continual weakening due to frequent occurrence of minor earthquake events.
7. Discovery of New Faults
The fields of geology and seismology continually explore the mysteries of soils
and earth crust. New faults are discovered regularly because not all faults are
traceable from the surface. Some faults considered “inactive” previously are
re-categorized into “active” due to suspected movements along those faults;
e.g. the Rose Canyon fault in San Diego, CA. considered “inactive” until 10
years ago, is now classified as “potentially active” by the California Division of
Mines and Geology (CDMG). New traces of the Rose Canyon fault are found
in downtown San Diego, where none were known before. Since existing
buildings in downtown San Diego were not designed for activity on this fault,
technological risk in the existing buildings has surfaced and requires
assessment.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Much of the University of California-Berkeley campus is criss-crossed by
traces of the Hayward fault. Had these traces been discovered prior to design
of many campus buildings, they would have been constructed taking into
account the effects of this fault system. Many such examples exist throughout
California reinforcing the point, that discovery of such active geological faults
results in increasing the knowledge of technological risk of the existing stock of
buildings. It is important to note that the 1983 Coalinga earthquake, 1987
Whittier Narrows earthquake and the 1994 Northridge earthquake events
occurred on previously unidentified faults. (CDMG.1995)
8. Poor Enforcement of Building Codes
Building codes are enforced by local jurisdictions; i.e. cities and counties,
through their building departments. The quality of enforcement varies from
city to city due to varying capabilities of professionals employed in the building
department. Some of the bigger jurisdictions have qualified professionals to
review the design and construction of buildings, but many smaller jurisdictions
do not. In many states building officials do not review structural design, but
only occupancy and exiting requirements.1
Although the basic responsibility of code conforming design rests with the
design professional, and the responsibility of constructing the project to the
approved plans and specifications rests with the contractor, oversight by the
59
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
local jurisdiction is important and necessary as it represents a neutral party
protecting the public interest.
The quality of design professionals and construction inspectors also varies
from jurisdiction to jurisdiction. Thus, poorer quality of enforcement of codes in
the design and construction process adds to the technological risk component.
9. Actual Earthquake Versus Design Earthquake
Model building codes define a “design level” earthquake. Although it is
acknowledged that the magnitude of an actual earthquake may be much larger
(4 to 6 times) than the design level earthquake, provisions in the code are
written to prevent collapse in an actual earthquake. The reserve capacity of
the structural system is much greater than that required for design level
seismic loads.
During an actual earthquake the structure goes into an “inelastic” stage of
material behavior resulting in cracking and yielding of joints. Such behavior by
the structure is anticipated and is in fact required to dissipate earthquake
energy much faster. Unless the structure yields, the earthquake energy is not
dissipated efficiently. Once the energy is dissipated, the forces on the
structure are reduced significantly. It is this behavior of the structure that is
anticipated in writing code provisions to design a structure for a “design level
earthquake.”
60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The actual earthquake may be much larger than anticipated resulting in
significant damages. The level of added risk in the structure cannot be
quantified a priori because real behavior is demonstrated only in an actual
earthquake event. Structural engineers rely on probability and magnitude
estimates of seismic events provided by seismologists. Since a design level
earthquake is based on these probability calculations (mean or mean plus one
standard deviation), the magnitude of an actual event could significantly
exceed the anticipated magnitude though the probability of such an
occurrence may be low.
Finally, damage to the transportation network in a major earthquake could
render the affected area inaccessible; e.g. in the 1995 Kobe, Japan
earthquake, due to damage to railroad lines and bridges, all traffic in and out
of the city of Kobe had to be suspended for weeks. Damage to Osaka port
near Kobe rendered the port unusable crippling the import-export trade that is
the backbone of that area’s economy. (Mujumdar,1996, p.3)
Technological risk can comprehensively be grouped in four areas: Design and
construction errors, poor code enforcement, site uncertainties, and
environmental factors. Diagrammatically, technological risk areas are shown
in figure 3-3.
61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Design &
Construction Errors
Poor Code
Enforcement
Seismic Hazard
Technological Risk
Site Uncertainties
Environmental
Factors
Technological Risk
Figure 3-3
Over the years, technological risks in new structures have been reduced by
better designs, stricter regulations, monitoring of construction quality, and
incorporating research findings in new designs. Before new structural systems
which are not defined in the model codes are accepted, they go through a
rigorous review by local building jurisdiction officials through SEAOC or a peer
review process. Proposed systems are compared with knowledge of known
structural systems. The damage potential of these new structural systems can
thus be assessed. Similarly, site uncertainties are reduced by continually
mapping newly discovered active faults. It needs to be recognized that the
entire stock of physical facilities is a combination of old and new buildings and
that a large percentage of structures are designed to older codes. Estimates of
realistic damage in a major seismic event would have to reflect the continuous
changes in the overall distribution of old and new facilities.
62
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In summary, it is possible to assess with reasonable confidence, the
technological component of total risk under a given seismic event scenario
because sufficient knowledge now exists in the technological field.
Economic Risk
This component is partly dependent on the technological risk component. The
bigger the technological risk, the bigger the economic risk. However, total
economic impact due to a major seismic event goes beyond damage to
physical facilities and infrastructure. Economic risk measured by costs can be
broadly divided into direct costs and indirect costs. Direct economic costs are
those that are required to repair or replace the damage to buildings,
inventories, utilities and transportation networks and things which result from
the direct impact of these damages. Indirect economic costs, on the other
hand are those which have a long term economic impact on the economy of
the community and associated costs to bring the community back to normalcy,
such as unemployment and loss of market share. Costs are used as measures
of economic risk. Diagrammatically, economic risk is shown in figure 3-4.
A. Direct Economic Costs comprise:
1 Damage to transportation network.
2 Damage to utilities (water, sewer, gas, and electricity).
3 Damage to buildings and structures.
4 Loss of building contents, and household goods.
63
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Each of the components of Direct Economic costs is discussed below:
Seismic Hazard
Indirect
Economic
Loss
Direct Economic
Loss
Long Term
Economic Effects
Technological
Risk
Economic Risk
Economic Risk
Figure 3-4
1. Damage to the Transportation Network
The transportation network is comprised of highways, bridges, railroads,
airports, and harbor facilities. Because the network serving a community can
be divided into three parts based upon the legal jurisdiction: local, state, and
federal, the damage estimates come from these different sources. Damage
assessments to physical structures can be determined by estimating costs of
repair. The timeliness of damage assessments and the estimate of repair
costs varies due to jurisdictional boundaries. A community can mobilize its
resources quickly to assess the damage to the local transportation network.
However, many transportation arteries are trans-jurisdictional, e.g. a federal
highway serving a community is owned by the federal government but impacts
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the community and is maintained by the local or state government. The
damage assessment and repair costs of such transportation network
components will have to be assessed by that jurisdictional authority. Close co
operation between many jurisdictions exists which may expedite the process.
2. Damage to Utilities
Utilities, such as water, sewer, gas and electricity, and telecommunication
networks are considered in the direct economic cost because damage to these
utilities can disrupt daily life of people and businesses directly.
The majority of utilities in the United States are privately owned and operated
and provide services to many cities and counties. The extent of damage to
utilities is dependent upon the corporate philosophy towards seismic risk and
the commitment and risk management by the owners of these utilities. The
damage to utilities is also difficult to determine as most of the utilities are
underground, and the location of damage cannot be traced from the surface
readily. It is also difficult to retrofit older utility lines although some large
private utility companies have undertaken this task. Only the newer
developments in a community have utilities designed to current seismic codes;
however, these developments also may be connected to older main utility lines
and still are exposed to risk due to damage to older lines. The complicated
nature of utility networks leaves them vulnerable to damage in a major seismic
event.
65
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
However, since it is in the business interest of the owners of utilities to restore
service as soon as possible, they repair lines quickly. The economic costs to
repair damaged utilities can be estimated only after all damage locations are
identified and the extent of damages assessed.
3. Damage to Buildings and Structures
In most major earthquake events, the damage to buildings and structures is
most visible. The larger the stock of older buildings, the greater the damage.
Damage to buildings and structures can be grouped into three areas: cosmetic
(non-structural), repairable structural, and non-repairable.
Cosmetic damage is comprised of cracking, spading and dislocation of non-
structural elements such as partition walls, exterior stucco, retaining walls,
chimneys etc.
Repairable structural damage is comprised of cracks in structural systems with
minimal impact on the structural capacity of the system. Examples of
damages of this type are usually found in concrete and masonry structures. It
is presumed that structural repair of cracks and other damage will restore the
structure to its previous load resisting capacity.
Non-repairable structural damage is more time consuming to assess and
quantify. When cracking and other types of damage result in loss of significant
66
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
structural capacity (usually more than 10%), a structure may not be repairable
or it may not be cost effective to repair. It may be cost effective to demolish
the damaged structure and build a new one. Professional engineering
expertise is required to assess these types of damages. Costs of repair can
be estimated accurately. However, one needs to recognize that as hidden
areas in the structures are uncovered, additional damage may be noticed and
may impact the initial assessment of repair costs.
4. Loss of Building Contents, and Household Goods
This loss arises out of significant lateral and vertical excitation of building
structures causing damage to contents. When a building or a structure
undergoes seismic excitation, the floors, the roof, and the walls move vertically
and horizontally. Any cabinetry, shelves or equipment attached to floors, roofs
or walls undergo this movement, which may be amplified due to the physical
characteristics of the attached components, e.g. roof mounted equipment
undergoes amplified vibrations unless it is specifically isolated to prevent such
effects as excessive movements of buildings damage building contents.
Current provisions in building seismic codes are written primarily to safeguard
life and prevent major structural damage to buildings. Although provisions to
limit the lateral movement (drift) of the structure exist, their primary objective is
to limit major structural damage and a secondary objective is to limit human
67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
perception due to motion. These provisions do not necessarily prevent
damage to building contents
In recent significant earthquakes, non-structural damage such as driveways,
garden walls, garages etc. have caused injuries. This type of damage,
including building contents has exceeded damage to the structural systems
(Eguchi et al.,1993). The cost of damage to book- shelves and books at
California State University - Northridge library far exceeded the cost of
damage to the library structure in the Northridge earthquake of 1994. The
primary reason for such extensive damage was the toppling of book shelves
(private correspondence with the Chancellor’s office) which left the books on
the floor and were damaged due to dirt, water and objects falling on them.
Another aspect that arises from damage to building contents is the potential
injury to the occupants due to breakage of glass and other sharp objects. If the
building contents include harmful chemicals, their effects due to spillage
further add to risk.
B. Indirect Economic Costs
Indirect economic effects are difficult to quantify. Community services, which
cannot be provided due to damage to utilities and transportation infrastructure,
cannot be quantified accurately due to many unknowns that need to be taken
68
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
into account Lack of community services falls into the indirect economic cost
category. Components of indirect costs are listed below:
Indirect Economic costs:
1. Business interruptions.
2. Loss of wages.
3. Temporary relocation expenses.
4. Loss of market share.
5. Perception of poor economy in the future.
6. Future stricter regulatory environment.
7. Relocation of businesses to other areas.
Each of the components of indirect economic costs is discussed below:
1. Business Interruptions
In a major earthquake event, various businesses are interrupted from
conducting their normal operations. The sources of interruptions are: damage
to buildings, equipment and machinery, and production lines; utility outages;
damage to inventory; damage to supply and delivery routes; or inability of
workers to come to work. The extent and duration of interruptions depend
upon the type and size of business, its dependence on other suppliers and the
location of business operations.
69
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
If business operations are contingent on electric power and if the power is
disrupted due to an earthquake event, the loss to businesses could be
significant. The Governor’s Office of Emergency Services (OES) in California
stresses the need for a seventy two hour survivability for home safety. It is
anticipated that most essential services will be resumed to homes within that
time. To reduce dependence on electric power, many businesses keep
emergency generators for their own needs.
Businesses can reduce the impact of interruptions if they have control over
them. However, they are at the mercy of other interruptions if the cause of
interruption is external to; e.g. damage to supply routes and infrastructure that
is not in their control can cause a significant loss of revenue.
The location of the business plays an important part in determination of the
extent of business interruptions. If the business entity is located near major
transportation routes, employees can report to work quickly reducing the
duration of interruption. Similarly, if the business entity has multiple locations,
it can supply the merchandise or services to its distributors from other
locations. It can also increase production at non-damaged locations if capacity
to increase production exists.
The impact of a major earthquake event is more pronounced on manufacturing
industries as compared to service type businesses. Even within manufacturing
70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
industries, hi-tech businesses are more susceptible to interruptions due to the
sensitive nature of tools and machinery they use in their production.
Since the causes of business interruptions are numerous, it is best to rely on a
business owner’s perspective on the extent of losses considering various
scenarios. A business owner can estimate the magnitude of potential losses
due to limited duration of interruption. Aggregation of estimated business
losses in a community can provide the estimate of total potential loss due to
business interruptions.
2. Loss of Wages
This component of economic loss can be calculated as the loss of wages is
directly attributable to days of work lost. Wages are lost either due to the
inability of workers to come to the work place or due to closure of business for
a limited time. Should the businesses choose to reimburse their workers for
lost wages, the total loss of wages is transferred to the business as an
economic loss. Because business interruptions may be of a limited duration,
the aggregate loss of wages is expected to be a minor component of the
overall direct economic loss.
3. Temporary Relocation Expenses
When residential structures are damaged to the extent that an evacuation of
occupants becomes necessary, temporary relocation is required. Temporary
71
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
housing can be provided by community shelters or other accommodations; i.e.
hotels. The expenses for boarding and lodging are borne either by individuals,
rental property owners, non-profit organizations or local governments through
assistance by OES or FEMA. The cost of such assistance contributes to
additional direct economic to a community cost unless reimbursed by OES or
FEMA.
The extent of relocation expenses is dependent on the size and composition
of the community; e.g. a community with many retirees in older homes incurs
larger costs. If the community has many rental units and a younger workforce,
relocation expenses are shifted primarily to the owners of rental buildings. The
expense is also comparatively reduced as younger people are more mobile
and are apt to stay with friends for a few days. This component of risk
compared to other components is minor and is similar in magnitude to the
“loss of wages” component.
4. Loss of Market Share
When a community is impacted by a major seismic event, its economic and
social life is disrupted. Business entities that cannot supply products or
services timely suffer a competitive disadvantage as other manufacturers step
in to fulfill consumer needs. Although such a situation may be temporary in
nature, it could lead to permanent loss of market share.
72
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Smaller businesses cannot survive significant losses and may close down
after a major seismic event. Several smaller businesses closed permanently
after the 1994 Northridge earthquake event. These businesses were unable to
restart due to untenable financial losses. (Eguchi et al.,1997)
The impact of loss of market share and the extent of loss of market share are
difficult to estimate a priori. Although smaller businesses could be surveyed to
assess their potential closure and economic impact, it is only after an actual
seismic event that the impact can be realistically assessed.
5. Perception of Poor Economy in Future
Perceptions are many times more important than reality. It also takes
considerable time and effort to change perceptions as they are deeply rooted
in cultures and value system of individuals. While this attitude may be more
pronounced in individuals, businesses also demonstrate a similar aversion to
changes in perception.
Negative perceptions by business corporations can impact a community
significantly in economic terms. Many examples can be found of the attitude of
businesses trying to establish new locations for their operations. California
was perceived to be non-business friendly due to its political and regulatory
climate in the late eighties and early nineties. Many businesses, particularly in
the hi-tech and bio- tech industry, chose to locate to Colorado and Arizona as
73
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
preferred states over California. Negative business perceptions about
California exacerbated loss of jobs and extended the recent statewide
recession (1991-1995) even further.2 While such attitudes may not be directly
related to earthquake related costs, significant earthquake events such as
Loma Prieta and Northridge add to the cost of doing business in the state.
The inability of a community to provide needed services and potential long
term financial burdens to repair the infrastructure which could result in higher
taxes are issues of major concern to the business community. The initial
economic health and the pro business attitude of a region are important
determinants in reducing negative perceptions. A priori assessment of impact
can be made, but these estimates are very crude and serve only as an
indication of likely opinions as many uncertainties about perceptions are
beyond the control of the community.
6. Future Stricter Regulatory Environment
As better information on the behavior of buildings becomes available, building
codes are modified which results in designing future buildings to a stricter
code. Requiring retrofitting of certain types of existing buildings such as
unreinforced masonry, tilt up concrete construction and non ductile concrete
frame construction falls under local jurisdiction and may incur significant costs.
74
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Economic impact due to stricter regulations and enforcement can be
calculated as the added building costs can be estimated. The overall impact
on the community depends upon the type of activity in the community and its
location. If a community is growing and considerable building activity is
anticipated, added building costs will have to be considered even though they
may not be significant. However, if the local ordinances require retrofitting of
existing structures, the costs could be significant.
Relocation of businesses to a community or establishing institutional and
governmental operations in a community also generates demand for more
buildings and infrastructure. In such cases, the overall impact of stricter
regulations can be significant due to volume of construction. The added costs
can be estimated and taken into account in the overall economic risk
components.
7. Relocation of Businesses to other Areas
Businesses rethink their strategy to continue business in certain locations after
their operations have been adversely impacted. Some of the reasons for
relocation to other areas could be: lack o f tax and other incentives to justify
added investment, diminishing business activity in the community, and
relocation of suppliers or manpower resources. If earthquake events create
such a business climate, then business would tend to relocate to other areas.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
To minimize the impact of such activity which could result in a reduced
economic base, the attitude and economic health of a community are
important considerations. Associations such as the local chamber of
commerce can gauge an overall sentiment toward business relocation. Such
sentiment and its potential economic impact can be factored into overall
economic risk considerations.
A systematic study of economic risk has yet to be undertaken. Economic risk,
in current loss estimation models is quantified as a direct consequence of
technological risk and is limited primarily to physical structures. Currently
available loss estimation models do not have the ability to quantify long term
economic losses to a community. (King et al.,1997)
A community can at best determine acceptable levels of economic risk and its
preferences for short-term versus long term economic risk.
Societal Risk
The societal risk component is more widespread and of greater magnitude
than the technological and economic components and is extremely difficult to
quantify (Tierney and Nigg,1993). Societal risk has multiple dimensions:
values, culture, geography, risk information, and interaction with other
members of the society. Community response and earthquake preparedness
also influence the extent of risk to a community.
76
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Societal risk includes the reduced ability to attract future businesses and
citizens, and a weakened economic base. Although statewide economic health
may not be impacted, communities that suffer extensive losses create pockets
of a weakened economic base. Societal scars left on the community due to the
inability of the local government to provide needed services after a major
seismic event and the impact of personal tragedies and traumas on the long
term health of citizens, are also a part of societal risk.
Seismic Hazard
Values
= E =
Culture
IE
Geography
i : :
Risk Information
~ I ---------------
Interaction VWhin Society
Earthquake
Preparedness
and
C o m m u n ity Response
Filter
Societal
Risk
Societal
Scars
V\feakened
Economic
Base
Community
Behavior
to Risk
Societal Risk
Figure 3-5
77
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The magnitude o f societal risk also depends on the composition of a
community; Le. whether a community is comprised of younger families or
retirees; is a bedroom community or a vibrant business community. Figure 3-5
diagrammatically shows various aspects of societal risk.
Values and Culture
Values and culture are probably not neatly separable as each is dependent on
one another. When individuals take actions to minimize risk based upon cost
effectiveness, or in the context of economic conditions, their decisions are
based on personal values derived from their cultural background. However,
values and culture are both modified by the external environment in which
individuals reside.
It is extremely complex to determine community aggregate values and a
community culture. However, certain majority tendencies in a community can
be detected over a period of time. Such tendencies influence a community’s
propensity to determine a response to risk posed by a seismic hazard. Risk
perception, however, is a dynamic process and perceptions are not fixed in
time. Risks are re-interpreted by communities over a period of time.
An action taken to mitigate the hazard risk also changes its interpretation. In a
study on health related risks, the actions taken by the community as well as by
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the government agencies, changed risk perception and the extent of its
effects. (Fitchen, Heath and Fessenden-Raden, 1987)
Geography
Geography’s role in determining societal risk is pre-eminent. If a community, in
its opinion is located far from a hazard, then the community’s response to risk
will be minimal and a complacency towards risk mitigation may set in.
Another aspect in geographical location is its connection with a larger society.
If the transportation infrastructure connects the community by various
alternative routes to the larger society, the risk response is minimized as
compared to limited availability of alternate routes.
Risk information
The way risk information is presented plays a major role in deciding societal
risk. If it is presented in an expert’s language, which by necessity is technical,
most members will not understand it. Risk communication must be conducted
in laypersons’ language.
Probabilistic information needs to be presented where people can compare
earthquake risk with other risks they are facing. A probability of 0.01 per year
is perceived differently than a chance of 1 in 100, which can be compared with
other probabilities known to residents, e.g. probability of auto accidents.
79
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Who presents the risk probabilities is also significant, if presented in media by
experts and government authorities, it carries a different credibility than simply
media reports. Risk information is considered differently by the community if it
is overexposed in the media. In the latter case the urge to do something about
it seems to build in the minds and actions of people.
Interaction within Society
Interaction between members of a community provides a sense of unity and
togetherness. Even though the risk posed from a hazard may be of similar
magnitude, two communities will perceive and respond to it differently
depending upon the level of interaction in the community. The better the
community’s cohesiveness is, the better the response to risk and the lesser
the impact of risk.
Interaction within members of the community depends upon their education,
economic levels and the average age of the population in addition to cultural
and religious diversities. A community with similar views on social issues will
have better interaction as compared to one with diverse opinions. On the
contrary, the higher the level of education the lesser the cohesiveness. All
these issues are difficult to decipher independently of each other, however, a
community’s overall response to hazards is influenced by these factors.
The multiple dimensions of a community decide its behavior in risk mitigation
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
strategies. The factors that contribute to the social risk component are listed
below:
Societal Factors:
1. Damage to the valuable physical assets of the community
2. Medical costs (immediate and long term)
3. Loss of life and consequences of loss of life
4. Effect of weakened social support systems
5. General non-quantifiabie issues
f. Damage to the Valuable Physical Assets of the Community
Many communities have valuable assets, which cannot be replaced; e.g.
historic buildings, a beautiful coastline etc. The extent of resources to be
expended to preserve these requires a decision that needs to be taken by the
majority of the community. The preventative measures taken to preserve these
valuable assets have economic costs associated with them. The costs can be
determined by conducting technical studies by experts. These costs are
considered a contribution to the societal risk component rather than to the
economic cost component.
Not all such decisions can be decided by economic considerations alone.
Certain valuable assets may be preserved by the community irrespective of
the immediate financial burden. Thus, moral/ethical considerations play a part
in such decisions.
81
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2. Medical costs
Immediate Costs
Medical costs arise out of treatment of direct injuries and other necessary
medical attention given to the victims of a seismic event. The extent of the
costs is dependent on the size and the composition of the community. An
older population requires more medical attention than a younger population.
Generally, the costs are absorbed by medical insurance coverage and are not
directly a burden on the community. Increased personnel staff, continued
operations of medical facilities and provision of additional medical resources
are financial burdens on the community unless all the hospitals are privately
owned. The direct social costs can be easily estimated, as the records of
additional services are usually available at medical facilities. In past
earthquake events, the direct costs of medical treatment in the United States
have not been excessive.
Long Term Costs
These costs result from injuries that require long term medical care, injuries
can be physical or mental. The effect of trauma and the shock people undergo
after experiencing a catastrophic event, such as an earthquake, are not
quantifiable or assessable a priori. Not enough data exists from previous major
earthquakes to provide much guidance in this regard.3 After Northridge
earthquake event nearly 12% in the affected area considered moving out of
the area. Residual psychological side effects include sore throats, complaints
82
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of excessive tiredness and irritability. Psychiatrists say that 35% of those lost
loved ones or dire consequences in a quake will suffer long term effects.
(Romero and Adams, 1995, p.264)
This area of determining costs is controversial because specific assessments
cannot be made as to whether a medical condition was pre-existing. However,
it needs to be recognized that there is an added cost component due to long
term medical care. The extent of that component will depend upon the
composition of the community.
3. Loss of Life and its Consequences
Loss of life is always devastating not only to immediate family members, but
also to the community. Several studies have been completed in the past to put
a value on a person’s life. All such attempts have primarily resulted in
controversy as yet unresolved. It is possible to attach an economic value to a
life, but it is not possible to assess emotional and companionship loss
(Kelman,1981,p.38).
In addition, the impact of loss of life on the members of the family and on the
community are important considerations. In this study no attempt is made to
quantify this controversial component other than to acknowledge it.
83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4. Weakened Social Support Systems
After an earthquake event, either as an economic consequence or other
constraints, social support systems may be weakened. Weakened services to
the community may include libraries, paramedical care of senior citizens etc.
The reduced services will have to be restored by other institutions or the
services will be adversely impacted permanently. The economic value of such
services is difficult to estimate; they are among the social assets of the
community that are lost or diminished in value. This loss in value contributes to
the societal risk component.
5. General Non-Quantifiable Issues
Certain aspects of a community’s life cannot be assessed only by economic
considerations. These aspects include ethical/moral obligations of the
community, e.g. care of senior citizens, care of physically handicapped
individuals, child safety etc. A community needs to make a list of critical
facilities in which these people are housed. Appropriate measures need to be
taken not only to preserve these facilities but also to have social services
available for these individuals in a major seismic event. Economic costs of
such services can be estimated.
Each community will respond differently to a major seismic event; therefore,
the societal risk for each community will be different. Unless the societal risk
84
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
component is integrated into the overall damage evaluation, total seismic risk
assessment is not complete.
Although, the components of total risk can be identified and described, two
components are difficult to quantify; indirect economic effects and societal risk.
This study does not purport to quantify these components. The attempt in this
chapter has been to provide a comprehensive description of components of
risk.
In the next chapter acceptable seismic risk to a community and loss estimation
methodologies and their limitations are discussed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Notes
1 This is based upon writer’s experience as a professional engineer in the states of Massachusetts,
Tennessee, Rhode Island, New York, Ohio, and Hawaii.
2 Discussions by the writer with staff of the Governor’s Office of Planning and Research during 1994-
1998.
3 Based on the testimony by the Mayor of Watsonville, CA before California Seismic Safety
Commission hearing on 10th anniversary of Loma Prieta Earthquake, Sept.1999, Monterrey Bay, CA.
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4
LOSS ESTIMATION METHODOLOGIES AND ACCEPTABLE SEISMIC RISK
In this chapter, current loss estimation methodologies and their (imitations are
discussed. The discussion is limited to understanding the bases of
methodologies rather than the details therein. After a review of these loss
estimation methodologies, the concept of acceptable risk to a community is
presented and discussed.
LOSS ESTIMATION METHODOLOGIES
Three loss estimation methods are discussed: Probable Maximum Loss
method, Applied Technology Council (ATC) methodology and the HAZUS
method. Although other methods may be utilized for specific applications, the
three methods discussed here are widely used by seismic loss estimation
experts.
Probable Maximum Loss (PML) Method
The Probable Maximum Loss methodology was developed by Karl
Steinbrugge and Ted Algermissen 1 for the Insurance Service Office. The
method concentrates on determining monetary loss, which determination is
then used by insurance companies in settlement of potential claims, setting of
insurance premiums and underwriting insurance policies.
87
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The fundamental assumptions of this methodology are:
1. Earthquake magnitudes may be effectively converted into Modified Mercalli
Intensity 2 (MMI) patterns.
2. MMI attenuates as the distance from the causative fault increases.
3. Monetary losses are directly related to MMI and the type and value of a
structure.
4. Damage to building contents are estimated at 50% of the structural
monetary loss.
5. Fires following earthquakes cannot be estimated due to the many
variables related to sustained fire.
Methods developed and utilized by two insurance organizations based on PML
are presented below:
Allstate Research and Planning Center (Menlo Park, CA.)
This arm o f Allstate Insurance company developed a model for estimating
potential exposure in the residential insurance market due to a major
earthquake. The model is based on input from data on earthquake faults, their
magnitudes and epicenters and a basic deductible level before claims are
settled. The model uses a series of simulated earthquakes and utilizes ATC-
13 damage loss factors.3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The methodology concentrates on damage estimates to low rise wood frame
and low rise un-reinforced masonry buildings. This model is also used to
calculate expected long-term annual loss from a given fault by calculating
average losses from various magnitude earthquakes based on probabilities
associated with each magnitude. The model was tested using the damage
estimates for the 1987 Whittier earthquake. The loss predicted by the model
was within 5% of the actual loss to the company. The company believes that
its model is reasonably accurate. (CDMG,1990)
EW Blanch Co. Reinsurance Services
The Blanch Co. model is also based upon Steinbrugge’s work and the concept
of Probable Maximum Loss. Their model is called CATALYST. It defines the
probable maximum loss as the average probable maximum monetary loss to
structures, which will be experienced by 90% of the buildings, in a given class
in a specified PML zone (within 6 miles of a fault rupture) assuming an
earthquake of the magnitude 8.2 on the Richter scale. CATALYST estimates
risk based on expected damage ratios for certain building types at various
distances from an earthquake fault rupture. This model uses residential and
commercial data along with distance from the fault rupture of the quake; it
provides estimates for 21 Maximum Credible Earthquakes4 (MCE) along 10
major California faults.
89
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Applied Technology Council (ATC) Methodology
Applied Technology Council (ATC), a non-profit organization, started initially
by the Structural Engineers Association of California (SEAOC) to conduct
applied research, derives its funding from various governmental and non
governmental grants. Its prime purpose is to develop source documents that
translate and summarize useful information to practicing engineers. Its focus
has been on seismic engineering.
Through a series of reports, particularly ATC-13, ATC-20 and ATC-21, the
council presents a methodology to assess damage to physical facilities due to
earthquakes. In the ATC-13 report, the council developed expert opinions on
earthquake damage and loss estimates for existing industrial, commercial,
residential, utility and transportation facilities in California. This report also
includes damage probability matrices for 78 classes of structures and
estimates of time required to restore damaged facilities to pre-earthquake
usability. The report also emphasizes the inventory information considered
essential for estimating economic losses and the methodology used to
develop the required data. Damage loss factors developed under ATC-13 are
used by loss estimation experts to determine probable maximum loss (PML).
The ATC-20 and ATC-21 reports are not directed toward estimating losses;
they concentrate on determining the usability of buildings following an
90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
earthquake. The methodologies developed under these reports are, therefore,
valuable in assessing the seriousness of risk posed by buildings damaged in
an earthquake, rather than estimating losses.
Limitation of PML and ATC Methods
As can be noted, the primary emphasis in the PML and ATC methodologies
has been to assess the potential damage to a structure and its non-structural
components and to estimate contents’ loss as a percentage of structural loss.
The cost incurred to repair facilities to their pre-damage condition is
considered the estimate of loss. Because the PML method was developed
primarily for insurance companies and the ATC method to assess building
specific damage, both methodologies do not take into account societal impact
or long-term economic impact. These methodologies to conduct loss
estimation due to a major seismic event concentrate on technological and
resulting economic risk components only. Some experts consider that the
ATC-13 method represents a poor quality damage algorithm. Both PML and
ATC methods are therefore limited in their scope and are intended to assess
damage estimates for physical structures only.
The two methodologies described below are not directly based on PML and
thus offer other modified loss estimation techniques.
91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
EQE Engineering Model
EQE, a private company specializing in earthquake loss estimation, has
developed a model known as EQ Hazard. The model is based upon seismo-
tectonic modeling of faults, earthquake shaking intensity at sites, vulnerability
ratios based on site intensity and determination of the mean and high fractile
insured loss.
The methodology attenuates peak ground acceleration to the site and then
converts it to a MMI.5 The vulnerability is computed on the site specific MMI
and ATC-13 damage loss ratios. Probable maximum loss is estimated using
the maximum credible earthquake with a return period of 475 years from each
potential earthquake source.
Stanford University Model
The department of civil engineering at Stanford University developed an expert
system known as the Insurance/Investment Risk Assessment System (IRAS).
The model is intended to be a user friendly, knowledge-based expert system
for use by engineers, real estate investment analysts, insurance underwriters
and decision makers. IRAS can be used to assess individual risk, risk to a
portfolio of properties, or regional risk.
The Stanford methodology allows the choice between PML or ATC-13
methods. Loss-damage estimates are based on scenario earthquakes. An
92
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
improvement in this model is that it considers building specific information.
This expert system now resides in a private company domain (RMS Inc.) and
is no longer publicly available.
Both the EQE engineering model and the Stanford University model go
beyond facility specific damages. However, even in these models, the
emphasis is on damage to facilities, though on a wider geographic basis. An
effort to develop a more comprehensive methodology to estimate seismic
losses was undertaken by the National Institute of Building Sciences (NIBS),
an arm of the Federal Government. Through a contract with a large team of
consultants including ATC, NIBS produced a methodology in 1997, known as
HAZUS. This methodology attempts to take into account indirect economic
losses along with technological and direct economic losses.
HAZUS Methodology
HAZUS is intended to be used primarily by state, regional, and community
governments. The basis for all its default data is the geographic information
system (GIS). A zip code track is used to build the general building data, but
essential facilities such as hospitals and lifelines are built in to the data as
specifics of a site. The framework of the methodology comprises of six
modules. The modules are interdependent and output from one module is
treated as input to another module. The modules range from a ground motion
data module to an indirect economic loss module. Physical facilities including
93
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
infrastructure are aggregated in the inventory module. For a chosen
earthquake event, specific ground motion, ground failures and other earth
science hazards are generated. The impact of these earth science
occurrences generates the probabilities of direct damage to facilities and
induced damage such as fires following an earthquake.
The methodology estimates direct and indirect economic losses. Indirect
economic losses in HAZUS are defined as changes in employment and
personal income. A default database is built-in; however, specific data bases
can be incorporated in the software to improve the quality of loss estimation
(Whitman, et al.,1997). Indirect economic losses are calculated as a result of
direct economic losses; the calculations rely on effects of supply and demand
functions. It is not clear in the methodology as to how the indirect losses are
calculated for a community which is primarily a residential one without a lot of
business activity.
Limitations of HAZUS
Although HAZUS has captured a nationwide default database and is a
significantly better method for loss estimation than previously available
methods, its limitation is that a default database must be relied upon unless
community specific data is available. Specific data can be used as input
instead of the default data base. HAZUS also does not consider societal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
factors explicitly. Lifeline disruption is not coordinated with the direct and
indirect economic loss modules. (Brookshire, Chang et al.,1997)
One can argue that, if all the existing buildings, physical facilities, utilities and
infrastructure were retrofitted to comply with the current seismic codes, then
the potential life safety risk would be virtually eliminated. Since life safety risk
is almost removed, the cost of such an endeavor would approximate the total
cost for virtually eliminating life safety risk. Theoretically, the argument is
powerful; however, in reality the cost to retrofit all facilities may be judged
unacceptably large and therefore this is not a viable option to citizens or
decision- makers. (Tierney and Nigg,1993)
Although the life safety risk may be virtually eliminated, the behavior of
buildings and structures designed to latest building codes cannot be assumed
to be damage free in a major earthquake because such an assumption is not
realistic. Possible disruption to business functions is an important factor and,
therefore, estimates of losses related to functionality may outweigh the
concerns of building-facility damage for some decision- makers. Seismic risk
cannot be eliminated due to the various factors involved as noted.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The loss estimation methodologies are designed to predict aggregate
estimates of potential loss. The seismic risk may be reduced by actions of the
community. To determine seismic risk to a community, its input is necessary.
Reasons for Community Input
In this study, four reasons are proposed for seeking input from the community.
First, disaster assistance by Federal and State governments may be limited
and more of the burden will have to be shared by the community (FEMA,
1997). Second, Federal and State government partnerships create a culture of
accountability for Federal grants. Communities will also have to be
accountable for the grants (FEMA,1997, p.1). Third, the community’s interest
is best protected by actions which are a result of conscious decisions taken by
the majority in the community and, fourth, community composition, economic
and social capacity and geographical locations vary for each community
resulting in differing responses to the same seismic risk.
“To effectively manage risk, we must seek new ways to involve the public in
the decision-making process. They (the public) need to become involved early,
they need to be informed if their participation is to be meaningful."
(Rucklehaus, W. D., 1983, p.1028) Universal acceptability of risk cannot be
determined. What is an acceptable seismic risk to a particular community may
not be acceptable to another community. A proposed definition of acceptable
seismic risk to a community is presented.
96
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The model proposed in this study emphasizes two components: determination
of acceptable seismic risk by stakeholders, and community input through
procedural fairness.
Acceptable Seismic Risk to a Community
All acceptable risk problems are decision problems and therefore require a
choice among competing alternatives and varying courses of action to deal
with risk. A properly designed and executed decision process facilitates
identification of the most acceptable choice. One might call the risk associated
with the most acceptable choice the “acceptable risk.” (Fischoff, Slovic, and
Lichtenstein, 1981)
A critical aspect of acceptable risk is to determine who is impacted by the
acceptable risk decision? From whose viewpoint is risk to be judged; its
victims, experts, representative organizations, the community at large, or the
government? Each viewpoint defines its own widely different level of
acceptability. (Starr, 1976)
In this study it is argued that acceptable seismic risk ought to be decided by
the community at large after taking into account expert opinions on the
technological risk component and economic consequences arising out of it.
Therefore, the following definition of acceptable seismic risk is proposed.
97
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Definition of Acceptable Seismic Risk
Acceptable seismic risk is that amount o f risk which could be estimated
and voluntarily accepted by an individual, a fam ily, a group, a
community or society, taking into account technological, economic,
social and political considerations.
Since the basic focus of this study is acceptable seismic risk to a community,
voluntary acceptable seismic risk in the community context is defined as, that
level which is acceptable to the majority in the community. Individuals of
course have the choice to reduce their personal risk beyond the community-
based acceptable risk.
Since seismic risk in this study is posited to be a social and cultural construct,
assessment and determination of acceptable seismic risk must have
community input. Decisions related to rank ordering of risks and the extent of
their reduction ought to be aided, ideally, by the best possible information.
Such information should include the magnitude and the associated costs of
marginal reductions, to ensure as much as possible equity of treatment
between different elements of a risk conscious society (Starr,1976). It is hoped
that a community based decision on acceptable seismic risk would lead to
creation of incentives towards mitigation of seismic hazard by the government
and may even lead to regulations related to seismic safety.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Although acceptable risk can be defined universally with global dimensions, in
this study it is emphasized that its determination should rest at the local
community level. Thus, determination of acceptable community seismic risk is
possible in the United States since local governments are empowered to make
decisions related to the welfare of their residents within larger mandates by
State and Federal governments. A similar decision process would work in
other societies where local communities have political, legal and economic
powers resembling those in the United States. The quality and efficiency of
community based decisions would partly depend upon the extent of a
community’s dependence on state and federal governments for economic
assistance and legal constraints.
Dependency of Acceptable Seismic Risk
As the level of analysis is the community, no attempt is made to address
considerations related to an individual, or a family although they are included
in the general definition of acceptable risk. Seismic risk is an independent and
a dependent variable, but acceptable community seismic risk is a dependent
variable. Its dependency originates through variables that are internal as well
as external to the community.
Internal variables may be defined as those aspects and characteristics that are
specific to a community; e.g. composition of community, economic base and
manpower resources. External variables are those that are outside the
99
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
jurisdictional authority of the community, but the community is deemed to be
dependent on these external resources for its normal functioning, such as,
transportation networks serving the community and utility services supplied by
sources outside the community. External resources generally fall in the
infrastructure category. External variables also include risks posed to the
community by external causes e.g. inundation by damage to a dam outside
the community.
Several generic complexities are pointed out by Fischoff and Lichtenstein
(1981) in determining acceptable risk. These complexities may be summarized
as; uncertainty in defining the problem, difficulties in assessing the facts,
difficulties in assessing relevant values, uncertainty about the human element
in the decision-making process and difficulties in assessing the quality of
decisions that are produced as outcomes.
However, such complexities are facts of life for decisions on acceptable risk-
related problems. Since decisions are made considering risk, costs, and
benefits; subjective aspects of the human element are integral to the decision
process and require value tradeoffs. It should be recognized that acceptable
risk is also contextual and time dependent. As the values change over a
period and more information on risk becomes available, the most attractive
choice (acceptable risk) may also change over a time horizon.
100
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In order to provide the best possible information, it is important to discuss the
complexities inherent in separating facts and values. (Simon, 1946)
Facts and Values
Information is created if someone has a need to use it. Individual scientists
create data, but it is the community of scientists and other interpreters who
create facts by integrating data (Levine, 1974). In so far as the world is
understood through the prism of science, the facts it creates shape
perspectives (App!ebaum,1977; Henshel,1975; Markovic,1970). Levine further
states that legal requirements are an expression of society’s values that may
strongly affect its view of reality.
Some argue that science can “anesthetize” people to think about the
unthinkable, e.g. setting an explicit value on human life in order to guide policy
decisions, even though such values are implicitly set by decisions that are
made. Even flawed science may shape values (Wortman,1975). Special
interests also shape them. Toxicological studies, even with low statistical
power, have protected chemicals more than people (Page, 1978,1981). The
value bias is further compounded when scientific caution also balances
regulatory caution. “Values are why people care about risk management. It is
natural to involve the public in decisions about their risks by asking them about
101
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
their values. In addition, if values are explicit, we can communicate why one
alternative is chosen and why others are not.” (Keeney,R.,1987,p.128)
Separating values and facts is intellectually challenging, and the absence of
this discipline may lead scientists to be overconfident in their data and
politicians to claim expertise. However, facts and values are intertwined in the
very definitions of problems. Values shape facts and facts shape values
(Levine,1974; Price, 1974). To determine facts, scrutiny from all sides is
needed to improve the quality of the analysis.
Since the level of acceptable seismic risk is to be decided by the community,
the variables affecting its determination need to be discussed. In this study,
four variables are identified; interpreting scientific data, assessment o f risk-
cost-benefit analysis, group decision making and institutionalizing the process
o f change. These variables are discussed along two dimensions; the time
context and the total resource capacity o f the community.
VARIABLES OF ACCEPTABLE SEISMIC RISK
Interpreting Scientific Data
Presentation of data in a format and language that the lay public can
understand is the responsibility of experts if their recommendations are to be
lent credibility. The public relies on technical information to make judicious
choices. Since seismic events, including their magnitudes and frequencies of
occurrence, are probabilistic in nature, it is necessary for the public to
102
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
understand basic probability concepts and assumptions about what we do not
know. Based upon his studies of primitive societies and their understanding of
risk, Wynne (1987) argues that the difficulty in grasping probability rests with
social insulation. Even experts experience similar difficulty interpreting
probabilistic information.
Furthermore, public perception is dependent on the origins of risk; e.g. if a risk
is originated by a minority, the majority is angry and attitudes toward such risk
become political (Shogren,1990, p.13). The public also interprets data on
voluntary and involuntary risks differently. Individual behavior demonstrates
the acceptability of voluntary risk at higher levels than involuntary risk; e.g. the
public accepts one in 1000 as the rate of death in private transportation
vehicles, but it is willing to accept only one in a million as the death rate in
public transportation. (Starr, 1976)
Even though the origins of a seismic hazard risk rest in nature, the magnitude
of its effects are dependent on the mitigation efforts undertaken by the public.
In that regard, seismic hazard risk is subject to similar considerations as other
man made risks.
The problem of interpreting data on major seismic events is compounded
further since the annual probability of occurrence is low (0.002 in some parts
103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of California and 0.0004 in parts of the East Cost). Such low frequency of
occurrence does not carry as much weight as other events of frequent
occurrence; e.g. accident rates in private transportation vehicles.
Scientific data interpretation by the public is also impacted by geographic
proximity to the site of the event. Although scientific data might be credible, its
significance for a community will vary depending on its geographic proximity to
the event. The community may, therefore, interpret the same data differently.
This issue is particularly troubling because two communities, only 50 miles
away from each other, may interpret seismic events data differently although
the impact of a major earthquake along an active fault may be the same for
both communities.
Another aspect that is important (but not necessarily related to scientific data)
is based on immediate past experience; the public tends to have a false sense
of safety if a hazard event did not impact them personally in a major way
(Shrader-Frechette,1991). As an example, the probability of death in a major
seismic event is interpreted differently by school authorities because two of the
most recent seismic events, i.e. Loma Prieta, 1989, and Northridge, 1994, did
not entail fatalities to school students. The public tends to forget that both of
these events occurred at non-school hours and thus posed no threat to
students, teachers, or staff.
104
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In a number of ways experts fail to inform the public. One is by not telling the
whole story about hazards they know best because they fear information will
make the public anxious. A second way is to assume dissemination is not their
job. Third, some experts may have a vested interest in keeping things quiet
(Hanley,1980, p.5). It is extremely important, however, that known information
be conveyed to the public in the language they understand so that it leaves
little room for different interpretations of the same data.
Considering time contexts, scientific data changes as more information
becomes available or re-interpretation of previous data is made with the
advancement of science. Similarly assessments of potential damage are
subject to major revision over time. Many examples can be cited. The potential
damage to existing non-ductile concrete buildings changed significantly after
the 1971 San Fernando seismic event in which this type of construction
proved to be vulnerable. The behavior of structures, built with steel moment
frames, in the 1994 Northridge seismic event changed the potential damage
estimates for these types of existing structures. In 1997, additional geo-
technical data were introduced nearly doubling the seismic force on buildings
within 15 km. of an active fault. The loss-potential of buildings within these
zones has drastically changed due to recent knowledge.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
If scientific data on the probabilities of major seismic events and their impact
on the community are well understood by the decision-makers in the
community, then hazard mitigation efforts are impacted by the resource
capacity of the community. It is important to realize that interpretation of
scientific data is very critical if a community is to decide on the extent of
resources to be mobilized to mitigate the impact of a seismic hazard.
Assessment o f Risk- Cost-Benefit Analysis
The Risk-Cost-Benefit analysis (RCBA) methodology employs the technique of
converting risks, costs and benefits into monetary terms and then determining
if the benefits outweigh the risks and costs. RCBA is a powerful tool; but it is
based on the presumption that monetary value alone represents the real costs
or benefits, and that the magnitude of the monetary benefits is more important
than their distribution.
Risk-Cost- Benefit analyses are performed ex-ante because they reflect
preferences at the time of the decision and allow evaluation of risk. However,
economists currently do not believe that interpersonal comparisons of utilities
are scientifically valid (Whittington and MacRae, 1990, p.541). When Risk-
Cost-Benefit analyses are conducted considering only technological and
economic issues of physical structures, they can be performed free of
personal value judgments. However, when the societal factors are included,
the analyses cannot be conducted free of value judgments.
106
permission of the copyright owner. Further reproduction prohibited without permission.
There is a basic controversy in RCBA, which relates to the issue of whose
preferences are to be counted in conducting the analysis. Risk-Cost-Benefit
analysis is useful only to the extent that a general consensus exists that the
value assumptions are legitimate (Trumbull, 1990). The basis of cost-benefit
analysis is the potential Pareto optimality principle.6 This principle underlies
the argument that projects that result in potential Pareto improvements should
be pursued, even though transfers to make the actual Pareto improvements
cannot be made. The potential Pareto optimality principle is based on a
utilitarian logic. Aggregation of utilities by necessity implicitly compares utility
across individuals (Weimar and Vining, 1989). Therein lies the problem.
Unless the individuals whose preferences are to be counted are identified,
their preferences cannot be aggregated. Policy analysts and politicians,
however, use cost benefit analysis to make interpersonal comparisons of
utilities. Trumbull (1990, p.211) argues that a meaningful society should
include all individuals who are affected by the project. Such universalism,
however, does not help in identifying the individuals whose preferences should
be counted. The resolution of the issue as to whose preferences should be
counted needs to be decided by the public through discussion and public
debate based on ethical and philosophical grounds (Whittington and MacRae,
1990). In this study, stakeholders whose preferences should be counted are
identified.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
RCBA methodology is value laden because the data on risk and benefits and
costs is factually incomplete and imprecise. However, in assessing
technological and economic factors of acceptable seismic risk, the RCBA is
the most appropriate methodology to use. In its use it is important to raise an
ethical question; i.e. should seismic risk evaluation be primarily scientific or
predominantly political? Risk evaluation cannot be both wholly scientific and
wholly sanctioned by democratic procedures. The benefits of conducting a
risk-cost-benefit analysis are that the analysis directs resources to more highly
valued uses and forces a consideration of costs and benefits beyond the
immediate environment of a project (Trumbull, 1990). One of the basic
dilemmas to be resolved in risk-cost-benefit analysis is the time period for
which such analysis should be conducted.
RCBA as a tool can be improved and made more effective if it encompasses
the following:
1. Compare diverse risks considering probabilities and consequences based
upon different assumptions and scenarios to competitively price the
alternatives.
2. Rank order community preferences based upon their value judgments.
3. Explain explicit decision criteria and allow public debate on choices of
alternatives.
4. Assign different weights to RCBA components by acquiring input from
different groups in the community.
108
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Different RCBA’s can be conducted using different event scenarios so that the
public and the policy makers can make informed decisions. This improved
RCBA methodology will allow democratic decision-making and facilitate
understanding intractable risk decisions. The strength of improved RCBA is
that it provides a framework that can be interpreted in terms of many different
value systems.
The outcome of RCBA is closely tied to the resource capacity of a community.
In fact, the input from the resource capacity data could alter the RCBA
outcome. RCBA’s also change with time not only because risk changes, but
costs and benefits also change with time. RCBA’s should therefore be
considered dynamic in nature and valid only at the point in time when they are
conducted.
The role of a policy analyst is to be concerned with real decision making within
the policy process; and, as such, the role needs to go beyond economic
analysis based on the potential Pareto principle. Analysts must allow for a
dialogue with other professions and the public to consider ethicai grounds and
distribution effects in developing policy recommendations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Group Decision Making
When an individual is faced with alternative choices, a decision to follow a
certain course of action can be made based upon one’s own utility function.
However, when a group has to decide on a course of action, a different
dynamic takes place, making group decisions complex phenomena.
If organizations with differing economic and social interests are involved in the
decision process, the complexities increase multifold. The difficulty in group-
decision making is further increased by the significant variations in types of
organizations engaged and situations they encounter.
The ‘first virtue’ of social institutions is justice and fairness. To arrive at fair
social institutions certain non-subjective negotiating is necessary
(Rawls,1971 ).7 It is this precept which is basic to socially responsible group
decision-making. However, such an ideal situation is not readily achievable.
Therefore, in this study two theories are considered to be most relevant in
understanding the complex phenomena of decision-making: the theory of
cooperation, and the theory o f exchange. It is argued that people need to
understand the basic tenets of both theories. To arrive at a commonly
accepted decision by stakeholders, cooperation is necessary. However, to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
reach a cooperative stage, each party in the group needs to perceive that it is
getting positive benefits in exchange forgiving up something of value.
Cooperation in organizations was first discussed by Barnard (1938) in an
attempt to motivate workers and to create a pleasant view of the work
environment. The concept of cooperation is based upon having a mutual
interest among participating parties towards a successful outcome.
Following Barnard’s work, Simon (1946), and Cyert and March (1956)
developed the concept of cooperation in organizational behavior and defined it
as patterned relationships. Simon argues that, for cooperative behavior to be
stable, parties to the agreement must have knowledge of each other’s
strategies. This view is in contrast to that stated by Rawls. The theory of
cooperation also argues that cooperation is difficult to sustain if there is no
repetition of interactive actions (North,1996), particularly when information on
participants is lacking and when the number of participants is large.8 It can be
argued that, to be able to form a common notion, as a minimum, two parties
must know about the beliefs and wants of each other in a given set of
circumstances. (Schofield, 1985, p.209)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In a community setting where the common goal is to maximize benefits with a
consequent reduction in costs associated with risks, it can be assumed that
beliefs and wants of stakeholders to a large degree will be known.
Exchange theory considers group behavior in terms of rewards exchanged
and costs incurred in different interactions. This behavior can be considered
as an exchange based on evaluations of costs and benefits. The theory base
is micro-economics; it proposes that, for a behavior to occur, various material
and nonmaterial inducements must be offered. When the rewards exceed the
costs by at least a minimum level of expectation, exchange takes place.
(Kelly,1980, p.480 )
It can be assumed that the group decision will be made fairly according to a
Rawls’ concept, because stakeholders are deemed interested in the well being
of the community as a whole. All of them are facing a common hazard risk,
and the community is going to benefit by reduction in risk consequences.
Both the time context and resource capacity of a community are important
issues in group decision-making. Values change over time and the
composition of groups also changes, bringing new elements in exchange and
cooperative processes. The determinants of costs and benefits required for
group behavior also change with time. Resource capacity of a community,
which is also a major factor in group- decisions, leads to different negotiating
1 1 2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
tactics and awareness of the constraints by those engaged in decision
making.
The decision as to the acceptable seismic risk in a community setting requires
cooperative exchange processes; and, therefore, it is proposed that underlying
ideas from both theories need to be utilized in the determination of acceptable
seismic risk. This determination has to be achieved in spite of what Raiffa
(1980, p.341) found and stated; “ Our societal dialogues are becoming
increasingly strident, more adversarial, more litigious and less constructive.
We are not engaged in collaborative and collegial problem solving exercises
that seeks gains for a ll or for almost all.”
Institutionalizing the Process o f change
This task is one of the hardest to accomplish in a societal context, although
individuals routinely revise their opinions and actions based upon experience
and new information. The process of change becomes an identifiable and
integral part of the course of action by individuals.
A society also undergoes a similar transformational process on a routine
basis. Experience has shown that a society does change its opinions and
consequently modifies its actions to mitigate risk situations when new
evidence or data are introduced. However, a methodology to institutionalize
113
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the process and capture such changes in society is missing. An active group
of citizenry and specifically of stakeholders is a necessary precondition for
institutionalizing a process of change.
Media reports influence and change opinions about technological risks as
presented by experts as new research data becomes available and is widely
disseminated. How do such media reports affect the perceptions of societal
groups towards risk acceptability? Whether such reports are purely media
hype or deserve further attention and study is a decision that ought to be
taken by societal groups through deliberations. The roles played by local
governmental institutions and regulatory agencies in actively informing or
suppressing information from the public are important considerations in
institutionalizing a process of change. As the resource capacity of a
community also changes over time, a continual review of acceptable levels of
risk needs to be undertaken by active citizenry groups.
Raiffa (1980) posits that many decisions taken by a society on acceptable risk
can change over a period of time as acceptable risk is time dependent. Little is
known about many potential risks. Each year only a trifle more may become
known resulting in inaction or displacement of one risk with another one.9
As an example, after the 1971 San Fernando earthquake in Southern
California, the weakness of non-ductile concrete frame structural systems in
114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
building structures in resisting earthquakes became evident.1 0 An
institutionalized process to change building code regulations exists through
code writing bodies.
March (1996, p.309-319), in an article in the Psychological Review, suggests
that experimental learning which conforms to standard learning models is
shown to lead learners to favor less risky alternatives when outcomes are
positive. Learning to choose among alternatives with risky outcomes, on the
other hand, may result in favoring more such alternatives in the short run.1 1
It is proposed that a community establish an active group of stakeholders on a
permanent basis. This group should keep abreast of new information and
developments in the area of seismic risk that affect the community. This group
should also maintain active communications with political leaders in the
community and with rescue and relief agencies. Finally, this group should
initiate and formalize a public participation process on a regular basis. A
culture of change to respond to new realities is needed to institutionalize the
process.
Ethical I Moral Dimension
As a civil society, it is our obligation to take care of certain segments of the
population. This obligation goes beyond the norms of economic rationality.
115
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
This study posits that such obligations are to be considered the ethical/moral
values of a civil society.
Certain aspects within a community can be described and hopefully agreed
upon as moral/ethical obligations; e.g. taking care of children, the elderly and
the sick. Such issues cannot be rationally discussed only along the economic
value dimension. It is the responsibility of local and state governments to
devise laws and regulations to effect actions within a community to protect
weaker sections of a community.
in a community, identification of obligations that need to be considered on
moral/ethical grounds is necessary. A list of critical facilities and supporting
infrastructure must be made. To protect these facilities in a major seismic
event, provisions must be made by enacting laws, creating an implementation
structure and monitoring actions. Such critical facilities could include housing
for the elderly, community nursing care facilities, childcare centers, schools
and hospitals.
Since there is no economic incentive to have these critical facilities remain
functional during a major seismic event, the best course of action is enacting
regulations. These regulations must go beyond the normal structural safety
aspects; they need to address non-structural components, building contents,
and utility support systems. It is important that these facilities remain functional
116
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
during an earthquake because of their critical nature. In the case of
homeowners, a certain minimum level of structural safety should be provided
in building codes.
In case of other physical property owners, the acceptable risk to them is the
deductible amount in the insurance policy. In the case of financial institutions
acceptable risk may be zero, because they would want insurance to protect
their investment. They are in a position to transfer risk to insurance
companies.
In lifeline support systems like utilities, the acceptable risk to the community
would be the economic cost due to the unavailability of their services for a
limited period of time. Similarly for the transportation network, it is the
acceptable inoperable time, which is of importance to a community.
How many deaths and serious injuries are acceptable? These are complex
and difficult issues to debate and resolve. However, societal values need to be
addressed and a debate within a community needs to occur so decisions by
the majority can be taken.
Different communities may decide to put different values on aspects of a
similar nature. The decisions, although different in different communities,
determine the extent of societal values that are important for the majority.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
There is no reason to universalize the issue as community specifics play a
major role. What is acceptable to the majority in the community should be
taken as the community value.
Decisions about acceptable seismic risk policy should go beyond an economic
analysis. Policy analysts must develop a dialogue with other professions and
the public to consider ethical connotations and distribution effects in
developing policy recommendations.
Informing citizens of potential risk and obtaining their participation is a matter
of guaranteeing free choice on acceptable seismic risk. To achieve explicit
consent, citizen participation in negotiating solutions is critical. Once it is
realized that the process of hazard assessment and management is highly
value laden and politicized, then negotiation rather than mere expert decision
making becomes a virtual necessity for ensuring freely informed consent,
particularly to reach decisions on acceptable levels of risk. (Shrader-
Frechette,1991, p.206-209)
The next chapter discusses attributes of acceptable seismic risk and issues
related to identifying stakeholders, their selection, consideration of their
opinions and the resource capacity of a community. Collective decisions by
representatives of stakeholder groups are posited to be decisions taken by the
community as a whole.
118
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Notes
1 Karl Steinbrugge a consulting structural engineer in private practice and Ted Algermissen, a
seismologist working with the U.S. Geological Survey published a series of articles on earthquake loss
estimation. The first article was published in 1980; it concentrated on earthquake loss studies for the
Los Angeles metropolitan area and the San Francisco Bay Area. Subsequently the authors developed
probable maximum loss methodology for the insurance industry.
2 ItaIian seismologist Mercalli developed the scale in 1902 to measure earthquake intensity. The scale
has a range of twelve degrees, from I to XII. It is a descriptive scale and based upon the observed
damage, an earthquake event magnitude can be estimated. It was modified in 1931 by Wood and
Neumann to fit the conditions in California. Since 1931, in the U.S. it is known as the modified
Mercalli intensity scale.
J ATC-13 - Applied Technology Council developed this report in 1985. The focus of the report is to
estimate the economic impact of a major California earthquake. The proposed methodology estimates
percent of physical damage versus 7 levels of earthquake intensity for 78 existing facility classes in
California, including 36 building structure classes. It tabulates damage loss factor estimates from more
than 70earthquake experts. The damage factor is defined as the ratio of dollar loss to replacement,
expressed as percentage. Experts provided a low, best and high estimate of damage at modified
Mercalli Intensities of VI through XII. Best estimate was defined as most likely damage factor whereas
low and high estimates were defined as 90% probability bounds of the damage factor distribution.
4 The Maximum Credible Earthquake defined here, is the maximum possible earthquake that could ever
occur on a fault system. Although the definition of Maximum Credible Earthquake has undergone
changes, the definition stated here is used in the CATALYST model.
5 To compare earthquakes worldwide, a measure which does not depend on the density of population
and type of construction was needed. The Richter Scale developed in 1935 measures the magnitude
based upon the peak ground acceleration. To convert Richter magnitude into MMI, the damage caused
by a particular magnitude earthquake has to be compared to the description of a particular intensity on
the modified Mercalli intensity scale.
6 This principle underlies the argument that projects that result in potential Pareto improvements should
be pursued, even though transfers to make the actual Pareto improvements cannot be made.
7 Rawls (1971) states that if people were all rational, caring only about self-interests and negotiating
with each other without the knowledge of anyone’s social or economic positions, special interests,
talents or abilities, processes could arrive at fair social institutions.
8 Axelrod (1984) argues that, for collective action to occur under repeated active plays, a strategy of tit-
for-tat works. Hardin (1982) states that collective action depends not only on the size of the group but
also on the ratio of costs to benefits as perceived by its members.
9 Raiffa further states “committees should continue to be concerned about risk assessments and leam
how to make better evaluations. A more experimental, adaptive societal approach, is needed, one that
remains loose, flexible and resilient.”
1 0 Building codes were subsequently changed to reflect this information as it became available.
119
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 1 March further claims that the fact that human beings exhibit greater risk aversion for gains than to
losses in a wide variety of situations may reflect accumulated learning rather than inexplicable human
traits or utility functions.
120
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 5
ATTRIBUTES OF ACCEPTABLE SEISMIC RISK, STAKEHOLDERS, THEIR
SELECTION AND RESOURCE CAPACITY OF A COMMUNITY
Previous chapters discussed acceptable risk to a community and its
dependence on internal and external variables in general. It is further
proposed, that acceptable risk is also influenced by the resource capacity of a
community.
This chapter discusses various attributes of acceptable risk, stakeholders, and
criteria for their selection, and the resource capacity of communities. It has
been argued by a number of scholars that acceptable risk is a political issue
(Clarke, 1989; Petak,1993; Otway and Thomas 1982), this study proposes that
the community must ultimately be responsible for determining acceptable risk
within the context of political considerations. The public must decide how safe
is safe enough, how safe is fair enough, how safe is voluntary enough and
how safe is equitable enough? (Shrader-Frechette,1993, pp.32)
ATTRIBUTES OF ACCEPTABLE SEISMIC RISK
Attributes of acceptable seismic risk are broadly divided into seven groups:
Life Safety Considerations, Business Interruptions, Infrastructure Damage,
Operability o f Critical Facilities, Economic Impact, Community Assets, and
Community Preparedness. Each of these attributes has an impact on the level
121
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of acceptable seismic risk by a community. Many of the attributes are
interdependent, but each attribute fails into one of the three components of
seismic risk; i.e., technological, economic, and societal.
Life Safety Considerations
Life safety of the public is of paramount importance and cannot be
compromised. The intent in this study is not to put a value on human life as
that is an extremely complicated and controversial issue. The study does
reflect the viewpoint that life safety considerations are an important issue
which must be adequately addressed. Although individuals may take greater
risks, the primary concern is to provide adequate life safety provisions through
regulatory mechanisms.
Life safety concerns are present when building structures are occupied,
whether the occupancy is on a permanent basis like residences or on a
temporary basis like that in the workplace or other places of communal
gathering. Building codes require that structures which house people be
designed to provide a safe exit for occupants in emergencies. This provision is
interpreted by the design profession to mean simply, that a structure should
not collapse or be damaged to such an extent that occupants cannot exit
safely.
122
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Because various structural systems have shown deficiencies in withstanding
an earthquake force in major earthquakes, those buildings constructed with
these structural systems cannot guarantee a safe exit opportunity for
occupants. The structural ability of these buildings to withstand major
earthquake forces needs to be augmented. Depending upon the number of
such structures in a community, the overall risk to life will vary.
The cost incurred to retrofit these deficient structures can be considered a
“cost to provide life safety.” One of the decisions community stakeholders
have to make, within the community’s resource capacity, is trade-offs among
other needs and cost of life safety. Regulations to mandate a retrofit of
buildings of a certain age help provide life safety; however, such regulations
impose costs on owners of those facilities. Voluntary retrofit efforts instead of
mandatory requirements may reduce costs, but will also increase life safety
risk as not all buildings deficient in seismic safety resistance will be upgraded.
Life safety considerations need to be informed by an objective assessment of
existing facilities and costs to retrofit them.
Business Interruptions
Business operations in a community suffer as a result of damage to their
facilities, damage to transportation infrastructure, damage to utilities, and the
impact an earthquake event has on the health of employees. As can be seen,
this attribute is dependent upon the functionality of supporting systems in
123
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
addition to damage to business facilities. Acceptable risk due to potential
damage to business facilities can be determined by business owners.
However, business interruptions due to damage to support infrastructure are a
community dependent aspect.
In addition to the cost to the owner of the business resulting from damage due
to an earthquake, the community also incurs cost due to loss in wage-related
expenditures and taxes. One of the contentious issues to be resolved
between business owners and support system decision-makers is the duration
of in-operability of infrastructure. For example, a business owner may upgrade
his facility to limit the business interruption to three days; however, the
transportation infrastructure necessary to get his product to the market place
may not be repaired for a week. Such a situation results in business
interruption beyond three days. On the other hand, if the community has to
guarantee serviceability of supporting services within three days, the
community will incur additional cost.
Community stakeholders ought to discuss these issues pertaining to this
attribute and reach an understanding acceptable to stakeholders, which
include business owners.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Infrastructure Damage
Infrastructure damage can create havoc to economic and social life in a
community. Infrastructure can be classified in three broad areas; utilities,
transportation, and communications. A disruption to any of these creates
different levels of economic loss and social disruption.
Damage to utilities i.e. electricity, gas, and water brings personal, business
and public life virtually to a standstill even though for a short duration. In
addition, breakage of gas lines and electricity transmission lines can result in
fires, and breakage of sewer lines creates problems related to contamination.
A community needs to decide the acceptable duration to have non-functioning
utilities. In emergency preparedness guidelines, having basic supplies to last
72 hours is a usual criterion; however, each community can have its own
criterion of acceptability. Once the criterion is agreed upon, cost to ensure
services within the specified period has to be assessed and discussed with
utility companies. Current deregulation of utility services has further
complicated this issue.
Damage to transportation networks can result in a community being cut off
from other communities if the main transportation arteries are inaccessible.
Generally, in a network of transportation alternate routes are available;
however, geographic specificity and network connection with surrounding
communities impacts the acceptable level of non-functioning of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
transportation infrastructure. A community may be heavily dependent on state
and federal authorities if state and federal roads provide access a community.
Transportation networks generally involve many communities and a dialog
with other communities is essential.
Communications network damage can result in societal panic due to feelings
of isolation. However, given the current state of information technology
advances several alternate sources of communications are available.
Availability of such alternative sources may reduce the feeling of isolation.
Operability of Critical Facilities
Critical facilities that are necessary to remain functional during and after an
earthquake fall into two categories: those that are considered emergency
operations centers and those the community desires to remain operational.
Facilities that fall under the first category are typically: fire stations, police
stations, emergency management centers, schools and hospitals. Those
facilities except hospitals and schools, which are state funded, are designed to
code provisions that require them to remain operational during and after a
major earthquake event (Essential Services Act of 1986). These facilities
which are not state funded may be designed to comply with local building
codes. School facilities are used as emergency shelters and are designed to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Field Act provisions ensuring higher standards of seismic safety. Hospitals are
also designed to higher seismic safety standards.
The facilities that fall under the second category are, for example: community
nursing care facilities and senior citizen housing. Among reasons that can be
cited for the desirability of having these facilities remain functional during and
after an earthquake are: community nursing care facilities house patients who
cannot exit safely without assistance, and the necessity to continue care of
patients. Although hospital building regulations require continuous operation
of facilities, the regulations are written for new facilities. Older facilities built
prior to the 1972 Hospital Safety Act do not conform to these regulations.
Currently enacted law under California Senate Bill, SB1953, requires all
hospitals to provide continued operability during a major earthquake event by
the year 2025. The cost of such an endeavor is not known. Although private
hospitals will have to bear the costs of upgrading, community-owned hospitals
have to be retrofitted with community resources.
Public schools designed to the Field Act do not require continuous operability.
In past earthquakes, public school facilities have been used as emergency
shelters for those inhabitants whose residences were damaged by an
earthquake event. It is prudent to expect that these facilities can be used for
community gathering purposes. It is in the interest of the community at large
to retrofit older school facilities so they can be used for these purposes.
127
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
People who reside in senior citizen housing facilities need assistance in
varying degrees for their daily functions. In general, this segment of the
community needs to be taken care of by others. It is the moral responsibility of
community stakeholders to provide needed assistance in the form of having
these facilities functional with minimal damage.
All of the considerations discussed under “Operability of Critical Facilities”
have costs associated with them. Community stakeholders need to consider
this attribute carefully in formulating acceptable seismic risk.
Economic Impact
All damages as a result of a major earthquake event result in economic
losses. Direct economic impact is generally viewed as that resulting from
damage, i.e. cost of repair and replacement. Indirect economic impact results
from the effects of direct economic impact, e.g. business interruptions and loss
of wages.
To a large degree, direct economic effects are quantifiable. Apart from
economic losses suffered by owners of physical property that are perhaps
reimbursable by their insurance companies, the reinvestment needed to repair
community infrastructure must also be assessed.
128
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Aggregation of physical property damage losses, and retrofit costs of all public
utilities, transportation and communications networks can be used as a proxy
for direct economic impact. However, all economic effects can not be
foreseen, therefore the estimate is approximate at best, but it does provide a
measure of potential economic impact.
Indirect economic effects result from erosion of the business base because
businesses decide to move to other communities, adjustments to changes in
demand and supply, losses in production, reduction in expenditures and
reduced health and social services in the community. After Northridge
earthquake, due to shortfall between the total economic losses and outside
aid, the Los Angeles area consumer spending was projected to have been
depressed by 0.3-1.0% from a no-quake baseline costing 20,000 to 60,000
jobs. (Romero and Adams,1995, p.263)
Since indirect effects are difficult to quantify, a debate by stakeholders in the
community needs to take place to consider this attribute along with others so
that an informed decision can be made based on knowledge of the effects.
Community Assets
It is important to preserve certain community assets that are deemed
important by stakeholders. Some assets may be non-replaceable, and a
decision to preserve these cannot be made on an economic value basis only.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Examples of such assets include art collections in a museum, historic
buildings and unique man-made facilities specific to a community. Community
stakeholders have to make a list of such assets and decide through an
informed debate, those assets that need to be preserved and the cost
associated with such preservation.
Community Preparedness
This attribute is one of the most challenging and encompasses several
aspects of community behavior. A community which is better prepared to deal
with a hazard minimizes its impact thus reducing seismic risk. Community
preparedness is a subject by itself that is not discussed here. However,
certain aspects of the preparedness are presented. Ready and able voluntary
groups to provide social and paramedical services and the ability to set up
emergency medical clinics, alternate routes of delivery of services, search and
rescue crews, and emergency communications network comprise prime
components of preparedness.
Financial and manpower resources are required to undertake these
emergency preparedness measures. Each city and county in California is
required to set up an emergency management center. This community
preparedness attribute must be considered seriously as it can relieve some
societal tensions and reduce societal disruptions as a result of an earthquake
event.
130
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
STAKEHOLDER ANALYSIS
To determine acceptable seismic risk, preferences of stakeholders in the
community have to be known and a procedure of rank-ordering them needs to
be developed. In this study, the methodology proposed to determine these
preferences is through expressed opinions o f experts from each of the
stakeholder groups. A questionnaire is used for this purpose. The method
assumes that expert opinions are a good measure of expressed preferences
and are a reliable indicator of acceptable risk (Fischoff,1983). This method
helps in deciding the rank order of preferences of stakeholders taking into
account their value judgements. The details of the methodology are discussed
in chapter 6.
Because aggregation of various utility functions across numerous actors is
extremely complex unless we know who these actors are, a basic question to
be resolved is; whose preferences should be considered? This study identifies
seven groups of stakeholders whose preferences ought to be considered. The
selection of stakeholders is discussed in the next section. It is posited that
these groups represent the technological, economic and societal interests of a
community.
Once the preferences are rank ordered, appropriate weight needs to be given
to each preference. Clarke (1989) defines five steps in determining acceptable
131
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
risk. His focus, however, is on organizational decision-making rather than
community decision-making.
Preferences can be assigned to the three components of risk: technological,
economic, and societal. The combination of these three components will
determine the risk acceptable to a particular community. Acceptable risk
depends upon who owns what, economic value and tolerable levels of
inoperability of lifeline support systems.
STAKEHOLDER SELECTION
Selecting stakeholders who represent a community is critical. Since the
composition of a community varies, it is not easy to determine the groups that
should be identified so that one obtains a fair representation of the community.
A stakeholder must represent one of the following areas:
1. Serviceability of community infrastructure and utilities.
2. Emergency preparedness and recovery effort.
3. Physical property ownership.
4. Technical expertise in assessing potential damage to physical
facilities.
5. Economic base of community.
6. Representation of societal interests.
132
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
It is entirely possible that depending upon the composition of a community, the
areas represented can be expanded and considerations particular to that
community can be added. To qualify, a stakeholder must have influence in two
areas. The qualitative rating must be high at least in one area and medium in
at least one other area. See Table 5-1. For a typical community, groups of
stakeholders can be identified based upon these criteria. An informal survey
was conducted among loss estimation experts. Based on their judgments,
qualitative ratings of stakeholder influence have been determined. Although a
stakeholder group may have expertise in an area unless their influence
impacts the decision on seismic risk, the group may not score high; e.g.
engineering expertise is high in designing utilities and infrastructure but their
influence in making them seismically safe may not be high.
Groups of stakeholders cannot be neatly separated as these individuals are
interrelated, and actions undertaken by one group impact other groups. This is
the prime reason why input is required from all identified groups to formulate
acceptable seismic risk. Although the groups are interrelated, they are
discussed here separately in terms of actions related to their own decision
making processes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 5-1. INFLUENCE OF STAKEHOLDER GROUP
Community
Infrastructure
& Utilities
Medium Medium
Physical
Property
. Medium Low Medium ' High Low . Low Low'
Ownership
■ Economic B aisai
of Community
High High / Medium . High Medium Medium Low
Stakeholder Groups
The groups of stakeholders that can be Identified are: controllers of financial
resources, political decision-makers, regulatory agencies, physical property
owners, lifeline support system providers, engineering and scientific experts,
and rescue and relief agencies. Each group is discussed along three
dimensions: Organizational or associational structure, behavior, and
interdependencies with other stakeholder groups.
134
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Controllers of Financial Resources
Financial resources are controlled by three distinct entities: Government,
private businesses, and financial institutions. The government controls public
revenues, and private businesses control resources within their organizations.
Financial institutions are private lenders to businesses and can augment their
financial capacities by collaborating internationally, if required.
Government
Government expenditures occur at state and local levels. At the state level,
legislators control allocations of resources through their annual budgetary
process or emergency budget allocations. Allocation of funds for seismic
retrofit or other seismic mitigation goals has to be judged along with priorities
for numerous other programs. Many businesses in small towns reported not
getting as much attention from government relief agencies after Northridge
earthquake, because their communities did not have same political power as
larger towns(Romero and Adams,1995, p.31). Incentives to vote for allocation
of significant funds for seismic retrofit of public buildings are almost non
existent due to the lack of a statewide constituency. Since the governor has to
ultimately approve and sign the annual budget, the governor can also
influence allocation of resources for seismic mitigation. However, unless this
issue is perceived as politically important, little motivation exists to include
funds for seismic mitigation in the budget.
135
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In 1993, the California legislature enacted a law to assess and retrofit public
buildings. The perception was that workers in these buildings needed life
safety protection in the event of a major earthquake since the buildings were
old and were built to older codes. A $250 million bond was passed to finance
this program and the work of retrofitting state buildings is underway. This
dollar amount is woefully inadequate to retrofit nearly 14,000 state owned
buildings. Although a beginning has been made, funding of such programs is
unusual. One of the main responsibilities of the state government is to assist
communities in their recovery efforts after a major earthquake event. Since
resources are limited and recovery efforts are expensive, state government
needs to stress investing in ex-ante mitigation efforts and increasing the
resource capacity of communities.
Based upon public interest considerations it is the responsibility of the
legislature and the executive branch to address issues of societal importance
for the good of the public at large. Political leaders must rise above election
dominated actions.
Private Industry
Private industry is a large and complex conglomeration of various interests.
Members range from large businesses to very small businesses, and from
heavy manufacturing industries to high technology service industries. It is not
possible to discuss private industry as a unit. However, there is an underlying
136
of the copyright owner. Further reproduction prohibited without permission.
common feature to the organizational behavior of businesses that stems from
a singular consequence; i.e. “adverse impact on business” due to a major
seismic event. The impact varies with the size and location of a business, the
type of business and the type and range of its products or services.
Some small businesses may suffer damage so extensively that it becomes
impossible for them to continue in business afterwards as their assets are
wiped out, and their ability to restart their business is limited. Many small
businesses do not own their buildings and therefore have to depend on
building owners for mitigation. These businesses cannot afford the financial
consequences of a major seismic event; e.g. many small business entities
went out of business following the 1994 Northridge earthquake event. (Eguchi
etal., 1997)
Both large and small businesses are interested in limiting the duration of
interruption, damage to their real property assets, business assets and
inventories. If the product of a firm depends on sources outside the impacted
area for inputs and outputs, the effects of such externalities need to be
considered.
Direct interdependencies of large businesses are with government agencies,
and suppliers and vendors. Their indirect interdependency is with consumers.
For small businesses, the interdependency is directly with their consumers
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and suppliers. The actions taken by business entities are influenced by such
considerations.
Adverse impact on business is ultimately a financial concern, and all actions
by businesses emerge from that concern. Large business operations can
transfer risk by purchasing insurance, but smaller businesses may not be able
to afford the insurance premiums, even if insurance is available to cover all of
their losses; so they bear the risk of major damage. Since the concerns of
private industry cannot be heard through a single voice, it is extremely difficult
to determine who has standing to assess acceptable community risk.
Impacts on businesses result not only from disruptions to their own operations
but the consequent impact on the community; e.g. if a business is inoperable,
people are out of work and the economic base of the community is impacted.
The opinion of businesses in a community towards acceptable risk should be
taken into account. Business losses in the community should be aggregated
and the overall impact on the business community should be considered in
assessing the standing of businesses.
Financial institutions
Financial institutions operate in a competitive market environment. Their
primary business is to lend money and derive returns on their loans. Large
financial institutions are aggressive lenders as they can absorb more risk due
138
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
to their large and diversified asset base. These large institutions are also in a
position to distribute risk among other financial institutions. Smaller financial
institutions on the other hand tend to be locally based and more conservative.
These institutions also take pride in being recognized as part of the community
and are likely to develop close relations with business owners. In either case,
i.e. whether a large or a small financial institution is the lender, it is the ability
of businesses to repay their loans that is important. The prior experience of a
financial institution with the business owner is important in determining future
lending. For physical property improvement, loans can be secured with the
property as collateral. It is in the interest of lenders that damage to structures
housing businesses be minimized to protect their investment.
In California today, it is not unusual for a lending institution to ask for a
“probable maximum loss report”, prepared by a seismic expert before loans
are committed. The requirement for a seismic report varies with each lender.
Loans on commercial buildings generally require such a report. If the loan is
for business operations or for building inventories, continuation of the business
after a major seismic event is even more important. Financial institutions do
have significant standing in determining acceptable risk. The insurance
industry has a dual function. They are lenders as well as protectors of portfolio
assets.
139
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
What should be the standing of these controllers of financial resources in
assessing acceptable seismic risk? Since after every major seismic event
communities are financially impacted, opinions of controllers of financial
resources are important and must be taken into account.
Homeowners’ concerns are limited to the protection of their personal assets
and their immediate family members. These owners are not considered
controllers of financial resources, but homeowners are impacted directly and
suffer losses in major earthquake events and therefore their opinions also
need to be taken into account. The opinions of homeowners are represented
best by elected local government officials and regulatory agencies.
Political Decision Makers
This group of actors is comprised of state legislators, local government
officials, and the federal government.
The discussion in this section is focused on political action rather than control
of financial resources. If the financial aspect is separated from political action,
motivation for these actors to act is “self interest rightly understood ” as stated
by Tocqueville in the nineteenth century.
Since the interdependencies of elected representatives are with the electorate
and the executive branch, the decision-making process for seismic risk can be
140
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
considered in terms of a theory of exchange. According to this theory, an
exchange takes place only when parties to the exchange perceive that they
have benefited by exchange. Exchange theory posits that behavior is a
function of various material and non-material inducements that are offered.
California state legislators are of the opinion that life safety related seismic risk
has been addressed by building codes and any additional legislation to
mitigate seismic risk would cost more than the consequent benefits derived.1
The Federal government under its disaster emergency policy provides disaster
assistance to local and state governments after a major disastrous event. The
Federal government has realized that the cost of assistance has grown
dramatically in recent years (PEER, 1998) although some may consider this
assistance as a redistribution of locally collected tax dollars. The Federal
Emergency Management Agency (FEMA) has recently developed a
public/private partnership strategy to redefine its involvement in mitigating
seismic risk
Local elected officials make decisions related to their particular community.
Their decision-making authority is moderated by financial resources in the
community and the state and federal government policies of assistance.
Local political decision-makers, however, can significantly influence the
141
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
regulatory environment, community asset preservation policies, and
community preparedness.
All participants who have political decision-making authority and responsibility
need to act with a common approach. “We need to develop a common goods
approach, wherein the individual good is inextricably bound up with the
community good as a whole.” (EERI,1997,p.10)
Political-decision makers have important standing. Their opinions result from
considerations of the economic capacity of a community, the extent of likely
state and federal government assistance and political and social
considerations. For example, if a community has a bigger capacity to absorb
seismic risk, assistance by the state and federal governments may be
reduced. Political decisions at the state and federal government level for local
assistance can be based on ‘community specific’ needs.
Regulatory Agencies
Regulatory agencies play a critical role in establishing the level of seismic risk.
Generally, the rcie of regulatory agencies is to enforce laws enacted by the
legislative branch of the government. In the area of seismic risk however, very
few laws exist and the general health and welfare of the public is addressed
through building codes for the design and construction of buildings. The
American Association of State Highway and Transportation Officials
142
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(AASHTO) regulations govern transportation structures, and various codes
written by other professional societies, such as the American National
Standards Institute (ANSI) and the American Society of Mechanical Engineers
(ASME) also address the general health and welfare of the public.
Model building code 2 provisions are written by professionals who are experts
in their specific fields. The provisions are generated as a result of deliberations
and discussions in professional associations of specific disciplines. These
provisions, as a written body of material, are submitted to these professional
organizations comprised of city and county building officials, to vote on and to
adopt for enforcement. The code provisions can be modified and enhanced by
local governments to suit their specific needs. In California the intent of the
provisions cannot be lessened.
The organizational structure of regulatory agencies is primarily dictated by the
need to enforce code provisions. A hierarchical structure is headed by a
building official reporting to the city council. Various enforcement and oversight
staff personnel report to the building official. This structure is used in
regulating private buildings of various occupancies. Public buildings do not
necessarily go through the same process of oversight. Conduct of staff in
regulatory agencies is governed by their duties and responsibilities in
enforcing laws and regulations. These personnel do exercise considerable
power in granting or delaying building permits.
143
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Interdependencies of regulatory agencies such as code enforcement entities
are many. These agencies coordinate development of regulatory provisions
with other jurisdictions, depend on guidance from city councils or officials of
local government and interact with the public and special interests in their
community. The regulatory agencies need to balance these different interests
while keeping the mandate of protecting the heath and welfare of the public.
After a major earthquake event, regulatory agencies are also called upon to
assess damage to buildings and structures and to take appropriate actions to
protect the life safety of the public.
Regulatory agencies have a significant stake in determining acceptable risk;
their opinions are best considered in modifying code provisions. Regulatory
agencies can also influence local government officials in enacting laws to
mitigate seismic risk.
Physical Property Owners
The focus here is the group of people who are physical property owners;
however, homeowners are not considered a part of this group. The members
of this group are investors in residential properties and business real property.
This group is important because physical property represents a significant
portion of the private assets in a community. Revenues for the community are
generated from taxes on these properties. More importantly, severe damage
to properties will result in economic dislocations beyond the immediate
144
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
damage due to business interruptions, a need to provide temporary housing
for renters and a lack of income for those who work in businesses. These
dislocations will have an overall effect on consumer buying power.
Behavior of this group of investors varies with size, location, and type of
property and the remaining economic life of the properties under
consideration. This group of property owners has very little interdependency
with others for decision-making purposes.
The utility function of each owner is different as owners attach different
weights to property specific variables. For example, for an owner of a rental
housing complex where the majority of the mortgage is paid, the main source
of income may be rental income. In that case, the owner is likely to protect the
investment even though the building may be old. That property owner’s level
of acceptable risk would be smaller than an owner who has a large unpaid
mortgage. How can these diverse utility functions be aggregated?
Perhaps one way to aggregate the risk to physical property owners in a
community is to determine the non-insured part of the assets. An owner may
have purchased insurance to cover $20,000 in losses with a deductible of
$5,000 (the maximum amount the owner is responsible for). Aggregation of
these amounts will indicate total acceptable risk by this group in a community.
However, the acceptable risk by those who are impacted by damages to
145
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
physical properties may be different. Regulatory agencies will view a
deficiency in a structure as a potential seismic risk, and residents would
demand a safe exit in case of significant structural damage. The opinions of
physical property owners on acceptable seismic risk have to be modified by
these other considerations.
Life Line Support System Providers
This category of stakeholders includes utility companies that provide water,
wastewater, gas and electricity services, communications and transportation
networks.
This group represents a combination of private utility providers and public
entities responsible for transportation networks. Because privately owned
utility companies are regulated by government agencies, their structures and
hierarchies tend to be similar to those in the public sector. However, the two
sectors operate under different sets of goals, missions, and policies. Actions of
private utility companies are strongly motivated by return on investment and
maximizing the value of company stock. Conversely, public entities like the
department of transportation in California are mandated by statutes to provide
certain transportation related services to the public at large.
Most utility companies are not local community owned; they usually service a
wide geographical area. However, their distribution network in a community
146
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
may be locally owned, particularly in the case of water systems. A community
may be adversely impacted due to inoperable utilities even though the
earthquake effects may not be felt locally. The damage to the transmission
network may be at some distant location from the affected community. The
distribution network may also be severely damaged at some distant location
resulting in an inability to service local community customers.
Similar situations exist in transportation networks. State highways are under
the transportation department’s control, but operation and maintenance of
local roads are left to cities and counties. Surface inaccessibility could result
from damage to these networks anywhere along their various routes.
The interdependence of utility providers and transportation system providers is
heavy; e.g. a utility line may be damaged due to damage to roads; or because
of damage to roads, the utility companies may not be able to repair their
service facilities/equipment.
Acceptable seismic risk has two components in this case. First, the amount of
risk these providers can absorb, and second, the amount of risk a community
should accept for the limited operability of these systems. The first component
of risk is decided by each provider; the community does not exert much
influence on these decisions. However, decisions taken to address the first
component of risk by these providers do impact the economic and social life of
a community.
147
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The tolerable level of service interruptions ought to be determined in a
community based on public information and debate by the public in
conjunction with providers of these services. Lifeline system providers have to
work with community customers to agree on the acceptable level of service
interruption.
Engineering and Scientific Experts
Engineers and scientists have been studying the problem of earthquake risk
for quite some time. Through laboratory research and observations of actual
behavior of buildings and other structural systems in major earthquake events,
considerable advances have been made in understanding the technological
component of seismic risk. Laboratory research is concentrated primarily in
technical disciplines such as structural and geo-technical engineering.
From an organizational perspective, engineers and scientists operate as
individual entities providing technical expertise. Although some professional
associations, like the Earthquake Engineering Research Institute (EERI), the
Structural Engineers Association of California (SEAOC) and Building Seismic
Safety Council (BSSC) are engaged in promoting seismic safety, the seismic
engineering discipline is practiced on an individual, technical expertise basis.
SEAOC has developed and updated the Blue Book, which is considered an
authoritative document on seismic analysis and design of building structures.
148
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The behavior of engineering experts is primarily governed by the needs of
clients for their expert technical advice. Due to the individualistic, professional
nature of the discipline, many different opinions exist on seismic risk; technical
issues are debated in meetings of related professional associations. Through
such debates, building code provisions are generated and eventually enacted
by model code authorities although the code provisions may be modified by
local building officials. Interdependencies are limited to client-professional
relationships and interaction of the professionals within professional
associations. Society in general, public policy makers and business owners
depend on engineering and scientific experts to determine seismic risk and to
write regulatory provisions into model codes to mitigate that risk.
Engineers and scientists have considerable standing in defining seismic risk.
At present seismic risk is defined as “life safety”. As an engineering definition,
life safety” means, that during the most probable major seismic event, the
structure will perform adequately for occupants to exit safely.3 The life safety
concept is not directly related to overall building damage; it assumes that
although the structure may be damaged significantly, it will remain standing
and allow occupants to exit safely. Other risk levels are immediate occupancy,
continued operability and collapse prevention. However, the basic emphasis is
on life safety considerations.
149
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
SEAOC has made efforts to define structural system performance objectives
including life safety (SEAOC, 1994, p.8). The concept is that, based upon the
requirements of an owner, a structural engineer can guide design efforts to
meet the level of building performance objectives desired by the owner. The
concept is powerful, but it has problems in real life applications. For example,
should a business owner define his performance objective as a maximum
business interruption of two business days, structural engineers can try to
design the structure with added levels of safety so that operations are
interrupted minimally, but they cannot guarantee such a definitive
performance. At best, a certain higher level of probability can be generated;
e.g. the performance objective has an 80% chance of being realized. There
are also numerous professional liability issues that need to be resolved before
the concepts can be put to use in practice.
In a community engineers and scientists should be called upon to estimate the
potential probable damage levels to various structures and consequent costs
associated with their repair or replacement, but the “acceptable” magnitude of
damage ought to be left to a decision by the community. Engineers and
scientists have considerable standing in determining the technological
component of risk.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Rescue and Relief Agencies
This group of actors is active after a major seismic event occurs. These actors
are not involved in determining acceptable seismic risk, but they suffer from
the consequences of the acceptable seismic risk determined by others.
Although it might appear that the greater the acceptable risk, the lesser the
demand on rescue and relief agencies; but quite the contrary is the case.
Because property owners and the community as a whole are prepared to
accept larger risk, a greater degree of damage to facilities and a greater
overall impact on the community may ensue. The demand on rescue and relief
agencies therefore, increases.
The majority of agencies that provide rescue and relief operations are non
government agencies. These agencies do not have a direct stake in damage
losses. Some organizations have the resources and capabilities to strengthen
their workforce by assistance from other agency offices. However, this is not
normally the case with relief agencies that do not operate statewide or
nationwide.
To a greater degree, the organizational behavior of these agencies is driven
by ethical and moral considerations, community standards, legal responsibility,
politics, and their dedication to help others in need. Many of the workers are
volunteers. Some of these relief agencies are supported financially by states
or the federal government. However, the behavior of even these agencies has
151
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
been demonstrated to be acting first and collecting expenses afterwards.
Relief agency interdependencies are with other relief agencies; they must co
ordinate their efforts and with government agencies that provide funding.
Those relief agencies that receive funding from the government tend to restrict
their rescue and relief efforts due to constraints associated with funding for
rescue relief operations.
Other groups involved in rescue and relief operations are: police and fire
services, emergency operations centers, and medical facilities. However,
since fire and police services are usually a part of the local government, they
are not discussed separately; their mission is to provide services within their
capacity irrespective of immediate cost. Cooperative arrangements with other
fire service agencies and police departments are also common, boosting their
service capacity. Police agencies are also not discussed separately as their
goals and missions are similar to those of local and state governments and are
not cost driven.
Medical facilities deserve a separate discussion. The majority of these facilities
are privately owned; they behave as immediate caretakers, but eventually
collect a fee for their services. The overall capacity of medical facilities to
service the injured is important.
152
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Establishing local emergency operations centers is a fairly recent undertaking.
Each community center is expected to coordinate its activities with the Office
of Emergency Services of the State of California. Although rescue and relief
agencies play a major ex post role after the event, their standing and opinions
in determining acceptable seismic risk are weak.
There are two other entities that are not part of any group but have an
influence on the level of seismic risk: the California Seismic Safety
Commission and the Federal Emergency Management Agency.
California Seismic Safety Commission
California Seismic Safety Commission (CSSC) is an advisory body established
by the legislature to advise them on overall statewide seismic safety policy.
CSSC is interested in overall public policy dealing with seismic safety issues
and the impact of these policies on the population of California. The Governor
appoints the members of the commission. The members are supposed to
represent various interest groups and professional disciplines. Representation
on the commission is also given to both houses of the state legislature.
Although, the commission is an independent body, in reality it is influenced by
the political views of the current administration since the Governor appoints
most of the members. The Commission reports directly to the Governor;
regularly hears testimony on seismic issues from various public interest
153
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
groups, scientists and engineers; and proposes legislation related to seismic
safety.
Since CSSC is an advisory body, its decisions and findings are not binding on
any agency. As the commission has no enforcement power, its
recommendations go largely ignored. CSSC has no cooperative or exchange
relationship with any particular entity; their interest in the seismic safety of
structures and buildings is morally, ethically and legally driven. Since it is an
independent commission, its findings and recommendations bear the mark of
objectivity and are not viewed as special interest driven.
Federal Emergency Management Agency
FEMA influences policy decisions in two distinct ways: generating ‘building
code type’ documents and providing financial assistance after a major seismic
event. For example, after the 1994 Northridge earthquake in California FEMA
provided nearly $12 billion funding to various public facility owners
(EERI.1996). Some small grants were also provided to individuals.
The entity, which creates ‘building code type’ documents, is comprised of
nationwide seismic experts and is called the Building Seismic Safety Council
(BSSC). Various projects of BSSC are funded by FEMA. Provisions prepared
by the NEHRP (National Earthquake Hazard Reduction Program) influence
model building codes in the nation. Thus, acceptable performance of
154
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
structures is addressed indirectly by influencing the technological component
of risk. The major focus in these documents is on “life safety.”
Because FEMA provides disaster relief assistance (under a separate act of
congress) after a declared major disaster, acceptable risk by communities is
also directly influenced. The more funding that is given by FEMA, the less the
desire in a community to develop a risk absorbing capacity. Communities will
have to bear more risk in future major seismic events as less funding for
assistance will be available from FEMA and more strings will be attached to
such assistance. FEMA influences acceptable seismic risk indirectly through
its disaster assistance funding authority.
It is clear that the motivation for each group of stakeholders is different and is
influenced by economic, political, and technological considerations. How
should the standings of various stakeholders be accounted? It is not possible
to propose a uniform method to combine these disparate weightings. Table 5-
2 identifies the relative impact of each group on the three components of risk.
The extent of impact (weight) is defined in qualitative rather than quantitative
terms.4 The level of impact is defined into three categories: weak, modest and
strong. Table 5-2 allows us to identify the importance of the standing of
different groups in attempting to determine acceptable risk. Different
stakeholders can be identified and advice can be sought from them in the
category of their expertise.
155
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 5-2: IMPACT OF STAKEHOLDERS ON RISK COMPONENTS
Controllersof
FfnaitcialResources
Regulatory Agencies
LifelineSuppbrt
System Providers
Rescue andRelief
Agencies '
Weak Weak
RESOURCE CAPACITY OF A COMMUNITY
After acceptable seismic risk is determined, the cost to bear that risk has to be
weighed against the resource capacity of a community. Acceptable seismic
risk is an independent as well as a dependent variable and therefore difficult to
establish without an iterative process. The resource capacity of a community is
comprised of: financial resources and human resources. Total resource
capacity is a combination of the two. An economically well off community or
one with many businesses in its jurisdiction may have larger financial
resources available than a community with fewer businesses. However, a
larger but economically poor community can provide human resources to
assist in post disaster operations.
Finally, acceptable seismic risk as an independent variable and as a
dependent variable will have to be assessed against the resource capacity of
156
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
a community. Based upon resource capacity, acceptable risk may have to be
modified through an iterative process.
ECONOMIC CAPACITY
One needs to consider the economic capacity of a community to undertake ex-
ante mitigation and community preparedness efforts prior to the event and
economic recovery capability, after the event. The damage scale presented in
Figure 5-1, is conceptual in nature but demonstrates that damage due to
seismic hazard risk is nonlinear. Risk is plotted as a probability of different
magnitudes of earthquakes. The scale denoting economic capacity is also
conceptual in nature. The economic capacity of a community and therefore
the cost of mitigation it can bear is independent of risk level.
Based on the curve presented in Figure 5-1, it can be noted that depending
upon the economic capacity of a community, only some risk levels will be
mitigated with ex-ante actions alone. As the risk level measured in terms of
damage potential increases, the capacity to fully mitigate an earthquake’s
effects fall short.
Consider a community likely to experience an earthquake event with a
probability of occurrence of 0.002.
157
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
E
o
____________________________ Community e
f Preparedness O
o
. . . m
______________________________ Ex-ante
E Mitigation
Measures
1 in 50 1 in 1 0 0 1 in 250 1 in 500 1 in 1 0 0 0 1 in 2000
years years years years years years
Earthquake Occurrence
Resource Capacity o f a Community
Figure 5-1
If Community A has invested in ex-ante mitigation measures represented by
point A on the potential damage curve, and community preparedness is
virtually nil, the potential cost to bring the community back to its pre-event
stage would be represented by area bounded by points A, E, and D.
On the other hand, if community invested further in mitigation measures to
bring the damage resistant level to A’, the cost to bring the community to its
pre-event stage would be reduced to an area bounded by A’, E’, and D. For
158
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
an investment increase from E to E \ the reduction in future damage repair
cost is significant Furthermore, if the community invests financial and
manpower resources in community preparedness, the damage resistance
level is raised to point B and the potential cost to the community is reduced to
an area bounded by A’, B, F, and E1 . Because the damage potential curve is
steep, the consequent reduction in repair costs is significant with small
increases in investments in mitigation measures and community
preparedness.
The amount of investment a community can afford to make towards these risk
reduction measures is a matter for stakeholders to decide. A decision
methodology for considering stakeholder opinions is the subject of the next
chapter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Notes
1 Based upon numerous private discussions between the author and the key legislators of the California
legislature.
2 Currently three model codes exist However, from the year 2001, only one building code for the
United States, to be known as the International Building Code will exist
3 The basic life safety considerations are described in the Uniform Building Code 1997 edition.
160
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 6
PROPOSED METHODOLOGY FOR ANALYSIS: THE ANALYTICAL
HIERARCHY PROCESS (AHP)
The previous chapter discussed the attributes of acceptable seismic risk and
criteria for selection of stakeholders. This chapter focuses on three items;
identification of experts who are considered suitable for representing
stakeholders, Multi Attribute Decision Analysis methods (MADA), and the
Analytical Hierarchy Process (AHP), which is a particular type of MADA. AHP
is used in this study to analyze expert opinions.
IDENTIFICATION OF EXPERTS
Experts are selected based upon their knowledge and experience in one of
the critical areas identified in the previous chapter or based upon their position
and authority to influence decisions related to the health and welfare of the
community.
Experts in the following areas are identified:
Community emergency preparedness, civil engineering related to
infrastructure and utilities, structural engineers identifying technological risk
and as advisors to physical property owners and controllers of financial
resources, owner’s representatives, city managers and public works officials.
161
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(Descriptions of their associations, qualifications, and positions are provided in
Appendix ‘A’ without identifying their names).
A questionnaire (discussed in the next chapter) eliciting their opinions was
sent to experts in each area. The questions covered: life safety, physical
property damage, business interruptions, operation of critical facilities,
economic aspects, community assets and community preparedness.
Decision-making is a process of choosing among alternative courses of action
in order to attain goals and objectives. Nobel Laureate Herbert Simon wrote1
“that the whole process of managerial decision-making is synonymous with the
practice of management.”
Irving Janis 2 provided evidence in his book, Crucial Decisions, that “A poor-
quality decision-making process, characterized by simplistic strategies, is
more likely than a high-quality process to lead to undesirable outcomes.” He
further states, that “...when stakes are high, executives use analytical
techniques for solving problems.” Such techniques, though, depend on
quantitative methodologies and do not account for qualitative factors so
important in vital decisions.
When faced with a complex problem, a decision-maker has many factors to
consider in selecting the best alternative solution. Making tradeoffs between
162
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the objectives relating to a decision is a difficult and poorly understood aspect
of decision-making. The greater the number of objectives, the more complex
the decision process.
Multi-criteria decision-making involves structuring a problem and thinking
through many issues to establish a solution. The analysis involves analyzing
multiple tangible factors that are diverse and even conflicting, and intangible
factors such as a decision-maker’s subjectivity. Both tangible and intangible
factors have to be integrated together in the final decision. All decisions
involve both quantitative and qualitative factors. The ability to synthesize
quantitative and qualitative factors in a decision is extremely important.
Multi-attribute decision methods facilitate such decision-making processes.
MULTIATTRIBUTE DECISION ANALYSIS (MADA) METHODS
Multi-attribute decision analyses (MADA) methods consider both qualitative
and quantitative non-financial attributes in addition to common financial
measures (Norris and Marshall, 1995, p.2). Such methods are particularly
helpful when decisions are based upon expert opinions which are qualitative
and go beyond financial considerations.
Capital investment decisions are often evaluated using traditional economic
measures, such as net present value, life cycle cost, rate of return, or
payback. These economic worth methods determine the costs and benefits of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
an investment over a period of time. The American Society for Testing and
Materials (ASTM) has published standard methods for calculating economic
worth measures in the evaluation of investments in buildings and building
systems.3
These economic worth methods consider only monetary benefits and costs
associated with investment choices. However, even in selection of physical
facilities such as a building, decisions have to be made considering other
aspects, such as location/accessibility, site security, aesthetics, and
uncertainty that cannot be evaluated in monetary terms.
Non-financial characteristics may be either qualitative or quantitative.
Examples of non-financial quantitative characteristics are those which are
readily measured but require judgment to monetize. The number of deaths in a
major earthquake event is a quantifiable item, but there is no agreed upon
method or technique to convert data on deaths into a monetary value.
Qualitative impacts on the other hand are generally impractical or too costly to
measure. Unique community assets, which make a particular community
desirable to live in, cannot be measured quantitatively.
Because non-financial characteristics can be very important, decision-makers
need a method to account for these attributes or characteristics when
choosing among alternatives. Methods that accommodate non-monetary
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
benefits and costs are classified as MADA. These methods apply to problems
where a decision-maker is choosing or ranking a finite number of alternatives
which are measured by two or more relevant attributes. (Norris and
Marshall, 1995, p.4)
Elements of a MADA Problem
Norris and Marshall (1995,p.3-5) identify four characteristics common to all
MADA problems; they are stated below briefly:
Finite set of alternatives
MADA problems involve analysis of a finite set of discrete and predetermined
alternatives.
Tradeoffs among attributes
There is no single attribute that dominates tradeoffs among various attributes
or substitutes as a surrogate for all attributes. Also, there is often an
underlying tradeoff relationship among attributes.
Incommensurable Units
The attributes in a MADA problem generally are not measurable in the same
units. Some attributes may either be impractical, impossible or too costly to
measure. If all relevant attributes can be expressed in terms of financial costs
or benefits, then application of MADA is not required.
165
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Decision Matrix
A decision matrix indicates both the set of alternatives and the set of attributes
to be considered in a given problem and it summarizes the “raw” data
available to the decision-maker at the start of the analysis. Alternatives are
arranged in rows and attributes are arranged in columns.
In addition to a decision matrix, additional information from the decision-maker
is required to arrive at a final ranking. The decision matrix provides no
information about relative importance of the different attributes to the decision
maker, nor about any minimum acceptable or target values for particular
attributes.
Simplifying Assumptions
There are two basic assumptions made in MADA methods in order to simplify
procedures and analytical techniques. First, uncertainties are neglected; i.e.
uncertain values are represented by their expected values rather than
probability distributions. Second, imprecision in the decision matrix data is
neglected; i.e. ratings such as good or bad are converted to scalar numbers
rather than expressed in ranges.
Several MADA methods are available (Norris and Marshall, 1995; Forman,
1998). Different methods require different amounts and types of information
about attributes and alternatives above and beyond the basic data in the
166
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
decision matrix. Methods requiring additional information place heavier
demands upon decision-makers, but they are able to combine, evaluate and
trade-off the decision matrix data in more sophisticated ways than simpler
methods.
MADA methods that are used to solve three types of problems are of particular
interest: Screening alternatives, ranking alternatives or choosing a final ‘best’
alternative. MADA methods are also required to allow tradeoffs between low
and high performance attributes. The methods that allow for such tradeoffs are
known as compensatory methods. (Norris and Marshall,1995, p.9)
Compensatory Methods
Compensatory methods generally require that attributes be measured in
commensurate units, or that the methods have procedures for normalizing
data which are not initially commensurate in order to facilitate attribute tradeoff
analyses. Among these methods those which also allow a decision-maker to
assign weights to different attributes are of interest to this study.
Four compensatory methods are summarized; and two methods, the Non-
traditional Capital Investment Criteria (NCIC) method and the Analytical
Hierarchy Process (AHP), are described further. The four compensatory
methods are; weighted product, TOPSIS, distance from target and additive
weighting.
167
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Weighted Product
In this method the performance scores for attributes are raised to the power of
the attribute importance weight. Then, instead of summing up the resulting sub
scores across attributes to yield the total score for the alternative, the product
of the scores yields the final alternative scores. This method, however, tends
to penalize poor performance on one attribute more heavily than does the
additive weighting method.
Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) 4
The principal consideration in this method is that, the final chosen alternative
should be as close to the ideal solution as possible and as far from negative-
ideal solution as possible. The ideal solution is deemed to be a composite of
the best performance values exhibited in the decision-matrix by any alternative
for each attribute. On the contrary, the negative ideal solution is the composite
of worst performance values. Proximity to each of these extremes in
performance values is measured by: the square root of the sum of the squared
distances along each axis in the “attribute space.” Optimal weighting of each
attribute is allowed. The solutions can be depicted graphically.
Distance from Target
In this method target values for each attribute are chosen; these values need
to be exhibited by any available alternative. Proximity to the target point is
measured by the square root of the sum o f the squares method. The shortest
168
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
distance to the target is the chosen alternative. Weighting of attributes is
allowed. This method also allows graphical depiction.
Additive Weighting
The additive weighting method facilitates the screening, ranking, and choosing
of alternatives by developing a cardinal numerical score for each alternative. In
this method, the score of an alternative is equal to the weighted sum of its
cardinal evaluation ratings, where the weights are the importance weights
associated with each attribute. The Analytical Hierarchy Process (AHP) is a
particular approach to this methodology. AHP is described in detail separately.
MADA methods that require cardinal weighting of attributes can utilize paired
comparisons as a means to establish the attribute weights; however, the only
two methods that allow pair-wise comparisons, converted to attribute weights
in a pre-specified approach, are NCIC & AHP. These methods are further
described in this chapter. One of these two methods, AHP, is described in
detail as that is the preferred method used in this study.
Non-traditional Capital Investment Criteria (NCIC) Method
This method was developed in 1991 by Boucher and MacStravic. The intent of
this method is to tie multi-attribute evaluation of alternatives more closely to
traditional present worth analysis.
169
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
NCIC requires a decision-maker to specify a baseline alternative against which
all other alternatives are judged. For each alternative, the decision-maker
assesses the increments of value added or subtracted by the differences in
performance between the alternative and the baseline.
Specific Requirements
NCIC requires structuring a problem by using a hierarchy of attributes and
sub-attributes. The method acknowledges that some attributes may be
relevant for only some of the alternatives; it allows the decision-maker to
construct a separate hierarchy of attributes for each alternative if needed.
NCIC also requires that each hierarchy contain an attribute called annual
benefits. Thus a net benefit (total benefits minus costs) expressed in monetary
terms is required.
In order to accomplish this, the final scores in this method are presented in
monetary units rather than scoring units which some other MADA methods
entail. In making decisions about acceptable seismic risk, it is neither desirable
nor possible to express final scores in monetary units. Therefore, even though
the NCIC is a multi purpose and powerful MADA method, it is not discussed
further or used in this study.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Analytical Hierarchy Process (AHP)
The Analytical Hierarchy Process (AHP) was developed primarily by Saaty.5 It
is a mathematical process that assists individuals and groups in making
decisions based on multiple criteria. It is a type of additive weighting method,
and it is among the most widely used methods for decision making.6 Additive
weighting methods are recognized for their simple and intuitive logic, multi
purpose functionality, and their incorporation of compensatory tradeoffs
among attributes. The basic logic of additive weighting methods consists of
four principles as described below:
Cardinal Numerical Scores
Cardinal numerical scores are used which characterize the overall desirability
of each alternative. These desirability scores can then be used to rank the
alternatives, to identify the most preferred alternatives or to select a single
most preferred alternative.
Cardinal Attribute Weights
The relative importance of attributes to decision-makers is assumed to be
constant across alternatives and is described using cardinal weights, which
the decision-maker assigns to each of the attributes. The weights are
generally normalized so that they sum to one.
171
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Contributions to Desirability
The performance of each alternative with respect to each attribute must be
expressed numerically, and the numerical attribute scores must be
comparable across attributes. An alternative’s characteristic with respect to
each attribute contributes a calculable amount to the total desirability score of
the alternative. This contribution is calculated as the product of the attribute
score and the attribute importance weight.
Additivity
The separate contributions of each attribute to the total desirability score of an
alternative are considered to be additive. The overall desirability score for an
alternative is defined as the sum of the individual attribute’s contributions.
In general AHP is a method of breaking down complex unstructured situations
into components; arranging the parts into a hierarchic order; assigning
numerical values to subjective judgements on the relative importance of each
variable, and synthesizing judgments to determine which variables have the
highest priority and should be acted upon to influence the outcome of the
situation.
AHP focuses on the achievement of objectives. The key focus is on objectives
rather than alternatives. Tradeoffs among attributes and their relative
importance are key elements of the AHP process. AHP allows decision-
172
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
makers to model a complex problem in a hierarchical structure (Figure 6-1)
showing the relationships between the goal, objectives (criteria), sub
objectives and alternatives.
This method allows application of data, experience, insight and intuition in a
logical and thorough way. AHP also allows both objective and subjective
considerations in the decision process.
AHP is composed of several previously existing but unassociated concepts
and techniques such as hierarchical structuring of complexity, pair wise
comparisons, redundant judgments, an eigenvector method for deriving
weights and consistency considerations.7
Saaty combined these processes and concepts with a result, which far
exceeds the sum of its parts. The methodology has been utilized widely in
decision-making as well as in resource allocation problems.
Problems with Weighting
Rather than dealing with nominal, ordinal, or interval data, AHP uses ratio
scale numbers that provide more flexibility and accuracy.
173
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Hierarchical Structure
Goal
Objectives
Sub-
Objectives
Alternatives
JL
JL
J L
□ C
3 C
31 IE
3E 3C
3E
3 a r —ic
it
3 I ir
3 t
I— II— 1 [ I I II I [ JL
3E
I I ir
1
3 r— ii------ii i
Figure 6-1
Experiments have proven time and time again that the human brain is limited
in both its short-term memory capacity and its discrimination ability to about
seven things. According to James Martin (1973), if a person “has to choose
from a range of 20 alternatives, he will give inaccurate answers because the
range exceeds the bandwidth of his channel for perception. In many cases,
seven alternatives are the approximate limit of his channel capacity.” 8
174
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Not only have psychologists demonstrated that humans have difficulties with
considering more than about seven plus or minus two factors at a time, but
there is a mathematical basis for this phenomenon. Thomas Saaty (1980) has
shown that to maintain reasonable consistency when deriving priorities from
paired comparisons, n, the number of factors being considered must be less
than or equal to nine.
one of the difficulties with a weights and scores methodology stems from the
arbitrary assignment of weights, e.g. if a weight of ‘8’ is assigned to a
particular preference, does it really mean ‘8’ or could it be T or ‘9’. To
overcome this difficulty, qualitative statements are used, e.g. ‘A’ is ‘more’
preferable than ‘B’. It conveys the meaning that ‘A” is preferred to ‘B’ but does
not necessarily inform by how much! Since ‘more’ may mean different things
to different people.
AHP uses pairwise relative comparisons and incorporates redundancy, thus
reducing errors and providing a measure of consistency of judgments. A set of
judgments can be expressed verbally, numerically, or graphically. Although the
words themselves may not be accurate, use of redundancy permits accurate
priorities to be derived from verbal judgments.9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
AHP thus allows conversion of qualitative factors into quantitative factors.
There are three important ways in which AHP extends the weighting process:
pairwise comparisons, the principal eigenvector method and hierarchy.
Pairwise Comparisons
Decision-makers often find it difficult to accurately determine cardinal
importance weights for a set of attributes simultaneously. As the number of
attributes increases, it is necessary to convert the problem into making a
series of pair-wise comparisons for better results. For each pair of attributes,
the decision-maker specifies a judgment about how much more important one
attribute is than other.
Table 6-1 Verbal scale for pairwise comparison of attributes in AHP
Verbal judgment Numerical equivalent
Extremely more important 9
Between very strongly and extremely more
important
8
Very strongly more important 7
Between strongly and very strongly more
important
6
Strongly more important 5
Between moderately and strongly more important 4
Moderately more important 3
Between equally and moderately more important 2
Equally as important 1
176
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Two types of comparison judgments can be used: Numerical and Mediated.
Numerical judgments ask a decision-maker to specify by how much an
attribute is preferable to another (3 times, etc.) However, a precise scoring
may not be feasible when responding to questions. In such cases, a mediated
judgment is used. A verbal statement such as “strongly preferable” is
converted to a numerical equivalent. A nine point scale (see Table 6-1) is
proposed by Saaty to convert verbal judgments into numerical equivalents.
Principal Eigenvector Method
The next step in the additive weighting technique is to convert the paired
comparison data into attribute weights. This is accomplished by a particular
technique called the principal eigenvector method.
A matrix of pair wise comparisons is a square matrix, meaning that is has an
equal number of rows and columns. A matrix of n rows and columns has n
eigenvectors. The principal eigenvector is the eigenvector that has the largest
absolute value. The eigenvector method calculates a vector of cardinal
weights which are normalized to sum to one. This calculation is performed with
thecomputer program 1 0 used in this study.
Hierarchy
Saaty (1980, 1988) defines a hierarchy as “an abstraction of the structure of a
system to study the functional interactions of its components and their impacts
upon the entire system.”1 1 AHP formalizes the hierarchy to keep the number of
177
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
paired comparisons manageable and to facilitate handling problems with
numerous or multi-faceted attributes. Primary attributes can be broken down to
sub-attributes called “leaf attributes. It is recommended in the AHP
methodology to keep the attributes to seven or less. Detailed attributes can be
assigned to the sub-attribute level within a particular attribute. Attributes can
also be grouped to keep the number below seven.1 2
Numerical and Comparable Attribute Scores
The overall desirability score for an alternative is defined as the weighted sum
of that alternative’s “attribute scores.” There are two ways to compute
numerical and comparable attribute scores: normalizing quantitative data and
scoring alternatives using a matrix o f pairwise comparisons.
Normalizing Quantitative Data
When the desirability score for each attribute cannot be simply added because
the units are not commensurate; e.g. price and warranty period, then
normalization of the data in the decision matrix is necessary. This can be
achieved either by division by the sum or division by the maximum value.
Before the normalizing process can begin, it is essential to identify those
attributes for which higher values are preferred. Such attributes are identified
as benefit attributes. Those for which lower values are preferred are identified
as cost attributes, (e.g. price)
178
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
in normalizing benefit attributes the division by sum entails dividing each value
within a column by the sum in that column. Similarly dividing by the maximum
value entails dividing values within a column by the maximum value in that
column.
AHP recommends the use of division by sum method with the caution that
both methods should be tried out; and if the selection of a particular method
influences the final outcome, then the pairwise comparison based method is
more useful.
Scoring Using Pairwise Comparison
When a decision matrix includes qualitative data, or when the relationship
between qualitative performance data and the desirability of different levels of
performance is non-linear, this method of normalizing the data should be used.
Alternatives are compared in terms of strength of preference rather than
importance with respect to the attribute of interest. The principal eigenvector
method can be used to calculate a vector of attribute values. However, the
values along the principal eigenvector must still be normalized either by the
division by sum or division by the maximum value.
Strengths and Limitations of AHP
AHP is a well tested method, which allows consideration of multiple,
conflicting, non-monetary attributes in decision-making. Application of this
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
method is also facilitated by available user friendly software.1 3 Critics have
pointed out certain limitations of the method.1 4 The two most cited limitations
are: First, after deleting a particular alternative, rank reversal among remaining
alternatives could occur, and Second, the attribute weighting could be
arbitrary.
Saaty and other defenders of AHP have responded to this criticism by stating
that these limitations are not a flaw of the method because real world decision
making also exhibits such properties occasionally. While the acceptance of
AHP is not universal, it is widely used and considered to be well suited for
solving a variety of practical MADA problems. For the problem of determining
acceptable seismic risk in a community, AHP is deemed to be particularly well
suited.
Life
Safety
Economic
Impact
Business
Interruptions
Infrastructure
Damage
Community
Preparedness
Community
Assets
Critical Facilities
Operability
Acceptable Seismic Risk
Attributes of Total Acceptable Seismic Risk
Figure 6-2
180
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The hierarchy of acceptable seismic risk attributes is shown in Figure 6-2:
In this hierarchy of total acceptable seismic risk, the attributes are divided as
follows:
Benefit Attributes: Life Safety, Critical Facilities Operability, Community Assets
and Community Preparedness
Cost Attributes: Infrastructure damage, Business Interruptions and Economic
Impact
The application of the AHP method to analyze acceptable seismic risk based
on expert opinions is detailed in the next chapter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Notes
1 Herbert A. Simon, The New Science of Management Decision, New York, N.Y., Harper & Row,
I960, pp. 40-43
2 Irving L. Janis, Crucial Decision — Leadership in Policy-making and Crisis Management, The Free
Press, N.Y., 1989, pp. 287
3 American Society for Testing and Materials, ASTM Standards on Building Economics, Third Edition,
Philadelphia, PA. 1994
4 TOPSIS was developed by Hwang and Yoon (1981) in Lecture Notes in Economics and Mathematical
Systems 186, Springer-Berlin and NewYork, 1981
5 AHP was developed by Thomas L. Saaty, The Analytic Hierarchy Process: Planning, Priority setting
resource Allocation, Pittsburgh,PA. University of Pittsburgh, 1980
6 Hwang and Yoon describe this method further in “ Multiple Attribute Decision Making”, 1981, p.91
7 Decision by Objectives, E.H. Forman, Expert Choice Software, McLean, VA.1998, p29
8 Decision by Objectives, E.H. Forman, Expat Choice Software, Inc. McLean, VA. 1998, p.32
9 Decision by Objectives, E.H. Forman, Expert Choice Software, Inc. McLean, VA. 1998, p.33
1 0 The computer program used is developed by Expert Choice Inc., Pittsburgh, Pa. 1999 version is used
in this study.
1 1 Multicriteria Decision Making: The Analytic Hierarchy Process, Thomas L. Saaty, Pittsburgh, PA.
University of Pittsburgh, 1988, p.5
1 2 The limit of seven is not a hard constraint of AHP theory, but it has been proposed by available
software packages which facilitate the use of AHP. Keeping the number of attributes per set small helps
limit the total number of pairwise comparisons as the total number of attributes becomes large.
1 3 Expert Choice, Inc. Pittsburgh, PA. 15213,1999
1 4 A partial summary of the criticisms of AHP is given by James Dyer, “Remarks on the Analytical
Hierarchy Process,” Management Science, 36(3), 1990, p.249-258
182
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 7
EXPERT OPINION SURVEY RESULTS, ANALYSIS OF SURVEY AND
DISCUSSION OF RESULTS
GENERAL
This chapter discusses the results of the expert opinion survey related to
acceptable seismic risk. Two types of questionnaires were sent to experts
identified in their field. The sample questionnaires are enclosed in Appendix B.
The first questionnaire asked them to provide opinions on the importance of
life safety in physical facilities and whether mandatory or voluntary retrofits of
physical properties should be required. This same questionnaire also solicited
opinions on the rank order of importance of types of physical facilities, e.g.
buildings, utilities, and transportation infrastructure. Critical facilities, which
need to remain operational during and after an earthquake event were
identified and experts were asked to rank order them. Similarly community
assets, which are deemed to be important, were listed, and the experts again
were asked to rank order them. A total of 44 experts were polled using the first
questionnaire; thirty of them returned the questionnaires, which represents a
return rate of nearly 70%. The second questionnaire stressed the importance
of pair-wise comparisons; experts were asked to make comparisons in terms
of numerical grades of various attributes.
183
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The second questionnaire was sent to 26 experts who had responded to the
first questionnaire. Some respondents (4 out of 30) for the first questionnaire
had difficulties in responding to some questions. Their responses were not
considered in the survey results as they were incomplete. The purpose of the
second questionnaire was to rank order specific attributes of acceptable
seismic risk according to a rigorous verbal scale (see Appendix B). These
experts were asked to numerically compare one attribute with respect to
another. In this questionnaire, no direct qualitative responses were allowed.
Definitions of various attributes were provided for clarity in making decisions
(see Appendix B).
The ambiguity perceived by some respondents toward some questions in the
first questionnaire was removed in the second questionnaire. Out of 26
respondents 20 returned the second questionnaire representing a return rate
of nearly 80%.
Considering that the second questionnaire was sent only two weeks after the
first questionnaire, the rate of return for the response to second questionnaire
is good. It is important to note that at least three experts in each of the
categories of stakeholders provided their opinions. It is therefore concluded
that a fair number of responses from experts in each field was received. All
numerical scores from questionnaires 1 and 2 were averaged to represent the
collective opinions of all experts related to a particular category or attribute.
184
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The averages of these scores are shown in Table 7-1 and Figures 7-1 through
7-4.
FINDINGS FROM QUESTIONNAIRE No. 1
The general findings from the first questionnaire are discussed below.
1. Except two respondents who ranked life safety considerations as slightly
less important than the highest rank of 9, all others ranked life safety
considerations as the most important issue. This finding is not surprising as
most casualties in a major seismic event occur due to the lack of life safety
provisions.
In response to whether “mandatory” seismic retrofit of physical facilities
should be required, 60% of the respondents either “strongly agreed” or
“agreed” to this proposition. Another 20% of the respondents either
“disagreed” or “strongly disagreed” and the remaining 20% “slightly agreed”
(see Fig. 7-1).
It appears, that mandatory retrofit as an option to upgrade existing facilities
is not supported by all the experts. It would have been expected that
owners of physical property would resist the mandatory retrofit due to the
financial burden; however, nearly 50% of those who did not support
185
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
mandatory retrofit requirements are structural engineering experts
concerned with building code regulations related to life safety.
Mandatory Retrofit of Physical Facilities
100%
80%
Strongly Agree Slightly Agree Strongly Disagree
F ig ire 7-1
3. All except one respondent either “agreed” or “strongly agreed” to create an
incentive structure for voluntary upgrade of physical facilities. This option is
preferred by an overwhelming majority of the experts (Not shown in the
Figure 7-1).
4. Seven critical facilities deemed important to remain operational were
identified. These were: urgent care hospitals, fire stations, police stations,
public schools, transportation management centers, non-urgent care
hospitals, and senior citizen housing. Experts were asked to rank order
these, rank 1 is the highest and rank 7 is the lowest. The results of ranking
are shown graphically in Figure 7-2. Although urgent care hospitals
186
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
emerged as the highest ranked facilities to remain operational, the
difference in numerical scores between urgent-care hospitals and fire
stations is only 10%. Fire stations ranked the second most important
facilities to remain operational. The difference in numerical score between
transportation management centers and public schools is only 2%. For all
practical purposes these two facilities could be considered equally ranked.
Senior citizen housing is considered least important. It probably can be
surmised that the ethical/moral obligation to have senior citizen housing
facilities operational during an earthquake does not weigh heavily as
compared with other facilities on the list.
Ranking of Critical Facilities to Remain Operational
(0
■ i °-
Z IB
k e
O 0 £
• > .
a .a
Hre Station PaSce Station Sc ho ate Hon-Urgent Sr. Gfeen Urgent Care Transportation
Care Hospitals Housing Hospitak Management
Centers
Rank 1.0 = Highest
Rank 7.0 = Lowest
Figure 7-2
187
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5. Experts were also asked to rank buildings, transportation infrastructure,
and utilities in terms of the relative importance to limit physical damage to
these facilities. While on an overall numerical score basis, buildings ranked
the highest and transportation infrastructure the lowest, the difference in
numerical scores between each category is less than 10%, suggesting that
no clear-cut preference exists in allocating resources to upgrade these
physical facilities. Perhaps it is of some significance to note that although
on an overall average basis building facilities ranked as the most important,
only 50% of the respondents ranked them as the most important, whereas
nearly 35% ranked them as the least important. This difference suggests a
wide divergence of opinions and explains why a clear-cut preference
among these categories of facilities does not exist. Ranking scores are
shown in Figure 7-3.
Ranking of Physical Facilities (Minimize Damage)
3.0 ------------------------------------------------------------------------------------
Buildings Transportation Utilities
Infrastructure
_ Figure 7-3
I
Rank 1.0 = Highest
Rank 3.0 = Lowest
188
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6. Community assets such as museums, religious facilities, historic structures
and those assets which are unique to a community such as a baseball
park, were rank ordered by experts as follows: Historic structures were
ranked highest, museums second, religious facilities third and unique
community assets last (see Figure 7-4). It is somewhat surprising that
unique community assets are ranked lowest in the category of overall
community assets. The difference in numerical scores between each
category is at least 20% thus indicating more defined preferences.
Although, numerically the historic structures ranked at the top, nearly 40%
of the respondents ranked museums at the top of community assets
category suggesting that divergence of opinion is still considerable.
Ranking of Community Assets
Unique to
Community
Historic Religious
Facilities
Museums
| Rank 1.0 = Highest
! Rank 4.0 = Lowest I
189
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FINDINGS FROM QUESTIONNAIRE No. 2
Based on the survey results from the second questionnaire the following
observations can be made. Scores for each attribute are shown in pair-wise
comparisons in Table 7-1. The software computer program prints out the
complete matrix, the scores are rounded to full digits without decimals in this
matrix (see Table 7-2).
Table 7-1: Importance of Attributes by Respondents
Economic
Impact
Community
Asset
Preservation
Physical
Facilities
Damage
Business
Interrupt.
Critical
Facilities
Operability
Comm.
Prepared
Physical
Facilities
Damage
Business
Interrupt.
Economic
Impact
Critical
Facilities
Operab
Comm.
Asset
Preserv.
Comm.
Prepared
7. Life safety scores ranged from 4.81 to 8.13 out of a maximum of 9.0 when
compared with other attributes, indicating that the life safety attribute
dominates all others and is considered by the experts to be at least
“strongly more important” at the lower end of score scale to “between very
190
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
strongly and extremely more important” at the higher end of the score
scale. Scores for each attribute are shown in Table 7-1.
8. Minimizing physical facilities damage ranged from a score of 1.88 to 5.69
compared to other attributes on a verbal judgment scale, it ranked only
“moderately more important” as compared to operability o f critical facilities
but ranked “between strongly and very strongly more important” as
compared with business interruptions. The comparison scores for each
attribute are shown in Table 7-1.
Table 7-2
A ccep tab le seism ic ris k - m inim , d am ag e cost- m axim ize life s a fe ty _______
Node: 0
Compare the relative IM PO RTANCE with respect to: GOAL
1 =EQUAL 3 =M O PERATE 5=STRO NG 7 -V E R T STRONG 9=EXTREM E
1 UFESAFE 9 a 7
£ )
5 4 3 2 1 2 3 4 5 6 7 8 9 PHYS.FAC
2 UFESAFE 9 i
&
7 6 5 4 3 2 1 2 3 4 5 6 7 8 9 BUSINESS
3 U FESAFE 9 8
£ >
6 5 4 3 2 1 2 3 4 5 6 7 8 9 ECONOMIC
4 UFESAFE 9 8 7 6
C
4 3 2 1 2 3 4 5 6 7 8 9 CRITICAL
5 UFESAFE 9 8
P
6 5 4 3 2 1 2 3 4 5 6 7 8 9 ASSETS
6 UFESAFE 9 8 7
C
5 4 3 2 1 2 3 4 5 6 7 8 9 PREPARED
7 PHYS.FAC 9 8 7
< 2
6 4 3 2 1 2 3 4 5 6 7 8 9 BUSINESS
8 PHYS.FAC 9 8 7 6 5
c
3 2 1 2 3 4 5 6 7 8 9 ECONOMIC
8 PHYS.FAC 9 8 7 6 5 4 3
©
1 2 3 4 5 6 7 8 9 CRITICAL
10 PHYS.FAC 9 8 7 6
5C *
3 2 1 2 3 4 5 6 7 8 9 ASSETS
1 1 PHYS.FAC 9 8 7 6 5
£>
3 2 1 2 3 4 5 6 7 8 9 PREPARED
12 BUSINESS 9 8 7 6 5 4 3 2 ■ 2 3 4 5 6 7 8 9 ECONOMIC
13 BUSINESS 9 8 7 6 5 4 3 2
©
2 3 4 5 6 7 8 9 CRITICAL
14 BUSINESS 9 8 7 6 5 4
©
2 1 2 3 4 5 6 7 8 9 ASSETS
15 BUSINESS 9 8 7 6 5 4 3
* L *
2 3 4 5 6 7 8 9 PREPARED
IB ECONOMIC 9 8 7 6 5 4 3 2
©
2 3 4 5 6 7 8 9 CRITICAL
1 7 ECONOMIC 9 8 7 6 5
O
3 2 1 2 3 4 5 6 7 8 9 ASSETS
10 ECONOMIC 9 8 7 6 5 4
3 C *
1 2 3 4 5 B 7 8 9 PREPARED
19 CRITICAL 9 8
P
5 4 3 2 1 2 3 4 5 6 7 8 9 ASSETS
20 CRITICAL 9 8 7 6 5
©
3 2 1 2 3 4 5 6 7 8 9 PREPARED
21 ASSETS 9 8 7 6 5 4 3
2 2 3 4 5 6 7 8 9 PREPARED
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9. The Business interruptions attribute scored between 0.87 and 2.94. A rank
below 1.00 denotes that business interruption is less important than
operability o f critical facilities and is only “moderately more important” than
preservation of community assets at the higher end of the score scale. Of
course this attribute is highly dependent upon the types and number of
businesses in a community; e.g. in the high tech manufacturing business
community of the South Bay in the San Francisco region such a finding
would be considered an anomaly.
10.The Economic impact attribute ranged from 0.86 to 3.93. The score of 0.86
when compared with operability of critical facilities suggests it to be less
important. At the high end of the score scale, economic impact is
considered “between moderately and strongly more important” than
preservation of community assets. Since economic impact which includes
short and long term economic effects is a rank below operability o f critical
facilities, this suggests that experts stress the immediate impact on the
community more strongly than long term economic effects.
11. Operability o f critical facilities scored between 3.87 and 6.75 suggesting
that it is “between moderately and strongly more important” when
compared with community preparedness] however, it is “very strongly more
important” than preservation of community assets. One could infer that
more emphasis is placed on community preparedness as compared to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
preservation o f community assets. However, when the community assets
preservation attribute is scored against community preparedness, it scores
only 1.44, suggesting that experts consider these two attributes to be
almost equally important.
Such inconsistency in expert opinions as expressed in pair-wise comparison is
analyzed in the computer program which calculates an inconsistency ratio
(IR).
AHP Model o f Acceptable Seismic Risk
A model of acceptable seismic risk is structured in a hierarchy as shown in
Fig. 7-5.
Since the computer software has certain limitations in assigning titles, the
abbreviations are fully defined in Table 7-3. On this page a tree diagram of a
hierarchy of attributes is shown for easy reference.
Seven attributes as defined in Chapter 5, comprise the hierarchy of attributes
at level 1. Three attributes, i.e. physical facilities damage, operability o f critical
facilities, and preservation of community assets have sub-attributes. The level
of sub-attributes is called level 2 in the structured hierarchical model. The goal
of the overall model is to minimize cost attributes and maximize benefit
attributes. Both benefit attributes and cost attributes are listed in Chapter 5;
they are repeated here for ready reference. Benefit attributes are; life safety;
193
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
operability of critical facilities; preservation of community assets-, and,
community preparedness. The cost attributes are: physical facilities damage,
business interruptions, and economic impact
Acceptable Seismic Risk- Minimize Damage Cost- Maximize Life Safety
Figure 7-5
The next step in the program is to determine the weight of each attribute. This
is accomplished by a “bottom up” process. The weights of sub-attributes at
level 2 are calculated first. For the overall weight for the physical facilities
damage attribute the scores of three sub-attributes (leaf attributes); buildings,
transportation infrastructure, and utilities are used.
194
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 7-3
Acceptable seismic risk- minim, damage cost- maximize life safety
GOAL-
-UFESAFE •
PHYSFAC
4
BUILDING-----
U m iH E -------
TRANS POR —
BUSINESS -
ECONOMIC ■
CRITICAL -
HOSPITAL ■
FIRESTA -
POUCES -
NONURGE■
TRANSPO -
SCHOOLS -
L SENIOR C -
-HISTORIC
I-MUSEUMS----
ASSETS ------ 1 - RELIGIO-------
•-UNIQUE-------
PREPARED —
Abbreviation Definition
GOAL
ASSETS Maximize preservation of community assets
BUILDING Retrofit buildings to provide life safety
BUSINESS Minimize Business Interruptions
CRITICAL Critical facilities to remain operational
ECONOMIC Minimize Economic Impact on the Community
FIRESTA Fire stations to remain operational
HISTORIC Historic structures in the community
HOSPITAL Urgent care hospitals to remain functional
LIFESAFE Life safety need to conform to building codes
MUSEUMS Museum housing non replaceable items
NONURGE Non urgent care facilities to remain accessible
PHYS.FAC Reduce physical facilities damage
POLICES Police stations to remain operational
PREPARED Maximize community preparedness
RELIGIO Religious facilities such as churches, temples, mosques
SCHOOLS School facilities to remain usable by the community
SENIOR C Senior citizen housing to remain accessible and operable
TRANSPO Transportation management centers to remain functional
TRANSPOR Minimize damage to roads,bridges^ailroads;ports and harbors
UNIQUE Unique community features
UnLITIE Minimize damaae to electric, water eas aid telecom services
Attribute Definitions
195
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The average scores for these sub-attributes shown in Figure 7-3 are
normalized to buildings sub-attribute. The normalized scores and the
descriptions of these sub-attributes are shown in Table 7-4. The normalized
scores are then used in the matrix to calculate the weight of each sub
attribute. The computer program accomplishes this calculation through
eigenvector analysis. The resulting weights are shown in Figure 7-6. The
buildings sub-attribute has a weight of 0.374 for the overall category of
physical facilities damage attribute. Compared to the overall goal of minimizing
damage costs and maximizing life safety, the physical facilities damage
attribute has a weight of 0.191 as shown in Figure 7-6. Similar calculations for
operability o f critical facilities show an overall weight of 0.108. Each sub
attribute weight in that category is shown in Table 7-4 and also in Figure 7-7.
Acceptable Seismic Risk- Minimize Damage Cost- Maximize Life Safety
Ranking Physical Facilities Attributes
Fig u re 7-6
196
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 7-4
Acceptable seismic risk- minim, damage cost- maximize Itfe safety
Node: 20000
Compare the relative IMPORTANCE with respect to: PHYS.FAC UTIUTIE TRANSPOR
BUILDING 1.1 1.3
U TIU TIE 1.1
R pggtem antts tim es more than column elem ent unless enclosed in f i
Abbreviation Definition
Goal Acceptable seismic risk- minim, damage cost-maximize life safety
PHYS.FAC Reduce Physical Facilities damage
BUILDING Retrofit buildingsto provide life safety
UTIUTIE Minimize damage to electric, water, gas and telecom services
TRANSPOR Minimize damage to roads.bridges.railroads.portsand harbors
Ranking Physical facilities Attributes
Calculations of weights for preservation o f community assets and its sub
attributes are shown in Table 7-6 and Figure 7-8.
Acceptable Seismic Risk-Minimize Damage Cost- Maximize Life Safety
Critical Facilities -S ub Attributes
Figure 7-7
197
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 7-5
Acceptable seismic risk- minim, damage cost- maximize life safety
Node: 60000
Compare the relative IMPORTANCE with respect to: ASSETS < GOAL
MUSEUMS RELIGIO UNIQUE
HISTORIC 1.2 1.5 1.8
MUSEUMS 1.3 1.5
RELIGIO 1.2
Rom d u T w n tis tjm tsm o re tta n column «tan«nt isitesxtndoscd in 0
A b breviation D efinition
Goal Acceptable seismic risk- minim, damage cost- maximize life safety
ASSETS Maximize preservation of community assets
HISTORIC Historic structures in the community
MUSEUMS Museum housing non replacable items
RELIGIO religious facilities such as churches, temples, mosques
UNIQUE Unique community features
Ranking Community Assets
Acceptable Seismic Risk-Minimize Damage Cost- Maximize Life Safety
Community Assets - Sub Attributes
Figure 7-8
198
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The next step in the process is to calculate the weight of each of the seven
major attributes. The numerical scores from respondents for pairwise
comparisons and the descriptions of objectives under each attribute are shown
in Table 7-7. It should be noted that two scores are denoted in parentheses
because they had average values of 0.87 and 0.86 respectively.
Table 7-6
A c c e p ta b le s e is m ic ris k - m inim , d a m a g e c o s t- m axim ize fife s a fe ty ________
Node: 0
Com pare the relative IM P O R TA N C E with respect to: GOAL
PHYS.FAC BUSINESS e s a iN io rtii: CRiTiC aL ASSETS ' ' PREPARED
UFESAFE 6.1 8.1 7.0 4.8 7.0 5.9
PHYS.FAC 5.7 3.8 1.9 4.6 4.1
BUSirESS 1.1 1.0 2.9 1.7
ECONOMIC 1.0 3.S 2.5
CRITICAL 6.7 3.9
ASSETS 1.4
atamantis_timasmorathan column alamant unianancftosad in 0
Abbreviation
Definition
Goal Acceptable seismic risk- minim, damage cost- maximize life safety
U FESA FE Life safety need to conform to building codes
PHYS.FAC Fteduce Physical Facilities damage
BUSINESS Minimize Business Interruptions
ECO NO M IC Minim ize Economic Im pact on the Community
CRITICAL Critical facilities to remain operational
ASSETS Maxim ize preservation of community assets
PREPARED Maxim ize community preparedness
Ranking Attributes of Acceptable Seismic Risk
Since the methodology does not allow for scores less than 1, the row attribute
needs to be reversed with the column attribute. However, since the scores are
close to 1.0 such a step is not warranted. The final weights for the attributes as
calculated by eigen-vector analysis are shown in Figures 7-9 and 7-11.
199
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acceptable Seismic Risk- Minimize Damage Cost- Maximize Life Safety
Final Weights of Various Attributes
Figure 7-9
Acceptable seismic risk- minim, damage cost- maximize life safety
Derived Priorities with respect to Goal
UFESAFE .485
PHYS.FAC .191
BUSINESS .067
ECONOMIC .077
CRITICAL .108
ASSETS .034
PREPARED .038
Inconsistency Ratio =0.08
Fig ure 7-10
Inconsistency Ratio (IR)
The computer program also calculates an inconsistency ratio. This ratio
reflects the percentage of time decision makers are inconsistent in making
200
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
judgments on a particular set of elements. The inconsistency ratio in this study
came out to be 0.08 (see Figure 7-10). According to Saaty and the
methodology, an inconsistency ratio below 0.10 is considered acceptable.1 A
perfectly consistent result of pairwise comparisons would suggest an
inconsistency ratio of 0.00. These calculations complete the process for
calculating weights (derived priorities) of the various attributes that go into
making the final decision.
Synthesis of Weights
The results are further analyzed based upon the purpose for which the results
are to be used. In this study, the goal is to prioritize various attributes to
reduce overall damage and maximize life safety. Given this purpose, the
program synthesizes each leaf attribute node with respect to the overall global
goal. The synthesis, shown in Table 7-7 is called a “distributive mode”
synthesis. It generates the weight of each leaf attribute. The inconsistency
ratio in this calculation is 0.07, which is below the acceptable value of 0.10.
These results demonstrate that the pair-wise comparisons are fairly
consistent. Graphically the scores for the weights of major attributes are
shown in Figure 7-10 and the scores of each attribute including sub attributes
are shown in Figure 7-11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acceptable seismic risk-minim, damage cost- maximize life safety
Synthesis of Leaf Nodes with respect to GOAL
Dislrtiiitive Mode
OVERALL INCONSISTENCY INDEX = 0.07
UFESAFE
ECONOMIC
BUILDING
BUSIfCSS
UTILITIE
TRANSPOR
PREPARED
HOSPITAL
FIRE STA
POLICES
NONURGE
HISTORIC
TRANSPO
SCHOOLS
SENIOR C
MUSEUMS
RELIGIO
UNIQUE
.485
077
.071
.067
.063
.056
.038
.025 ■
.023 ■
.016 ■
.014 ■
.011 ■
.010 ■
.010 ■
.008 ■
.008 ■
.007 I
.006 I
Figure 7-11
202
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 7-7
Acceptable seismic risk- minim, damage cost- maximize life safety
Synthesis of Leaf Nodes with respect to GOAL
Distributee Mode
OVERALL INCONSISTENCY INDEX = 0.07
LEVEL 1 LEVEL 2 LEVEL3 LEVEL4 LEVEL 5
LIFESAFE-485
PHYS.FAC=.191
BUILDING=.071
UT1UTIE-0B3
TRAN SPOR=.056
CRmCAL=108
HOSPiTAL=.025
FIRE STA=.023
POLICE S=.016
NONURGE =014
TRANSPO = 010
SCHOOLS = 010
SENIOR C=.008
ECONOMIC=.077
BUSINESS=.067
PREFARED=.Q38
ASSETS =.034
HISTORIC=.011
MUSEUMS =009
RELIGIO =.007
UNIQUE =006
NOTE: Levels 3, 4, and 5 do not exist in this study
ACCEPTABLE SEISMIC RISK TO COMMUNITY (ASRc)
For the community in this study, the benefit attributes; life safety, operability o f
critical facilities, community assets, and community preparedness have a
combined weight of 0.665. The cost attributes: physical facilities damage,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
business interruptions, and economic impact have a combined weight of
0.335.
The purpose, as defined in the overall goal, is to maximize benefit attributes
and minimize cost attributes. Since the community stakeholders through
expert opinions have expressed their priorities by assigning weights to each
attribute, the community is in a position to make decisions about allocating
resources appropriately.
Out of available resources, if ‘R’ resources are to be allocated to reduce
seismic risk to a level acceptable to the community, then nearly 48% of the
resources will be expensed on life safety considerations, nearly 20% on
reducing physical facilities damage potential and another 10% on keeping
critical facilities operational. The remaining 20% of the resources are divided
into keeping business interruptions to a minimum (nearly 7%), keeping
economic impact minimal (nearly 8%), preserving community assets (nearly
3%), and the remaining 4% in community preparedness efforts.
A general equation of acceptable seismic risk to a community can now be
written as:
Equation 7-1:
ASRc = [a(LS) + b(FD) + c(bi) + d(Ei) + e(CF) + f(CA) + g(cp)] +8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Where ASRc = A measure of acceptable seismic risk to a community.
Constants ‘a’ through ‘g’ are weights assigned to each attribute by community
stakeholders and ‘8’ is a variable which may be important to a specific
community but may not be one of the general attributes of acceptable seismic
risk. If all attributes are defined by the community, then ‘5‘ would be zero.
In this study, the constants (weights) for various attributes are found to be as
follows:
a = 0.485 LS - Life Safety
b = 0.191 FD - Physical Facilities Damage
c = 0.067 Bl - Business Interruptions
d = 0.077 El - Economic Impact
e = 0.108 CF - Operability of Critical Facilities
f= 0.034 CA - Preservation of Community Assets
g = 0.038 CP - Community Preparedness
and
‘5‘ = 0
Based on the derived priorities, when the community makes resource
allocation decisions, it still may not be able to meet the needs of the
community to mitigate seismic risk. In that case, a redistribution of allocations
may be necessary, and the feedback from stakeholders needs to be
205
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
incorporated into the decision matrix to derive new priorities. A feedback loop
shown in Figure 1-1, in the Introduction Chapter needs to be activated.
Another way to look at priorities is to divide the overall goal into two, as
follows:
Subject to ‘R’ (resources) < 1.0
Maximize
Total benefit attributes = [a (LS) + e (CF) + f (CA) + g (CP)] — Equation (7-2)
and Minimize
Total cost attributes = [b ( fd) + c (Bi) + d (El)] — Equation (7-3)
Explanation of Methodology Application
The application of the proposed decision methodology is described in
summary from below:
1. In a community, to identify attributes to be considered, consider the
composition of the community related to its economic base, infrastructure
dependence, community assets, its preparedness to face a major seismic
event and life safety vulnerability.
2. Select stakeholders that represent the community. The selection of
stakeholders is based upon the criteria detailed in Chapter 5. Care needs
206
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
to be exercised in selecting the stakeholders so they are a fair
representation of the community.
3. Determine various attributes of acceptable seismic risk. Decide on sub
attributes of various attributes. If necessary, decide sub-attributes of sub
attributes.
4. Prepare a questionnaire eliciting opinions of stakeholders on various
attributes and sub-attributes. The opinions need to be expressed by either
verbally or numerically comparing one attribute or sub-attribute at a time
with another at the same level of the hierarchy; i.e. a sub-attribute can be
compared only with another sub-attribute because they are at the same
level of hierarchy.
5. Structure a model hierarchy defining each attribute and sub-attribute. If the
number of attributes or sub-attributes is more than seven, grouping may be
necessary to limit their number to seven. This is the limitation of the
currently available computer software.
6. Based upon the stakeholder survey responses, calculate average
numerical scores for each element i.e. an attribute or a sub-attribute.
Verbal judgements are converted into numerical scores based upon a
scale proposed by Saaty (see Appendix B, Questionnaire No.2).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7. Input this data (i.e. scores) in the computer program to determine the
weight of each element based upon a matrix of pairwise comparison. The
computer program performs an eigen-vector analysis and generates
weights of elements. The weights of attributes at the lowest level of the
hierarchy need to be determined first. Moving up through the hierarchy
levels from the bottom up, weights of attributes at each level are
determined. These weights of attributes at level 1 are to be used in making
resource allocation decisions.
8. The weight generated for each attribute represents the collective weight of
stakeholders' opinions for that attribute.
9. Resource allocation decisions to mitigate seismic risk can be made
according to these weights (derived priorities).
10. Because resources in a community are limited, the required resources
needed for mitigation of specific aspects of seismic risk may not be met.
Also, once a decision to allocate resources is made based upon derived
priorities, community stakeholders may revisit the allocation and determine
revised priorities. In such cases, new weights (priorities) need to be
generated.
208
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Suppose a community has identified financial requirements based upon
technical experts’ opinions that it should upgrade its physical facilities and
provide life safety to building occupants at a cost of $20m and $10Cm
respectively- However, the maximum resources in the community are only
$50m for the overall seismic mitigation program. Then, according to the
derived priorities, only $9.55m can be spent on physical facilities needs and
another $24.25m can be used to provide life safety. The remaining sum of
$16.2m is allocated to other priorities. These resource allocations do not meet
the needs of the community as identified. Either reallocation of resources
needs to be done, or the acceptable seismic risk by the community needs to
be redefined based upon the derived priorities. Reallocation of resources will
define a different level of acceptable seismic risk.
The Analytical Hierarchy Process (AHP) is used as a methodology to account
for expert opinions in comparing various attributes of acceptable seismic risk.
It is posited that through this research a theoretical framework has been
developed to reach decisions on acceptable seismic risk to a community by
considering opinions of stakeholders.
The next chapter discusses the potential application of this methodology at
different levels of government, implications for further research and theory and
then the implications of the methodology for public policy.
209
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 8
SUMMARY, FINDINGS, CONCLUSIONS AND IMPLICATIONS
Introduction
The purpose of this dissertation is to develop a framework by which a
community can determine acceptable seismic risk to that community. One of
the more basic problems is that currently seismic risk is defined by technical
experts without necessary and valuable input from stakeholders who are
impacted by the hazard. To technical experts, seismic risk is tantamount only
to damages to physical facilities including infrastructure. That the extent of
potential damage to the physical facilities can be fairly accurately predicted is
indeed a remarkable achievement by the scientific and engineering
community. However, damage to physical facilities is only one component of
total seismic risk, albeit an important one.
If a local community is to build safer new buildings and structures and to
retrofit existing structures, what part of available community resources should
be expended on this endeavor? Any community with limited resources has to
make tough decisions on allocation of resources among competing claims.
How much seismic risk is acceptable to a community, and what measures can
210
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
be adopted to increase the risk bearing capacity of a community? To study
these issues, the following research questions are posed:
How should acceptable seismic risk to a community be defined? Who are the
stakeholders in defining acceptable seismic risk, and how should they be
selected? What are the various attributes of acceptable seismic risk? How
can the quantitative and qualitative attributes of seismic risk be combined to
obtain a single measure of acceptable seismic risk? Can the proposed
methodology be generalized to determine acceptable seismic risk at different
levels of society?
It is important to address these questions in order to develop a general
methodology for determining acceptable seismic risk to a community so that
informed public policy decisions related to mitigating seismic risk can be
made. Although the focus of the study is a community, it is believed that the
methodology developed is general in nature and can be utilized for making
informed public policy decisions at various levels of governments.
Discussion of Proposed Methodology
Current loss estimation methodologies, particularly HAZUS are powerful in
assessing damage to physical facilities and consequent direct economic
effects. Societal considerations and indirect long term economic effects are
not taken into account, most probably due to difficulties in quantifying them.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Loss estimates will vary with each community depending upon the actions
taken by community stakeholders.
It has been amply evident in past significant earthquake events that the
amount of structural and non-structural damage alone exceeds the capacities
of affected communities to absorb them. In the United States under the
disaster mitigation policy of the Federal government losses to individuals and
some businesses are reimbursed. However, as losses due to each major
natural disaster keep on mounting, the Federal government alone cannot
provide the needed disaster assistance to communities. The Federal
government has, as a result, developed a policy under which it expects to
share seismic risk with stakeholders. In 1998, it initiated “Project Impact:
Building a Disaster Resistant Community.” This program is a partnership
between the Federal government, local businesses, and communities. In 120
identified ‘Project Impact’ communities across the United States, more than
1000 businesses and non-profit organizations are cooperating to build safer
communities. This study has developed a framework to account for
preferences of community stakeholders.
As this study emphasizes that acceptable seismic risk to community must be
determined by stakeholders in the community, it is necessary to analyze
methods that allow the input of stakeholders. The steps utilized in proposed
methodology used in this study are;
212
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1. Establish selection criteria for stakeholders
2. Identification of stakeholders
3. Identification of attributes of acceptable seismic risk
4. Eliciting preferences of attributes by experts as proxies for stakeholders
5. Selection of Analytical method
6. Considerations of preferences in the analysis
7. Determining final weights o f attributes
The method used to analyze ranking of preferences should be able convert
qualitative data into numerical scores and analyze the data in a pair-wise
comparison mode. The Analytical Hierarchy Process (AHP), a specific type of
multiattribute decision analysis method, is particularly well suited for this
purpose and is used.
Organization of Chapter
This chapter summarizes the discussions on total seismic risk, acceptable
seismic risk, and various issues identified above. Conclusions and limitations
of the findings are then discussed. Finally, implications of the proposed
methodology for theory and further research are posed along with the
implications for citizens participating in making informed decisions.
In this study, it has been argued that seismic risk in general and acceptable
seismic risk in particular is a social and cultural construct. The stakeholders in
213
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the community, taking into account technical experts’ opinions, must make its
determination. It has also been argued that although the probability of
occurrence of a major seismic hazard event and its magnitude cannot be
controlled by a community, its impact on the community can be modified by
the actions of stakeholders in the community.
The impact of a major seismic event encompasses various aspects of
community life; it impacts life safety of building occupants, interrupts business
operations, affects lifeline systems of a community, disrupts its economic
health, destroys community assets, puts greater demands on rescue and relief
agencies and strains the economic recovery efforts of a community. The
extent of these impacts varies from community to community as the
composition, economic base and resource capacity vary for each community.
Total and Acceptable Seismic Risk
The premise in this study has been that acceptable seismic risk cannot and
need not be universalized but ought to be determined on a community specific
basis. This study has attempted to conceptualize acceptable seismic risk as a
community specific issue to be decided by the stakeholders in the community.
A community- based decision in allocating resources can be useful only if it is
an informed decision based upon available information and if stakeholders’
opinions are appropriately considered in decision processes. Total seismic risk
214
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
is comprised of three components: technological, economic, and societal. It is
argued in this study that unless all three components of total seismic risk are
taken into account, the evaluation and assessment of total seismic risk are not
complete. The technological component must be determined by technical
experts; the economic consequences by businesses, owners of physical
properties and public officials and the societal component by public officials
representing citizens of the community at large. Identified stakeholders in a
community must decide the amount of risk in each of the areas that is
acceptable to the overall community. A definition of acceptable seismic risk is
proposed:
Acceptable seismic risk is that amount o f risk which can be estimated
and voluntarily accepted by an individual, a family, a group, a
community or society, taking into account technological, economic,
social and political considerations.
A natural hazard such as a major seismic event is a low probability, high
consequences hazard. Although its magnitude and probability of occurrence
are determined by factors outside the control of human beings, the effects of
such an event can be controlled by humanly devised actions. Because seismic
risk can be modified as a result of community actions, acceptable seismic nsk
is both an independent and a dependent variable.
215
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Selection Criterion for Stakeholders
For community based decisions on acceptable seismic risk, it is extremely
critical that stakeholders be selected considering their stake in the outcome of
the decision and their expertise and interest in various components of seismic
risk. It is also essential that stakeholders be representative of the community
as a whole.
An informal discussion with seismic loss estimation experts resulted in
formulating the criteria for selection of stakeholders. The proposed selection
criteria for stakeholders demands that each group must score ‘high' in at least
one area of influence and at least ‘medium’ in another area of influence. There
are six criteria that are listed on next page. The influence of stakeholders is
assessed qualitatively as ‘low ‘medium’, and ‘high’.
Identification of Stakeholders
Based upon the criteria defined above, seven groups of stakeholders are
identified within six areas of expertise or influence. The public interest is
assumed to be protected by elected officials as they represent the electorate.
The stakeholders are; controllers o f financial resources, political decision
makers, regulatory agencies, physical property owners, lifeline support system
providers, engineering and scientific experts, and rescue and relief agencies.
216
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The areas of expertise are:
1. Serviceability of community infrastructure and utilities.
2. Emergency preparedness and recovery effort.
3. Physical property ownership.
4. Technical expertise in assessing potential damage to physical
facilities.
5. Economic base of community.
6. Representation of societal interests
Experts serve as proxies for stakeholders. Experts are selected from the
scientific and engineering community, public officials, emergency
preparedness coordinators, owner’s representatives, and lifeline support
providers from a wide range of alternatives available through professional
associations, the scientific community, and public official rosters. Chosen
experts are individuals well recognized for their knowledge, experience, and
ability to influence seismic risk outcome.
Attributes of Acceptable Seismic Risk
To seek expert opinions, it is necessary to first identify attributes of acceptable
seismic risk and put them in a questionnaire format to elicit expert opinions.
The attributes are: Life Safety Considerations, Business interruptions,
Infrastructure Damage, Operability o f Critical Facilities, Economic Impact,
Community Assets, and Community Preparedness.
217
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ranking of Preferences by Stakeholders
Two types of questionnaires were sent to experts identified in their field. The
first questionnaire asked them to provide opinions on the importance of life
safety in physical facilities and whether mandatory or voluntary retrofits of
physical properties should be required. This same questionnaire also solicited
rank orderings of the importance of types of physical facilities, e.g. buildings,
utilities, and transportation infrastructure. Critical facilities which need to
remain operational during and after an earthquake event were identified, and
experts were asked to rank order them. Similarly community assets, which are
deemed to be important were listed, and the experts again were asked to rank
order them. The second questionnaire stressed the importance of pair-wise
comparison; experts were asked to compare attributes in terms of numerical
grades.
Experts were asked to rank the importance of one attribute with respect to
another in a pairwise comparison mode. The opinions of experts as proxies for
stakeholders in the community were mathematically analyzed by the Analytic
Hierarchy Process (AHP). The AHP methodology takes into account
quantifiable and non-quantifiable scores. This methodology has been widely
used in decision making where the overall goal is to make the best decision in
a multi-criteria setting.
218
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The AHP methodology was chosen for this study for the following reasons:
1. Decisions on acceptable seismic risk need input from many divergent
stakeholders in a community.
2. Since appropriate allocation of resources to mitigate seismic hazard risk
is the goal, preferences of stakeholders must be ranked.
3. Acceptable seismic risk has several attributes, all of which need to be
considered in the overall decision process.
4. Not all preferences can be quantified; the pair-wise comparison of
attributes can be used based on qualitative statements which are then
converted into numerical scores.
5. AHP allows for consistent evaluation of various elements that go into
decision making.
6. AHP gives a theoretical grounding for consistent decision making.
7. Weights can be applied to different priorities.
8. By using pair-wise comparisons, community decision making can be
accomplished to arrive at the best level of acceptable seismic risk.
9. The methodology allows one to pinpoint areas where the community
can divert resources to use them most cost-effectively in enhancing the
resource capacity of a community.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Findings
Findings are described in detail in chapter seven. However, a summary of
important findings is given below:
1. More than 95% of respondents ranked life safety as the most important
attribute.
2. The majority of respondents preferred “mandatory” retrofit as compared
to "voluntary” retrofit of physical facilities.
3. Owners of physical property were not among those who did not want
‘mandatory’ retrofits.
4. An overwhelming majority would prefer creation of an incentive structure
to upgrade physical facilities.
5. Although seven critical facilities to remain operational were ranked in
order of importance, the difference between each rank was no more
than 10%, thus indicating no clear preference ranking. A similar situation
exists in ranking the preference between limiting physical damage to
buildings, transportation infrastructure and utilities.
6. Minimizing damage to physical facilities ranked considerably above the
need to limit business interruptions.
7. Immediate economic impact on the community is considered more
important than long term economic effects.
8. Operability of critical facilities ranked considerably higher than
preservation of community assets. Community preparedness ranked
nearly equal to preservation of community assets.
220
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9. An inconsistency ratio less than 0.10 indicates that inconsistency in
expressing expert opinions is within acceptable limits of AHP
methodology.
10. Out of the seven attributes of acceptable seismic risk, four are benefit
attributes with a combined weight of 0.665. There are three cost
attributes with a combined weight 0.335.
11. Since each attribute has a weight assigned through analysis of data, a
community is able to decide on the allocation of resources to mitigate
seismic risk considering attribute weights.
12.it is possible to write a general equation for acceptable seismic risk to a
community, with weights attached to each of the attributes.
Limitations of Methodology
In general, the following limitations of the methodology can be identified:
1. Currently available computer software allows a maximum of seven
attributes. More than seven attributes need to be grouped.
2. Decisions are made by the majority and not necessarily by a consensus
of the stakeholders.
3. If certain critical needs are not met, stakeholders may have to allocate
resources differently than weights assigned to attributes from analysis.
221
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In such a case acceptable seismic risk to a community is redefined by
stakeholders as the decisions do not follow the analysis.
4. Decisions on attribute preferences outside the immediate control of a
community, such as infrastructure, cannot be made without concurrence
of the owners of those services.
5. Although the methodology is general, attributes and their weights have
to be determined on a community specific basis.
Conclusions
1. Total seismic risk is comprised of three components; technological,
economic, and societal. A determination of total seismic risk needs to
incorporate expert opinions from all three disciplines, not just
technological.
2. A definition of acceptable seismic risk to a community is proposed. The
concept of acceptable seismic risk need not be universalized.
Acceptable seismic risk varies with each community depending upon its
specific characteristics.
3. Since acceptable seismic risk is posited to be a social and cultural
construct, it is important to have input from stakeholders for its
determination.
222
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4. It is possible to take into account the ranking of preferences expressed
by the stakeholders in the overall analysis of pair-wise comparisons of
attributes.
5. A generalized methodology is developed to account for preferences of
stakeholders in determining acceptable seismic risk specific to a
community. The weights of various attributes that comprise the total
score are more important than the total score itself. These weights allow
a community to focus their resource allocation to specific aspects of
acceptable seismic risk.
6. It is believed that the methodology developed can be utilized at different
levels of government to make informed public policy decisions
considering stakeholders input.
Implications o f Proposed Methodology for Theory
The proposed methodology has attempted to develop a process by which
community stakeholders and therefore preferences of community at large can
be considered in determining acceptable seismic risk to a community. The
concepts proposed in this study can be utilized in developing a theory of
acceptable natural hazard risk through community based participation.
• The concept proposed in this study does not simply consider public opinion
but stimulates an informed deliberative decision process.
223
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
• The decision process takes into account expert opinions in technological,
economic and societal disciplines.
• A basic concept developed in the study is that acceptable risk need not be
determined by wholly scientific or wholly democratic procedures. The
scientific findings through the opinions of scientific experts are valuable in
determining the extent of risk and its consequences, no matter how
narrowly defined these opinions may be. In developing a theory, these
scientific findings can be incorporated as part of a social decision process.
• The particular nature of natural hazards i.e. events with low probability of
occurrence with high consequences requires theoretical considerations
different than those for man made risks.
• A generalized theory can identify parameters of acceptable risk; it can
leave the determinants to specific cases.
Implications for Further Research
• Defining various attributes of acceptable risk and their dependencies is an
important consideration.
• Since the risk to a community in the case of natural hazards varies with
actions taken by its citizens, defining criteria and selecting stakeholders
requires careful consideration. Criteria for selection of stakeholders is an
area o f further significant research.
• Seeking expert opinions (as proxies for stakeholder opinions) and ranking
their preferences is another area for further research.
224
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
• Determining an acceptable level o f community risk requires tradeoffs
dependent upon the resource capacity o f a community. Tradeoffs can be
based upon results o f an analysis o f preferences; however, this will vary
with each community. The extent o f optimal tradeoffs without sacrificing the
heavily weighted attributes can be studied further.
Finally, an area of further research is, distinguishing between public risk and
private risk and the implications of private risk on overall community risk.
Implications for Policy for Citizens
Neither the local, state or federal government has sufficient revenue sources
to fund all the programs demanded by citizens. Competing programs vie for a
share of overall revenue. Programs such as seismic hazard mitigation often
take a back seat as a major seismic event has a low probability of occurrence,
and the distribution of probabilities vary geographically.
To allocate resources appropriately, an objective decision making process and
methodology are a necessity; otherwise, pork-barrel politics take over and
resources are either not available or misallocated for programs such as
seismic hazard risk reduction. AHP provides a mathematically based
methodology which takes into account subjective preferences of stakeholders
in a community and assists communities in making decisions based on
multiple criteria.
225
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
By using the framework developed in this study, public policy decisions on
allocation of resources can be made more effectively. The framework allows
flexibility to add or subtract attributes of a goal, add or subtract stakeholders;
and it has the ability to generate several levels o f attributes that may be
important to a community. Direct participation by citizens in major decision
making which impacts their welfare is plausible by using this framework.
Citizens are more likely to accept the outcome o f a policy decision if they
participated in making it.
Although, politicians will continue to make decisions in a political rationality
manner, they are ultimately responsible to the electorate. The framework
developed in this study allows citizens and other stakeholders to express their
opinions and assist decision-makers in taking those opinions into account.
A p p lica tio n fo r D iffere nt Levels o f S ociety
The methodology in this study has been used to derive priorities of
stakeholders in a community setting because such decisions are to be made
by communities. However, the framework fo r the decision methodology is
general in nature and can be used at different levels of society for decision
making related to acceptable seismic risk.
Consider acceptable seismic risk decisions at the state level. A state may
have different probabilities of occurrence of a major seismic event in its
226
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
different geographic parts. Decision methodologies for these different parts
can be grouped separately as the opinions of stakeholders in different parts
are likely to be different.
In each part of the state several cities and counties may be involved. All the
elected public officials from one part should be considered as representing the
interests of the public in that p a rt Life safety considerations may involve
bedroom communities, industrial communities and other community
compositions. Such divergence in life safety considerations can be captured
in sub-attributes and sub-sub-attributes o f life safety in the overall hierarchical
structure of the methodology.
For example, in a bedroom community, a majority of the physical property is
concentrated in single family homes. The construction material of these
homes is generally wood. Single family homes built with wood generally do
not pose a life safety risk. However, it is important in a bedroom community to
have minimal damage to infrastructure so residents can go to work. The
infrastructure damage attribute is likely to be more important than the life
safety attribute for these types of communities as single family homes pose
little life safety risk potential.
Such sentiments can be captured in expressions of stakeholders’ opinions.
Attributes o f acceptable seismic risk in a specific part of the state can be
227
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
broadly defined. For each of the divergent communities, sub-attributes and
sub-sub-attributes can be defined in detail.
Selection of stakeholders to represent a large geographical area is more
tedious and complex as compared to smaller geographical areas. However,
based upon the criteria of selection developed in the study, stakeholders can
be selected. Attributes of acceptable seismic risk are listed in Table 8-1 .The
list in the table serves as a “master list” from which attributes can be selected.
However, the list can be expanded to accommodate specific needs of a
community. Remaining procedures related to questionnaires, calculation of
weights or priorities of attributes and sub-attributes are the same as discussed
in community decision making in the previous chapter.
Depending upon the derived priorities, appropriate decisions can be made at
the state level on the extent and distribution of resources to be allocated to
mitigate seismic risk. There is a need to initiate an incentive structure for
communities to enhance their resource capacities to absorb more seismic risk
locally. It is hoped that community-based decisions on acceptable seismic risk
will lead to judicious use of limited community resources and better public
policy decisions at different levels o f government.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 8-1: Attributes of Seismic Risk
Level I Level II Level III
Life Safety
Considerations
(a) Life safety in single family homes
(b) Life safety in rental housing properties
(c) Life safety in public occupancy
buildings
(d) Life safety in industrial plants
(e) Life safety in public schools
(f) Life safety in hospitals
Physical
Facilities
(a) Buildings 1. Residential
2. Non-residential
(b) Transportation Network 1. Roads
2. Bridges
3. Railroads
4. Airports
5. Harbors
(c) Utilities 1. Gas
2. W ater
3. Electricity
4. Wastewater
5. Telecommunications
Business
Interruptions
(a) Small businesses
(b) International businesses
(c) Regional businesses
(d) Hi-tech businesses
(e) Industrial manufacturing
(f) Service industries
Operability of
Critical
Facilities
(a) R re stations
(b) Police stations
(c) Schools
(d) Non-urgent care hospitals
(e) Senior citizen housing
(f) Urgent care hospitals
(g) Transportation management centers
Economic
Impact
(a) Short term 1.Restoring functionality to
community life
2.Temporary service contracts
(b) Long term 1. Loss of businesses
2. Loss of community services
3. Change in perception
Preservation of
Community
Assets
(a) Unique to community
(b) Historic structures
(c) Religious facilities
(d) Museums
Community
Preparedness
(a) Immediate response ability
(b) Long term community assistance
capability
(c) Capability to reduce impact of hazard
229
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
BIBLIOGRAPHY
Applebaum, R. P. “The future is made, not predicted: Technocratic planners
versus public interests.” Society (May-June 1977): 49-53.
Applied Technology Council. Earthquake Damage Evaluation Data for
California (ATC-13). Redwood City, CA: Applied Technology Council, 1985.
— Procedures fo r Post earthquake Safety Evaluation o f Buildings (ATC-20).
Redwood City, CA: Applied Technology Council, 1989.
— Rapid Visual Screenings o f Buildings fo r Potential Seismic Hazards (ATC-
21). Redwood City, CA: Applied Technology Council,1988.
Axelrod, Robert. The evolution o f cooperation. New York: Basic Books, 1984.
Barnard, Chester I. The functions o f the executive. Cambridge: Harvard
University Press, 1938, 1968.
Bell, David. “Regret in Decision Making Under Uncertainty.” Operations
Research 30 (1982): 1367-1381.
Boucher, T.O. and MacStravic, E. L. “Multi attribute Evaluation within a
Present W orth Framework and its Relation to the Analytic Hierarchy
Process.” Engineering Economics 37 (1991).
Bowman, Edward, H. and Howard Kunreuther. “ Post Bhopal Behavior at a
Chemical Company,” Journal o f Management Studies, Vol.25 (1988): 387-
402.
Brookshire, David et al. “Direct and indirect Losses from Earthquake
Damage.” Earthquake Spectra (November 1997): 683-702.
Brown, M. “Communicating Information about W orkplace Hazards: Effects on
W orker Attitudes towards Risks.” in The Social and Cultural Construction
o f Risk, (Covello, V., and B. Johnson, Eds. 1987): 251-274.
Caldwell, L.K. and K.S. Shrader-Frechette. Policy fo r Land and Ethics.
Lanham, MD: Rowman and Littlefield Publishers, 1993, p.113-161.
California Department of Conservation (Division of Mines and Geology).
“Seismic Hazard Information Needs o f the Insurance Industry, Local
Government and Property Owners in California: An Analysis.” Special
Publication No. 108, Sacramento, CA, 1990.
230
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
California Department of Conservation (Division of Mines and Geology).
“The Northridge, Ca. Earthquake o f 17 Jan.1994.” Special Publication
SP116, (Woods,M.C. and W.R. Seiple, Eds.), Sacramento, CA, 1995.
California Legislature- Joint Hearing of the Senate Insurance and Judiciary
Committees. Hearing on Earthquake Insurance Issues. Government
Document 833S, Sacramento, CA, October 1995.
Camerer, Colin, F. and Howard Kunreuther. “ Decision Processes for Low
Probability Events: Policy Implications.” Journal o f Policy Analysis and
Management (Vol.8, N o .4 ,1989): 562-592.
Card, W .l. and G.H. Mooney. “W hat is the Monetary Value o f Human Life?”
British Medical Journal (Part 2 , 1977):1627-1629.
City of Kobe, Japan. The great Hanshin- Awaji Earthquake Statistics and
Restoration Progress Report (Feb A, 2000):1-4.
Clarke, Lee. Acceptable Risk? Making Decisions in a Toxic Environment
Berkeley:University of California Press,1989, p.60
Cohrssen, John and Vincent Covello. “Risk Analysis: A Guide to Principles
and Methods for Analyzing Health and Environmental Risk.” U.S. Council
on Environmental Quality, Executive Office o f the President, Washington,
D.C., (1989): 7.
— National Research Council: Committee on Risk Perception and
Communication - Improving Risk Communication, National Academy
Press, Washington, DC (1989): 32.
Covello, V. “Uses of Social and Behavioral Research on Risk.” Environmental
International (June 1984): 17-26.
Covello, Flamm, Rodricks and Tardiff, (Eds.) The Analysis o f the Actual versus
Perceived Risks, New York: Plenum Press,1983, p.235-249.
Covello, V. and B. Johnson. “The Social and Cultural Construction of Risk:
Issues, Methods, and Case Studies.” in The Social and Cultural
Construction o f Risk: Essays on Risk Selection and Perception, (Johnson,
B. and V. Covello, Eds.) Dordeecht, Holland : D. Reidel Publishing
Co.1987), vii-xiii.
Crouch, Edmund A.C. and Richard Wilson. Risk/Benefit Analysis. Cambridge:
Ballinger Publishing Co.1982.
231
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Cyert, R.M and March, J.G. “Organizational factors in the Theory of Oligopoly.”
Quarterly Journal o f Economics, (February 1956).
Douglas, Mary and A. W ildavsky. Risk and Culture: An Essay on the
Selection o f Technical and Environmental Dangers. Berkeley: University of
California Press,1982, p.5-13.
Earthquake Engineering Research Institute (EERI). Earthquake Basics B rief 3,
April 1997.
Edwards, W. and J.R. Newman. Multiattribute Evaluation: Quantitative
Applications in the Social Sciences. Sage Publications,1983.
Eguchi, R.T., James Goltz, Hope Seligson et al. “Real-Time Loss estimation
as an Emergency Response Decision Support System: The Early Post-
Earthquake Damage Assessment Tool (EPDAT).” Earthquake Spectra,
vol.13,No.4, (Nov. 1997): 815-832.
Eguchi, R.T. and Hope A. Seligson. “Lifeline Damage and Resulting Impacts.”
in Socio-Economic impacts, Monograph 5, National Earthquake
Conference, Memphis,TN, (1993): 69-99.
EQE Engineering. Portfolio Analysis and the EQE Earthquake Risk reduction
Program. (EQE Engineering Bulletin 19), San Francisco, CA.
Expert Choice, Inc. Expert Choice User's Guide. Pittsburgh, Pennsylvania:
Expert Choice Inc. 1999.
Federal Emergency Management Agency (FEMA). Seismic Considerations for
Communities a t Risk. Washington, DC, 1983.
— Seismic Considerations fo r Communities at Risk, Earthquake Hazards
Reduction Series 13. Washington, DC, October, 1990.
— Reducing the Risks o f Non-Structural Earthquake Damage: A Practical
Guide. W ashington, DC, September 1994.
— Create Results - O riented Incentives to Reduce the Costs o f a Disaste.
FEMA 03. W ashington, DC, 1997.
— State Capability Assessment fo r Readiness: Report to the US Senate
Committee on Appropriations. Washington, DC, December 1997.
— Report on Costs and Benefits o f Natural Hazard Mitigation, Executive
Summary. W ashington, DC, 1998.
232
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fischhoff, Baruch, R. Stephen Watson, and Chris Hope. “Defining Risk.”
Policy Sciences \/ol.17, (1984): 124.
Fischoff, B., Paul Slovic and Sarah Lichtenstein. “Fault Trees: Sensitivity of
estimated failure probabilities to problem representation.” Journal o f
Experimental Psychology: Human Perception and Performance Vol.4,
(1978): 330-344.
— “Weighting the risks.” Environment, 21 (4), (1979): 17-20, 32-38.
— “Images o f Disaster. Perception and acceptance of risks from Nuclear
Power.” in Energy Risk Management, (Goodman, G. and W. D. Rowe,
Eds.), New York Academy Press, 1979.
— “Perceived Risks: Psychological Factors and Social Implications.”
Proceedings o f the Royal Society (London) A376, (1981): 17-34
— Acceptable Risk. Cambridge : Cambridge University Press,1981.
Fischoff, Baruch. “ Acceptable Risk: The Case of Nuclear Power.” Journal o f
Policy Analysis and Management Vo 1 . 2, No.4, (1983): 559-575.
— “Challenges in Risk Assessment and Risk Management.” in The ANNALS
o f the American Academy o f Political and Social Science, (Kunreuther.H.
and P.SIovic, Eds.) Vol.545,Thousand Oaks, CA: Sage Periodical
Press,1996, p.75-83.
Fischoff, Baruch., P.SIovic, S. Lichtenstein, S. Read, and B. Coombs. “ How
Safe is safe Enough? A psychometric Study towards Technological Risks
and Benefits.” Policy Sciences \/o \.8 , (1978): 127-152.
Fisher, Ann, L.G. Chestnut and D.M. Violette. “ The Value o f Reducing Risks
of Death: A Note on New Evidence." Journal o f Policy Analysis and
Management, Vol.8, No. 1(1989): 88-100.
Fitchen, Janet M. “The Importance of Community Context in Effective Risk
Management.” In Risk Assessment and Management, (L. Lave, Ed.), New
York: Plenum Press,1987, p.677-685.
Fitchen, Jennifer Heath and June Fessenden-Raden. “ Risk Perception in
Community Context: A Case Study.” in The Social and Cultural
construction o f Risk (.Johnson, B. and V. Covello, Eds.), Dordeecht,
Holland: D. Reidel Publishing Co.,1987, p. 31-54.
Forman, E.H. Decisions by Objectives. Expert Choice Software Inc., McLean,
VA, 1998.
233
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Friedman, Donald G. Earthquake Losses under Workers Compensation and
General Liability; estimates fo ra “worst-case” event in G reater Los
Angeles. All Industry Research Advisory Council, Oak Brook, IL, October
1988, p.72.
Gutmanis.l. and John A. Jaksch. “High Consequence Analysis, Evaluation,
and Application of Select Criteria.” in Low Probability High-Cosequence
Risk Analysis, (W aller and V. Covello, Eds.), New York: Plenum Press,
1984, p.393-423.
Hadden, Susan G. Risk Analysis, Institutions and Public Policy. Port
Washington, NY: Associated Faculty Press,1984, p.1-17.
Hanley, J. “The silence o f scientists.” Chemical and Engineering News, 58
No.12 (1980): 5.
Hardin, Russel. Collective Action. Baltimore, MD: John Hopkins University
Press, 1982, p.22-30.
Harker,P.T.(Ed.) Socio-Economic Planning Issues 20/6. Special Issue on the
Analytical Hierarchy Process, 1986.
Harrington, Leigh. The Valuation o f Life Shortening Aspects o f Risks. EPA-
230-07-85-007 United States Environmental Protection Agency,
Washington, DC, September 1984.
Harvey, Charles M. “ Decision Analysis Models for Risks in the Distant
Future.” in The Risk Assessment and Management, (L. Lave, Ed.), New
York: Plenum Press, 1987, p. 509-516.
Henshel, R.L. “Effects of Disciplinary Prestige on Predictive Accuracy :
Distortions from feedback loops.” Futures, Vol.7 (1975): 92-106.
Hershey, John C. and Paul J. Shoemaker. “ Risk Taking and Problem Context
in the Domain o f Losses: An Expected Utility Analysis.” Journal o f Risk and
Insurance 47(1985): 1213-1231.
Hertz, D. B. and H. Thomas. Risk Analysis and its Applications. John Wiley
and Sons, 1984, p. 11-18.
Holder, R.D. “Some Comments on the AHP.” Journal o f the Operational
Research Society 41/1 1 (1990): 1073-1076.
Hwang, Ching-Lai and Kwangsun Yoon. Multiple Attribute Decision- Making.
Lecture Notes in Economics and Mathematical Systems. Berlin: Springer-
Verlag, 1981.
234
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Inhaber, H. “Risk with energy from conventional and non-conventional
sources.” Science 203 (1979): 718-723.
International Decade for Natural Disaster Reduction. Newsletter Issue No. 12.
Geneva, Switzerland, 1998.
International Decade for Natural Disaster Reduction. Partnership fo ra Safer
World in the 21st Century. Geneva, Switzerland, April 1998.
Jam's, I. Victims o f Groupthink. Boston: Houghton Mifflin, 1972, p.207-217.
Johnson, Eric J. and David A. Schkade. “Bias in Utility Assessments: Further
Evidence and Explanations.” Management Science Vol.35 (1989): 406-
424.
Kaprow, Miriam. “Manufacturing Danger: Fear and Pollution in Industrial
Society.” American Anthropologist (1985): 357-364.
Keeney, Ralph. “The role of Values in Risk Management.” in The Social and
Cultural Construction o f Risk, (Covello, V., and B. Johnson, Eds.),
Dordeecht, Holland: Reidel Publishing Co., 1987, p.126-234.
Kelly, Joe. Organizational B ehavior Its Data, First Principles, and
Applications. Homewood,IL: Richard D. Inwin Press,1980, p.476-485.
Kelman,Steve. “Cost-Benefit Analysis: An Ethical Critique.” Regulation Vol.5
No.1 (1981): 33-40.
-- “Adversary and Cooperationist Institutions for Conflict Resolution in Public
Policy Making." Journal o f Policy Analysis and Management Vol.11,
No.2, (1992): 169-206.
Kelso, William A. Benefit-Cost Analysis and Program Evaluation in Decision-
Making in the Public Sector. (Uoyd G. Nigro, Ed.), New York: Marcel
Dekker lnc.1984, p.28-30.
Kielcott, K. and Joanne M. Nigg. Mobility and Perceptions o f a Hazardous
Environment and Behavior, 14(1982): 131-154.
King, Stephanie e ta l. “Methodologies for Evaluating the Socio - Economic
Consequences for Large Earthquakes.” Earthquake Spectra (Nov. 1997):
565-583.
Kingdon, John. Agendas, Alternatives and Public Policies. New York, NY:
Harper Collins, 1995.
235
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Kolb, Jeffrey A. and Joel D. Scheraga, “ Discounting the Benefits and Costs of
Environmental Regulations.” Journal o f Policy Analysis and Management
Vol.9, No.3(1990): 381-390.
Kopp, R. J. “ W hy Existence Value Should be used in Cost-Benefit Analysis.”
Journal o f Policy Analysis and Management Vol.11, No.1 ,(1992): 123-130.
Krimsky, Sheldon and Alonzo Plough. Environmental Hazards:
Communicating Risk as a Serial Process. Dover, MA: Auburn House
Publication Co.,1988, p. 17.
Lee, Kai N. “Assessing Danger.” Book Review. Science. V o l.193, (July
9,1977): 139.
Levine, M. “Scientific Method and the Adversary model: Some preliminary
thoughts.” American Psychologist V ol.29 (1974): 661-716.
Lowrance, W illiam W . “The Nature o f Risk.” in Societal Risk Assessment:
How Safe is Safe Enough? (Richard C.Schwing and W alter A. Albers,Jr.
Eds.) New York: Plenum Press,1980, p.5-14.
March, James.G. “Learning to be Risk Averse.” Psychological Review Vol.103,
No.2 (1996): 309-319.
Markovic, M. Social Determinism and Freedom, in (H.E. K elfer and M.K.
Munitz, Eds), Mind, Science and History, Albany: State University of New
York Press, 1970.
Martin, James. Design o f Man-Computer Dialogues. Englewood Cliffs, NJ:
Prentice Hall, 1973.
May, Peter, and W alter Williams. D isaster Policy Implementation: Managing
Programs under Shared Governance. New York: Plenum Press, p.93-108.
McCann, John. H ow to Prepare fo r an Earthquake. Insurance Information
Institute, New York, 1990, p. 1-2.
McEachern, A.W . “Two Simple Versions o f Multiattribute Utility Analysis.” in
Decisions with Multiple Objectives: Preferences and Value Tradeoffs.
(Keeney, R.L. and H.Raiffa, Eds.), W iley and Son, 1976, p.65-90.
Mishan, E. J. “ Recent Contributions to the Literature of Life Valuation: A
Critical Assessment.” in The Value o f Life and Safety (M. W. Jones-Lee,
Ed.), New York: North Holland Publishers, 1982.
236
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Mitchell,R.C. and R.T. Carson. Using Surveys to Value Public Goods: The
Continent Valuation Method. Washington, DC: Resources for the Future,
1989.
Moore, P.G. The Business o f Risk. Cambridge, U.K: Cambridge University
Press, 1980.
Moore, Michael J. and W. Kip Viscusi. “ Doubling the Estimated Value o f Life:
Results using new Occupational Fatality Data.” Journal o f Policy Analysis
and Management Vol.7, No.3 (Spring, 1988): 476-400.
Morgan, M.Granger and Max Henrion. Uncertainty: A guide to Dealing with
Uncertainty in Quantitative Risk and Policy Analysis. Cambridge:
Cambridge University Press, 1990, p.163-181.
Mujumdar, Vilas. “Seismic Performance of School Buildings in 1995 Kobe,
Japan, Earthquake.” Division o f the State Architect, Sacramento, CA,
January 1996, p.1-10.
- “Evaluation o f Seismic Risk through Total Acceptable Cost Model.” Paper
presented at the 12th W orld Conference on Earthquake Engineering,
January 30 to February 4, 2000, Auckland, New Zeeland.
National center for Earthquake Engineering Research. Bibliography on Cost
Benefit Analysis and M itigation: 1971-June 1996. far Federal Emergency
Management Agency, 1996.
— Bibliography on Mitigation: Cost Benefit/Economic issues 1987-N ov.1996.
for Federal Emergency management Agency, 1996.
Noms, Gregory A. and Harold E. Marshall. Multiattribute Decision Analysis
M ethod for Evaluating Buildings and Building Systems. National Institute of
Standards and Technology. NISTIR 5663, Gaithersburg, MD, September,
1995.
North, Douglass, C. Institutions, Institutional change and Economic
performance. Cambridge: Cambridge University Press,1996.
Otway,H. “ Public Wisdom, Expert Fallability: Toward a contextual Theory of
Risk.” in Social Theories o f R/sfr (Krimsky.S. and D. Golding, Eds.),
W estport, CT: Praeger Publishers,1992.
Otway, Harry and Kerry Thomas. “Reflections on Risk Perception and Policy.”
Risk Analysis 2, (1982): 69-82.
237
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Pacific Earthquake Engineering Research Center (PEER). Defining Links
Between Planning, Policy Analysis, Economics and Earthquake
Engineering, Report 98-04. (Comerio, Mary and Peter Gordon, Eds.),
September 1998.
Page, T. “A generic view o f toxic chemicals and sim ilar risks.” Ecology Law
Quarterly, 7 (1978): 207-243.
- “A framework for unreasonable risk in the Toxic Substances Control Act ."in
R. Nicholson (Ed.), Carcinogenic Risk Assessment, New York: New York
Academy o f Sciences, 1981.
Palm, Risa and Michael Hodgson. After a California Earthquake: Attitudes and
Behavior Change. Chicago: The University of Chicago Press, 1992.
Petak, Wiliam J. “A Survey of Disaster Mitigation Research in the United
States.” The US-former Soviet Union Seminar on Social Science Research
on M itigathn fo r and Recovery from Disasters and Large Scale Hazards,
Moscow, Russia, (Apr.21-24,1993): 1-8.
— “ Guidelines for Developing a Seismic Safety Element for the General Plan.”
PAS Report No. M-12. American Society o f Planning Officials, Chicago,
1973.
Petak, William J and Arthur A. Atkisson. Natural Hazard Risk Assessment and
Public Policy: Anticipating the Unexpected. New York: Springer-Verlag,
1982.
Price, Don K. “Money and lnfluence:The Links of Science to Public Policy.”
Daedulus 103 (summer 1974): 97-114.
Raiffa, Howard. “Concluding Remarks.” in Societal Risk Assessment: How
Safe is Safe Enough? (Schwing and Albers, Eds.) New York: Plenum
Press, 1980, p. 339-341.
Rawls, J. A Theory o f Justice. Cambridge: Harvard University Press, 1971,
p.75-83.
Ravetz, J.R. “Public Perceptions of Acceptable Risks.” Science and Public
Policy 6, (Oct. 1979): 497-512.
Rayner, Steve. “Risk and Relativism in Science for Policy.” in The Social and
Cultural Construction o f Risk: Essays on Risk Selection and Perception,
(Johnson, B. and Vincent T. Covelio, Eds.), Dordeecht, Holland : D. Reidel
Publishing Co., 1987, p.14-23.
238
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
~ The Classification and Dynamics o f Sectarian Forms o f Organization:
Grid/Group Perspectives on the F ar Left in Britain. Ph.D. thesis,
University o f London, 1979.
Renn, Ortwin. “Concepts o f Risk: A Classification.” in Social Theories o f Risk,
(Krimsky, S. and D. Golding, Eds.), W estport, CT: Praeger Publishers,
1992.
Romero, Philip and Justin Adams. “Economic Impact o f the Northridge
Earthquake.” in The Northridge Earthquake, CA. Earthquake o f 17th
January 1994, Sp116 (Woods and Seiple, Eds.), Department o f
Conservation, Division o f Mines and Geology, Sacramento, CA, 1995,
p.263-264.
Rosenthal, Don,H. and Robert H. Nelson. “ Why Existence Value Should not
be Used in Cost-Benefit Analysis." Journal o f Policy Analysis and
Managements/ol.11, No.1 (1992): 116-122.
Rowe, William D. An Anatom y o f Risk. John Wiley and Sons, 1977, p. 193-210.
Rucklehaus, W .D. “Science,Risk and the Public Policy.” Science, Vol. 221,
(1983): 1026-1028.
— “Risk,Science and Democracy.” Issues in Science and Technology, Vol.1,
No.3 (spring 1985): 19-38.
Saaty, Thomas L. Fundamentals o f Decision Making and Priority Theory with
the Analytical Hierarchy Process. Pittsburgh, PA: RWS Publications, 1994,
p.337.
— Multicriteria Decision Making: The Analytic Hierarchy Process. Pittsburgh,
PA: University of Pittsburgh, 1988.
— “Risk - Its Priority and Probability:The Analytical Hierarchy Process.” Risk
Analysis 7/2, (1987): 159-172.
— The Analytic Hierarchy Process: Planning, Priority Setting, Resource
Allocation. New York: McGraw-Hill,1980, p.95.
Sacramento Bee. “Quake Jolts Makers o f Computer Chips,” Sept. 30,1999,
p.C3.
Scawthorn,Charles. Fire Following Earthquakes: Estimates o f the
Conflagaration to Insured Property in Greater Los Angeles and San
Francisco. A ll Industry Research Advisory Council, Oakbrooke, IL.1987,
p.83.
239
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Schofield, Norman. “Anarchy, Altruism and Cooperation: A review.” Social
Choice and W elfare 2, (1985): 207-219.
Shim, J.P. “Bibliographic Research on the AHP.” Socio-Economic Planning
Sciences 23, (1989): 3161-3167.
Shogren, Jason S. A Prim er on Environmental Risk Analysis, S taff Report 90-
SR46.Center for Agricultural and Rural Development Ames, IA: Iowa State
University, 1990, p .13.
Shrader-Frechette, K. S. Risk and Rationality. Berkeley: University of
California Press, 1991, p.206-209.
— Burying Uncertainty: Risk and the Case against Geological Disposal o f
N uclear Waste. Berkeley: University of California Press, 1993, p.31-38.
Simon, Herbert A. Administrative Behavior. New York: The free Press, 1977.
— “The Proverbs of Administration.” Public Administration Review (W inter
1946): 53-67.
Slovic, P., B. Fischoff and S. Lichtenstein. “Facts and Fears: Understanding
Perceived Risk.” in Societal Risk Assessment: How Safe is Safe Enough?
(Schwing, R. and W . Albers, Eds.), New York: Plenum Press,1980, p.181-
216.
Smith,Kerry,V., W iliam H. Desvousses, F.Reed Johnson and Ann Fisher.
“Can Public Information Programs affect Risk Perceptions.” Journal o f
Policy and Management, Vol.9, No.1 (1990): 41-59.
Stallen, P.J. “Risk o f Science or Science of Risk?” in Society, Technology and
Risk Assessment, (Conrad, J. Ed.), London Academic Press, 1980.
Starr, C. and C. W hipple. “Risk of Risk Decisions.” Science, 208, (1980):
1114-1119.
— “General Philosophy o f Risk-Benefit analysis.” Extract from Energy and the
Environment: A risk benefit approach. (Ashley.H., R.L. Rudman and C.
Whipple, Eds.), London: Pergamon Press, 1976, p. 629.
Starr, Chauncey. “Introductory Remarks.” in Societal Risk Assessment: How
Safe is Safe Enough? (Schwing, R. and W . Albers, Eds), New York:
Plenum Press, 1980, p.2-4.
240
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
— “Risk Management, Assessment and Acceptability.” Risk Analysis, 5
(1976): 97-102.
State Farm Insurance Company Bulletin. California Law and You: California
Earthquake Authority Fills Need. January 2000.
Structural Engineers Association o f California (SEAOC). Vision 2000
committee report, 1994.
Taversky, A. and D. Kahneman. “Choices, Values and Frames.” American
Psychologist, 39(4), (April 1984): 341-350.
— “Judgment under Uncertainty: Heuristics and Biases.” Science 185, (1974):
1124-1134.
Taversky, A.S. Sattah and Paul Slovic. “Contigent W eighting in Judgment and
Choice.” Psychological Review, Vol.95, (1988): 371-384.
Thaler, Richard, and William Gould. “ Public policy Toward Life Saving: Should
Consumer Preference Rule?” Journal o f Policy Analysis and Management,
Vol. 1, No.12 (1982): 223-242.
Thompson, M. and A. Wildavsky. “A Proposal to Create a Cultural Theory of
Risk.” in The Risk Analysis Controversy: An Institutional Perspective,
(Kunreuther, H. C. and E. V. Ley, Eds.), Berlin, Germany: Springer Verlag,
1982, p.150-152.
— “Choosing Preferences by Constructing Institutions: A Cultural Theory of
Preference Formation.” American Political Science Review, Vol. 81, No.
1, (1987):5-7.
Tierney, Kathleen and Joanne Nigg. “Earthquake in the Central and Eastern
United States: Social, Economic and Policy Aspects." in Socio-Economic
Impacts, Monograph 5, National Earthquake Conference, Memphis,
TN.1993, p.3-16.
Tribe, L.H. “Policy Science: Analysis or Ideology?” Philosophy and Public
Affairs, 2 (1972): 66-110.
Trumbull, W illiam N. “Who has Standing in Cost-Benefit Analysis.” Journal o f
Policy Analysis and Management, Vol. 9, No. 2 (1990): 201-218.
Turner, Ralph, Joann Nigg and Denise Heller Paz. W aiting fo r Disaster,
Earthquake Watch in California. Berkeley: University of California Press,
1986.
241
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Vargas, L.G. “Comparison o f Three Multicriteria Decision- Making Theories:
The AHP, M ultiattribute Utility Theory and Ranking Methods." Proceedings
of the 3rd International Symposium on the AHP, Washington, DC, 1994,
p. 19-29.
Weaver, S. “The Passionate Risk Debate.” The Oregon Journal, April 24,
1979.
Weimer, David and Aidan Vining. Policy Analysis: Concepts and Practice.
Prentice-Hall, 1989, p.233-244.
White, Louise, G. “Policy Analysis as a Discourse.” Journal o f Policy Analysis
and Management Vol. 13, No.3, (1994): 506-525.
Whitman, Robert e ta l. “Development of a National Earthquake Loss
Estimation Methodology.” Earthquake Spectra, (Nov. 1997): 643-662.
W hitting, Dale and Duncan MacRae, Jr. “Comment: Judgment about Who has
Standing in Cost-Benefit-Analysis.” Journal o f Policy Analysis and
ManagementVol. 9, No. 4 (1990): 536-547.
Witt, James Lee. Building a Public/Private Partnership in Emergency
Management FEMA, Washington DC, 1998.
Wortman, P.M. “Evaluation Research: A Psychological Perspective.” American
Psychologist 30 (1975): 562-575.
Wynne, B. Risk Management and Hazardous W aste. New York: Springer-
Verlag, 1987.
- “Risk and Social Learning: Reification to Engagement." in Social Theories
o f Risk. (Krimsky.S. and D. Golding, Eds.), Westport, CT: Praeger
Publishers, 1992.
of the copyright owner. Further reproduction prohibited without permission.
APPENDIX A
243
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix A
Questionnaire Respondents
Structural Engineering
• President of a well known seismic engineering consulting firm.
• Principal in a well reputed seismic engineering consulting firm.
• President of a specialized earthquake risk analysis firm and a
contributor to development of HAZUS (federal government
earthquake mitigation technology).
• Member of committee for ASTM that writes standards for
building risk evaluation.
• President of a structural engineering firm that advises owners on
seismic risk in their facilities.
• Former president of the Structural Engineers Association of
California and a principal in a seismic risk evaluation firm.
• Former Director of Seismic Mitigation Program - National
Science Foundation, and now a seismic risk evaluation
consultant and Chairman of ASTM - task force on guidelines for
building risk evaluation standards.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Civil Engineering
• Seismic civil engineering consultant and former executive
director of the
• Principal of a prominent geo-technical consulting firm in
Southern California.
• Executive Director o f the Seismic Safety Commission.
Public Officials
• Director of Planning and Building for a mid-size city in Southern
California.
• President of a regulatory body comprised of building officials
responsible for the Uniform Building Code provisions.
• A senior earthquake engineer for FEMA in Southern California
and a former university professor at the University of Michigan.
• Building official for major Southern California city.
• Building officials from two mid-size cities in Northern California.
• Representative of a major utility company and former President
of the Earthquake Engineering Research Institute.
Community Preparedness
• Director of the Southern California Earthquake Center.
245
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
• Executive Director o f the Earthquake Engineering Research
Institute and urban planner in the Bay Area.
Owner's Representative
• Office o f the Chancellor at UC Berkeley.
• Assistant Chancellor of Facilities for the UC system.
Industry Representatives
• Executive Director o f the Concrete Masonry Association of
California.
• Principal of a large construction company.
• Executive Director o f the Portland Cement Association for Codes
and Standards.
• Executive Director o f the American Iron & Steel Institute.
Seismic Researchers
• Director of the Pacific Earthquake Engineering Center at UC
Berkeley.
• Deputy Director o f the Mid-America Earthquake Center.
It is recognized that most of the identified experts perform duties in various
capacities; e.g. structural and civil engineers act as owner’s representatives,
seismic researchers act in a public official capacity. The disciplines identified
closely reflect their direct associations.
246
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX B
247
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix B - Questionnaire No.1
Existing Facilities
It is presumed there is a strong likelihood of a major earthquake occurring in
your area within the next 20 to 30 years. A major earthquake is defined as
one measuring 7.0 or higher on the Richter scale. (For your information, both
the Loma Prieta and Northridge earthquakes measured below 7.0 on the
Richter scale. The greater the magnitude, the greater the earthquake force.)
You may want to read all the questions quickly before answering each one
separately.
The questions are divided into four major areas:
• Life Safety
• Damage to Physical Facilities
• Operability of Critical Facilities
• Preservation of Community Assets
(Please rate or rank the following questions using a ■/ or numbers as appropriate.)
A. Life Safety
1. How important is it to provide life safety in physical facilities? Rank ‘9’
being “least important.”
□ 1 □ 6
□ 2 □ 7
□ 3 □ 8
□ 4 □ 9
2. Please rate, “To provide life safety, mandatory retrofit of existing
property is
necessary.”
□ 5
□ Strongly agree
□ Agree
□ Slightly agree
□ Slightly disagree
□ Disagree
□ Strongly disagree
248
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3. Please rate, “To provide life safety, voluntary retrofit o f physical facilities
should be encouraged.”
□ Strongly agree
□ Agree
□ Slightly agree
B. Damage to Physical Facilities
1. Rank in order of importance (1,2,3) those physical facilities that need to be
seismically retrofitted. Rank 3 being the least important.
Buildings in the community.
Transportation infrastructure serving a community.
Utilities servicing a community.
C. Operability o f Critical Facilities
1. Rank in order of importance (1,2,3,4,5,6,7) the continued operation of the
following critical facilities. Rank 7 is lowest.
Senior citizen housing.
Urgent care hospitals.
Non-urgent care hospitals
Schools (k-12).
D. Preservation of Community Assets
1. Rank in order of importance (1,2,3,4) the following community assets.
Rank 4 is lowest.
Historic structures.
Non-historic structures, but
unique to community,
(ballparks, art centers).
Religious facilities (churches,
temples, etc.).
Museums.
249
Transportation management
centers.
Fire stations.
Police stations.
□ Slightly disagree
□ Disagree
□ Strongly disagree
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix B - Questionnaire No.2
Acceptable Seismic Risk
- A Community Decision -
Exercise
S cenario
Your community is deciding on the acceptable level o f seismic risk. It is
anticipated your community will face a major seismic risk hazard in the next 20
to 30 years. The resources of the community are limited. The stakeholders
need to decide how much risk is acceptable and what priorities should be
allocated to various attributes o f seism ic risk.
(Please keep in mind that disaster assistance from state or federal sources will
be limited.)
Your community has, through societal debate, decided to build up the
maximum capacity possible to absorb as much seismic risk as possible, with
the goal o f minimizing the overall im pact of a seismic event on the community
and maximizing life safety considerations.
D ire ctio n s
The attributes o f acceptable risk are identified and defined on the next page.
You, as an expert, are asked to rate each attribute in comparison with the
others. The following scale is used for pairwise comparison purposes.
V e rtica l Scale fo r Pairwise C om parisons o f A ttrib u te s
Verbal Judgment Numerical Equivalents
Extremely more important 9
Between very strongly and extremely more important 8
Very strongly more important 7
Between strongly and very strongly more important 6
Strongly more important 5
Between moderately and strongly more important 4
Moderately more important 3
Between equally and moderately more important 2
Equally as important 1
250
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Your response is extremely critical to determine acceptable seismic risk.
In the attached table, please fill out a numerical rank for each row attribute as
compared with each column attribute. Please do not fill in the shaded areas.
Definitions of Attributes
1. Life Safety - defined as the ability to allow building occupants to exit
safely.
2. Infrastructure Damage - defined as damage to utilities, communications
networks, and transportation networks that would affect the community.
3. Business Interruptions - defined as interruptions to business operations
for a limited duration.
4. Economic Impact - defined as the cost o f repair of buildings and
infrastructure, and the long-term impact on the community’s economic
health, as well as short and long-term impacts on health of citizens.
5. Operability of Critical Facilities - defined as the need to have facilities
such as hospitals, schools, fire stations, and police stations in operable
condition during and after an earthquake.
6. Preservation of Community Assets - defined as, for example, museums,
libraries, religious worship facilities, sports arenas etc.
7. Community Preparedness - defined as the preparedness of community
groups to volunteer services in various social areas (e.g., assistance in
food delivery, paramedical services etc.).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Exercise
(complete only those squares which are not shaded)
Pairwise Comparison Table
Critical
Facilities
O perability
Community
Asset
Preservation
Comm.
Prepared
Physical
Facilities
Damage
Business
Interrupt.
Economic
Impact
Physical
Facilities
Damage
Business
Interrupt.
Econom ic
Im pact
Comm.
A sset
Preserv
Comm
Prepared
252
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Determining risk propensity of government program managers for high risk /high payoff projects

PDF

Charter schools' influence on public school administrators' innovative behaviors

PDF

Development of a family risk-factor measure that predicts imminent risk of placement and appropriateness for family-based, wrap -around services

PDF

Detecting the effects social and business pressures on small California trucking firm tax compliance

PDF

A configuration study of multiagency partnerships as practiced in Taipei City government

PDF

A financial audit model for entrepreneurial governments

PDF

Assessing United States information assurance *policy response to computer -based threats to national security

PDF

Adapting and applying a mission-focused strategic framework for emergency management

PDF

Evaluation of professional services consultants in rural government

PDF

Educational reform for private state -approved schools in California

PDF

Conversion of health care organizations from non -profit to for -profit status

PDF

Drug treatment providers' organizational responses to implementation of California's Proposition 36

PDF

Executive spending power: Flexibility in obligation and outlay timing as a measure of federal budgetary and policy control

PDF

Historical perspectives and future horizons of local government managers and the International City /County Management Association

PDF

Care management for the uninsured: A force field analysis of the business case

PDF

Do uses of human resource information technology (HRInT) tools in federal organizations improve their human resource management productivity?

PDF

Internal venturing in public agencies

PDF

Information overload: Exploring management of electronic mail

PDF

Bridled boldness: Reengineering the Federal Bureau of Prisons. Lessons learned

PDF

An analysis of health risk selection and quality of care under Medicare fee -for -service and Medicare managed care health care systems

Asset Metadata

Creator Mujumdar, Vilas Sitaram (author)

Core Title Determining acceptable seismic risk: A community participation-based approach

School School of Policy, Planning and Development

Degree Doctor of Public Administration

Degree Program Public Administration

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag engineering, civil,OAI-PMH Harvest,Political Science, public administration

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Clayton, Ross (committee chair), Newland, Chester (committee member), Petak, William (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-83511

Unique identifier UC11327931

Identifier 3018111.pdf (filename),usctheses-c16-83511 (legacy record id)

Legacy Identifier 3018111.pdf

Dmrecord 83511

Document Type Dissertation

Rights Mujumdar, Vilas Sitaram

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA