Click for next page ( 101


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 100
Working Paper B User Needs IDENTIFICATION OF THE USER CO~IUNITY Early in the panel's deliberations it became clear that one of the most important considerations in exarn~ning the different methods of estimating earthquake losses would be the users' needs. This required defining who the users were so that their particular needs could be reflected in the panel's assessment of different loss estimation methods. Many different groups and sectors potentially could have been included as users. A subpanel developed several sampling strate- gies, as well as research designs for a comprehensive study of user groups. For example, the panel could have considered the needs of such diverse elements as federal, state, and local adrn~nistrators and officials, insurance companies, bonding companies, social scientists, the engineering and scientific communities, public information insti- tutes, and other groups. Indeed, the different user groups in each of these sectors pose very complicated sampling and design problems. After considerable discussion with FEMA and USGS representatives, the user group was defined to include only state, county, and local public officials. Several factors led to this simple definition of users: ~ The scale of a study that would include all potential groups would be large. 100

OCR for page 100
101 o The time frame of the pane} was relatively short, prohibiting a large-scare study. Funding for such a large-scale study was not possible within the panel's budgetary constraints. ~ The funding agency's major clientele were public sector enti- ties invested with protecting the public health, safety, and welfare. Given this user group the panel's goal was to determine needs and to evaluate different methods in terms of meeting the users' requirements. It was believed that the better such studies met the requirements and needs of the user community the more likely the studies would be utilized in planning for, responding to, mitigating the effects of, and recovering from a major damaging earthquake. Even with this limited definition of the user community, the selection of state and local officials for inclusion in the study presented some significant problems that limit the extent to which the subpanel's recommendations may be generalized and that warrant discussion. Options for obtaining the views of users included a questionnaire survey based on a scientific national sample, a similar survey with a smaller sample, in-depth discussions with some very experienced users, and a workshop. The workshop option was selected primarily on the basis of time and budget. The results of opinions solicited in the workshop, presented later in this paper, should not be construed to be statistically valid as a representation of state and local users. The Method for Selecting Users The first step was to obtain a list of users that could be used in constructing an appropriate sample. After extensive consultation with USGS, FEMA, and the COSMOS Corporation, consultant to the panel, a user was defined to be an appointed or elected public official who could be involved in developing data for use in loss esti- mation studies or in making decisions, based on those studies, which resulted in a lowered risk to the community. This definition, although limited, included officials in such functional positions as mayors, city managers, planners, directors of public works departments, building code officios, county commissioners and managers, and emergency service personnel at the local and state level. The geographic scope of the list was limited to approximately a dozen higher seismic risk areas of the United States (Table B-1~. The panel could not develop what it considered an adequate list.

OCR for page 100
102 TABLE B-1 Participants in the Workshop, by Area and Level of Government Level of Government Area City County/Regional State California 3 emergency 1 regional earthquake 1 state earthquake services program manager program manager officials 2 building code officers 1 city manager 1 planner Central 1 county commissioner 1 state emergency United States services official Northeast 3 emergency 1 state emergency services services official officals . Puget Sound, 1 city council Washington member 1 emergency services official Utah 1 mayor 1 county residential 2 state emergency supervisor services officials Alaska 1 planner 1 state emergency services official Hawaii 1 state emergency services official South Carolina 1 state emergency services official Puerto Rico 1 state emergency services official Total 14 3 9 Eventually the list of local government users included some plan- ners, building code officials, and a few council members, mayors, and managers (Table B-1~. The state list of users was overrepresented by emergency service managers. Finally, the list was overly representa- tive of California users, which is not surprising given the fact that over a dozen loss studies have been conducted there. The workshop invitees were selected from lists of potential users supplied by federal agencies, and very few actual users are present in this pool at this time; there was a limited representation of the functional positions at the state and local levels, and geographic affi~- iations were not nationally representative. All of these factors make it necessary to address briefly the limitations of the data collected.

OCR for page 100
103 Data Sources and [mutations The pane! has relied on the results of the survey, the small group discussions, and the presentations by technical and user community members to suggest the users' needs. This information has been uti- lized to broaden the perspective of the panel in its deliberations. The workshop provided the key instruments for gathering data, and a few factors require discussion. The pane! fully recognizes the Irritations of these data sources and the fact that generalizing solely on the basis of the workshop cannot be done with much certainty. The major goal of the workshop held September 22, 1986 was not to train or even educate the participants in loss estunation studies. Rather, participants were invited to educate members of the pane] about the requirements and needs of the community which would or potentially could utilize loss studies once they were completed. The workshop was designed to provide several different approaches and methods by which panel members could determine user needs. The survey instrument administered to participants was designed to gain insight into participants' needs and familiarity with loss stud- ies. The instrument was administered twice during the workshop, but not to deterrn~ne the effectiveness of the workshop. The pane! was far more interested in the responses to the first questionnaire prior to participant exposure to the speakers, because of the focus on determining what state and local users (or potential users) of loss studies believed they needed to utilize such studies. Hence, the find- ings below focus almost entirely on the users' responses to the first questionnaire. Finally, small group discussions addressed four questions, three of which were common to all groups. The questions reflected issues briefly covered in the questionnaire but requiring additional atten- tion. These data sources are utilized in this working paper and are suggestive and informative. The pane! makes no claim that the findings can be generalized to the larger user community. If INDINGS The discussion in this section is based on findings about user needs from the three data sources discussed earlier.

OCR for page 100
104 Usefulness of Prior [ose Studies A major concern of the panel and the federal agencies focused on the widespread belief that previous loss studies were not being adequately utilized by the user community. Many of the pane} mem- bers who had been involved in some of these loss studies indicated their disappointment over the lack of use. As a result, workshop participants were asked about their exposure to such studies and the usefulness to their agencies and units of government. Twelve of the 26 workshop participants indicated either that they had never seen the results of a loss study or that the study in which they had participated was not yet completed. The remaining 14 participants were asked how useful the results of the studies were for a variety of activities: mitigation efforts, planning and preparedness efforts, response and recovery planning, land-use planning, building code design, and efforts to educate the public and elected officials. Participants did not indicate that the results were very useful for any one activity, but a majority of participants found these studies useful (either very or somewhat useful) for the spectrum of activities. The most important use of these studies, according to small group discussions and questionnaire results, was their use in educating elected officials and the public about the seriousness of seismic threat and the need to take action. There was clear agreement that such studies have been and should be used to advocate the unportance of seisrn~c programs in order to obtain greater emphasis on actions that reduce the ef- fects of an earthquake. Additional uses that received strong support among participants were public awareness and education programs and emergency response planning. General Barriers to Utilization Participants were asked why they believed loss studies were not utilized in developing public policy. Most important, and a general theme in the utilization issue, was the lack of involvement by state and local officials and policymakers in the entire study process. Par- ticipants indicated that too often the "experts" conducting the loss studies proceeded without regard to whether the users would under- stand what the results addressed or meant. In addition, some users indicated that the conflicts and disagreements among professional and technical experts had seriously undermined any efforts to utilize such studies. Workshop participants also stated that some reports

OCR for page 100
105 were completed in an untimely fashion and, when delivered, were written too technically. All of these factors seem to contribute to the final barrier to using these studies the lack of support among elected officials for taking action and making policy. If the above-identified barriers were removed it would not ensure greater support from policymakers, but it would help. If these loss studies are to be used, in part by advocates of seismic planning and policy, then officials must be involved in the loss estunation study process and the reports must be understandable, less technically presented, and timely. Defining the Seimn;c Hazard: The Earthquake Scenario An issue that has emerged in many loss studies emphasizes how helpful it Is to policymakers and planners to have a loss study based on the most damaging historical earthquake. Participants in the workshop strongly indicated their desire to have studies focus on major but likely earthquakes. In addition, participants (about two- thirds) believed it was either very or somewhat important to have different estimates of loss for different seasons of the year. Finally, all of the participants believed it important that losses be estimated for earthquakes occurring at different times of the day. In short, if the users' needs are to be met, loss studies should include these features: most likely earthquake to cause significant damage, seasonal estimates of loss, and estimates for the event occurring at different times of the day. Geographic Focus of Study Users at the state, region, county, and local level have differ- ent needs and requirements. In addition, recent research indicates that the key actors' functional positions influence their support for seismic planning and policy (Mushkatel and Nigg, 1987~. Nowhere are these different needs more manifest than in the data addressing the geographic focus of the studies. The small group discussions strongly indicated that the level of government one is employed by influences the desired geographic focus. Hence state participants wanted the loss studies to be for either states or regions, whereas local government officials desired a local focus. Local government participants used several examples of studies that were of such large geographic focus that they were of little value

OCR for page 100
106 to localities, particularly when the loss estimates and data could not be disaggregated to the local level. In addition, for individuals in some functional positions the most valuable ciata and Toss estimates would be site specific, which may be impossible or at a minimum very costly. Workshop participants also discussed some elements of the inven- tory data used in loss studies. There was agreement that a multitude of data from both public and private sources should be utilized in such studies. Yet there was also the belief that too frequently the data utilized were not maintained or accessible to the users and that in new studies firms or governmental entities had to recollect or re- discover much of the same data. Thus the users urged those doing Toss studies to take steps to standardize the process for collection, maintenance, and dissemination of inventory data. Types of Formation m [ose Studies Loss studies have produced much information about projected losses for different types of structures serving various purposes. Par- ticipants at the workshop were asked to rank the importance of loss estimates to 19 different structures and facilities along a four-point ordinal scale from very important to not at all ~rnportant. Over 90 percent of the participants indicated that estimates regarding emer- gency public facilities (96 percent) and hospitals (92 percent) were very important. Almost as vital were loss estimates for water dis- tribution systems (88 percent), electric power systems (80 percent), hazardous materials storage sites (80 percent), and highway systems (76 percent). The least important information concerning losses ac- cording to workshop participants were port facilities (14 percent) and government buildings (32 percent). These rankings are relative, and tests of statistical significance are inappropriate given the data base and sample. They suggest, however, that the participants seemed to focus on the response and preparedness components of the disaster and the ability of authorities to estimate losses to facilities critical for emergency response. This focus may be a function of the makeup of the participants (functional position) or of the fact that most loss estimate utilization has been identified somewhat with emergency response and preparedness. Specificity, Accuracy, and Credibility of [ose Estimates An issue discussed at length by the pane! is the importance of

OCR for page 100
107 accuracy In loss estimates and the trade-offs between accuracy (cer- tainty) and specificity. To the user community an important point is the credibility of the estimates. A consistent theme was that the earthquake scenario and the estimates of losses had to be generated from a recognized and credible source and, most important, be plau- sible. Small group discussions indicated that some estimates were based on extreme events with such high loss estimates that they had not been taken seriously. In some instances, public actors viewing the expected losses were so overwhelmed they felt local and state action would not be feasible or would not make any difference because the problems were intractable in light of the estimates. Given the amount of error loss studies potentially contain, steps must be taken to ensure the greatest amount of credibility to loss studies. The greater involvement of state and local authorities throughout the loss estimate study process will increase the like- lihood such estimates are taken seriously. Ideally, these Toss estimates could be both certain (estimation of total Tosses) and specific or detailed (Iosses to specific locations or sites). When participants were asked to select between specificity and certainty, a majority chose specificity (60 percent) as the more important to them for utilizing the information. This is especially important for hazard reduction programs. This desire for specificity is not surprising but may cause some difficulty because of the state of the art in loss estunates. Even more disturbing was the fact that of those state and local users at the workshop who had some familiarity with loss estimates, only 17 percent were very confident of the loss predictions. Obviously this lack of confidence contributes negatively to the credibility issue discussed above. One frequently mentioned problem was that the ranges of predicted losses in life and property were too great to be very useful for planning purposes. Finally, participants were asked on the survey to indicate how reliable different loss estimates must be for utilization. The results are difficult to interpret since state and local users might be willing to forego some information reliability if the type of structure or its purpose is sufficiently important. Keeping this potential trade- o~ in mind, participants indicated that it was most important to have very reliable information for dams (48 percent), electric power systems (44 percent), natural gas and water distribution systems (40 percent), and highway systems (36 percent). Participants indicated it was least important to have very reliable information about airports

OCR for page 100
108 (4 percent), radio and television facilities (8 percent), government buildings (12 percent), and residential structures (16 percent). Obviously state and local officials want as much specific and cred- ible information as possible. Yet these requirements, as reasonable as they seem, involve real costs. It is in this light that the infor- mation collected on users' ~nIlingness to spend takes on additional - slgnmcance. Cost, Willingness, and Ability to Spend The issue of the willingness and ability to spend scarce fiscal resources on loss studies by the state and local user community has several important dimensions. First, the proposed sharing of costs between FEMA and state and local governments for other programs may in the near future include the monies used to finance loss studies. lIence, the pane! determined it would be appropriate to investigate not only the needs of users, but also their willingness and ability to spend monies to obtain loss estimates. Furthermore, it is often thought that if a government spends some of its own resources for a study it is more likely to use the results. One of the questions included in the survey requested workshop participants to indicate what amount their office or agency would be willing to spend for an earthquake loss study. The most frequent response category selected was less than $75,000. Because of the way the question was worded it is impossible to deterrn~ne how many of the 71 percent who indicated they would spend $75,000 or less would spend nothing for such a study. Almost 20 percent of the participants did not answer the question at all, and only 10 percent indicated they would spend $225,000 or more. In short, state and local users perceive a lack of willingness or ability for their agencies and offices to expend monies for such studies. In addition, more than 80 percent of the state and local users indicated that their current budgets did not contain adequate funds for such a study. Questions about future budgets were not asked. Finally, despite this apparent inability or unwillingness to fund loss studies, users expressed support for sharing costs. When asked what percentage their governmental units should be responsible for, 59 percent of the participants indicated between 41 and 50 percent, 16 percent signified less than 10 percent, and 11 percent noted more than 50 percent.

OCR for page 100
109 Workshop ciata reveal support among state and local users for cost sharing, but they also show that current budgets are not suf- ficient to assume these costs. The last constraint on willingness to spend involves the total cost of the studies. Workshop participants (71 percent) indicated they would only spend less than $75,000 for a loss study. Hence, there seems to be a strong desire to hold down costs because current budgets are inadequate to finance the studies. The [ose Study Report and its Dilation One explanation for the lack of willingness and/or ability to spend is the users' lack of satisfaction with such studies. The survey data cannot test this explanation, but a consistent viewpoint that emerged at the workshop was that current loss studies are understood only with great difficulty by the user community. A major problem is the results are not presented in a way that makes clear their implications for seismic planning and policy development. Users at the workshop also criticized the presentation of the re- sults, most often citing them as being too technical. They supported the presentation of technical materials in an appendix, rather than in the body of the report. In addition, the problem associates} with inventory and other data bases reemerged. Users want the data to be accessible to them after the report is finished. Such accessibility would permit Aggregation to lower units of government or to a smaller geographic area. The participants often shared the perception that once such studies had been completed they were not disseminated adequately. Too little attention was paid to disseminating the findings to the potentially large community of users. Participants believed that more attention should be given to dissemination in the loss study process, and suggested that either state or local government agencies be responsible for the dissemination of findings to the users. As previously emphasized in this paper, participants strongly believed that to ensure the clarity and dissemination of study results for the largest possible user community, state and local representatives should be involved in the loss study process from its inception. CONCLUSIONS AND RECOMMENDATIONS The user needs subpane} concluded that some previous loss stud- ies may not have sufficiently taken into account state and local users.

OCR for page 100
110 This lack of attention and focus has manifested itself in studies hav- ing a geographic focus and an inventory data base that do not easily permit utilization at the local level. ~ addition, reports have been too technical to be readily understood by users. The earthquake scenario on occasion has produced loss estimates that lacked credi- bility and hence were not useful for planners or policymakers. Too often it seems the producers of loss studies have incorrectly identi- fied other producers of loss studies as being the users of their studies. Too often users have neither received the types of information they thought they were to obtain, nor have they received reports they could understand and disseminate easily. Data assembled from workshop discussions and the survey form the base on which the pane} has based its recommendations. It is important to reiterate that these data may not reflect the needs of the larger state and local government user communities. The pane! believes, however, they are suggestive of those needs. Within the methodological limitations discussed earlier, the following recom- mendations are offered. 1. Producers of loss estimation studies should involve their state and local clientele (the users) in the entire loss estimate study process. Loss estimation should and can be a vehicle of understanding the risk and potential losses from earthquakes. Therefore, the process by which such studies are conducted becomes more important than the actual results. The involvement of state and local users in the entire process of loss estimating will increase the likelihood that these important actors come to understand not only the manner in which the study is carried out but also the nature and extent of the seismic problem. Their involvement will facilitate the utilization and dissemination of the findings as well as the use of such studies for the purpose of advocating greater emphasis on seismic policy. 2. Loss estimate studies should clearly indicate the level of po- tential error In the estimates as well as the confidence of the producers of the estimates for the various components of loss estimates. A consistent workshop theme among users was the desire for credible loss estimates. The state of the art in such studies is not well advanced, and predictions of loss may be in error by a factor of 10. The user community needs to understand where error in prediction is most likely. In addition, it is important for the user community to be able to specify where the most accurate information is needed and to

OCR for page 100
111 know what accuracy is possible. When additional expenditures may result in Tower error factors, this information should be presented. It is also relevant when deciding if the Formation is sufficiently valuable to warrant additional resources. 3. Producers of studies should build an inventory base for loss estimates that can be d~saggregated to the smallest political and geographical unit. State or regional loss studies must present sufficient information for local planning, preparedness, and mitigation activities. By com- piling inventory data so that they can be d~saggregated and accessed by local units, producers win provide the opportunity for smaller units to use their studies. Furthermore, the computerization of data would permit updating and multiple use. For example, if a loss study identifies "suspects buildings in a regional area, each locale could be provided a list of these buildings and their locations to determine if local action is warranted. 4. Loss estunate studies should contain a scenario earthquake that is relatively probable and yet large enough to cause serious losses. Loss estunates should be provided for different seasons and times of the day. This recommendation is consistent with findings from workshop discussions and survey instrument results. About 70 percent of the users indicated these types of information are essential for planning purposes. The producers of loss studies should determune the impor- tance to users of the estimates of loss to different types of structures and functions. In addition, the importance of the certainty and re- liability of the different estimates to the users should be identified, and the studies should be oriented toward these needs. The users at the workshop ranked the importance of 19 differ- ent structures and functions and indicated how reliable loss estimate predictions should be for each structure and function. It is impor- tant to remember, however, that these rankings are only suggestive. IdeaDy, state and local decision makers who are involved in the loss study from its inception can provide producers with more refined definitions of their needs. 6. The dissemination of loss study findings should have greater

OCR for page 100
112 emphasis. State and local users of such studies should be responsible for dissemination to relevant agencies and the public. Previous dissemination of studies appears to have been unsatis- factory, and there is some indication that the dissemination process has been a barrier to utilization. To increase the likelihood of access and use the reports should be as nontechnical as possible. Method- ologica] discussions should be included in appendixes. More emphasis should be placed on the implications of the findings for set c plan- ning and policy adoption. The loss study reports must be aimed at the audience of users and not other producers. Methodological appendixes wid provide the information necessary for replication and validity checks. But the thrust of the report must be concentrated on those who wiD apply the findings the users.