REPEAT RESPONSE TO HURRICANE EVACUATION ORDERS
Return to Hazards Center Home Page
Return to Quick Response Paper Index
This material is based upon work supported by the National Science Foundation under Grant No. CMS-9632458. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
These two events offer a unique opportunity to study the impact of repeated "false alarms" - evacuations ordered based on expectations of hurricane landfall that ultimately proved to be wrong. The influence and credibility of forecasters and local officials, important factors in the evacuation response, could be affected by this recent history of repeated "misses," "near-misses," and "hits." The impact of false alarms on future evacuations is a widespread source of speculation and concern in the emergency management community, but very little research has directly studied the implications of these unnecessary evacuations (Baker, 1991).
Using these two hurricane events as case studies, this study employed short, face-to-face surveys to examine three specific questions:
The surveys involved face-to-face interviews with respondents at the entry to major stores (grocery and discount) in each community. This design was selected in order to focus on the residential populations of these heavily touristed communities. The survey included both open-ended and closed questions and required from five to ten minutes to administer. Less than two weeks after Hurricane Fran, survey teams were in the field in South Carolina (September 13-15, 1996). Because of the extent of damage, we were unable to get into the Wilmington area for interviewing until two weeks after that (September 28). We used a screening question to target residents and those who were in the area for both hurricanes. This approach yielded a total of 526 responses for a total response rate of 75%, which included 323 completed surveys - a completed survey response rate of 45%.
Hilton Head Myrtle Beach Area Wilmington TOTAL SC SC NC Survey Date 9/13 - 9/14/96 9/14 - 9/15/96 9/28/96 9/13-9/28/96 Total Surveys Not in area 32 177 4 213 Completed 128 143 52 323 Refusals 79 76 24 179 Total Contacts 238 396 80 714 Response Rate (%) 67.2 80.8 70.0 74.9 Complete Survey Response Rate (%) 53.7 36.1 65.0 45.2
On September 5, 1996, Hurricane Fran made landfall over Cape Fear just south of Wilmington, NC as a weak Category 3 storm (Mayfield, 1997). Fran's maximum sustained winds were 115 mph with a reported storm surge of 12-15 feet. Tropical force winds extended 290 miles seaward of the storm's center and up to 100 miles inland. The primary damage was in North Carolina with wind and storm surge damage ranging from Corncrake Inlet in the south to New River inlet in the north. The hardest hit barrier island communities included Carolina Beach, Wrightsville Beach, Topsail Beach, and North Topsail Beach. Inland, flooding from the already swollen rivers and saturated ground was extensive throughout the state.
In South Carolina, damage from Hurricane Fran was limited to downed trees, some flooding in low-lying areas, power outages, and some roof damage. A federal disaster declaration was issued on September 6 for North Carolina and Virginia. The federal disaster declaration for the four northeastern-most counties in South Carolina followed on September 30.
Location # Evacuated prior to # Evacuated prior to Total Interviews Bertha (% of Fran (% of community sample) community sample) Hilton Head, SC 64 (50%) 83 (65%) 128 Myrtle Beach, SC 48 (34%) 76 (53%) 143 Wilmington, NC 6 (12%) 17 (33%) 52 Total 118 (37%) 176 (55%) 323
Despite the difference in the storms, there was considerable consistency in individuals' choices. Of the 323 respondents, 43% did not evacuate for either hurricane, 34% evacuated for both, and 21% left in Fran but not in Bertha. The remaining 3% (9 individuals) left for Bertha, but remained for Fran. There were no differences in evacuation behavior attributable to the sociodemographic characteristics of respondents. Further, 75% of the respondents had previously experienced a hurricane (Question 23).
For the majority of residents, the timing of the evacuation advisories were just about right - not too early nor too late (Question 16). However, nearly 20% of respondents felt the evacuation orders for Bertha were premature while 15% believed that to be the case for Fran. Again there was considerable difference between the three communities with 35% and 18% of the Hilton Head residents suggesting that the evacuation orders came too early for Bertha and Fran respectively. In Myrtle Beach and Wilmington, only between 4 and 13% of the respondents felt that the warnings were issued too early.
First Reason Given Bertha Fran Bertha Fran % of those % of those % of total % of total who who survey survey evacuated evacuated population population (n=118) (n=176) (n=323) (n=323) Governor's order/advice 24.6 22.2 8.9 12.0 Local officials/local emergency responders 11.0 5.7 4.0 3.1 Weather Channel/ National Weather Service/local news 25.4 25.0 9.3 13.6 actions/advice from friends or family 6.7 7.9 2.5 4.3 Severity of storm/probability of "hit" 9.3 16.5 3.4 9.0 Other 24.6 23.9 9.0 13.0 Did not evacuate NA NA 62.8 44.9
What is interesting to note as well is the decline in the reliance on government officials from Bertha to Fran. This is partially explained by the frustrations of the Hilton Head residents over traffic jams and other difficulties with emergency operations encountered during the Bertha evacuation two months earlier. In fact, 47% of respondents reported problems in trying to leave the area for Bertha and more than half of the evacuees reported likewise for Fran (53%) (Question 6). Yet, the majority of residents felt the warnings and evacuation process was handled extremely well (32% in Bertha, 49.4% in Fran) (Question 14).
Reasons for non-evacuation centered on the perception of safety of the housing unit and/or location relative to the danger (Table 4). Non-evacuees also mentioned the probability of landfall and information from the Weather Channel as reasons given for staying.
Reason # Respondents (%) n=182 Severity of storm/probability of a "hit" 37 (30) Weather Channel 15 (8) Perception of safety of home/location 43 (24) Job requirements 12 (7) Other (contains no category greater than 4%) 61 (34)
Bertha Fran % responses % responses n=323 n=323 Governor's order 2.8 2.8 Media - TOTAL 39.4 52.5 Media subtotals Media - general 25.2 31.3 Media - local 3.4 5.0 Media - Weather Channel 6.8 10.9 Media - radio 4.0 5.3 Actions of friends/neighbors 0.9 1.6 Personal experience 1.2 1.6 Other 1.2 1.9 Not Applicable 54.3 39.7
For those who evacuated for Fran and not Bertha (67 respondents), the most reliable sources of information mirrored the overall results. In descending importance these were the media 42%, the Weather Channel (18%), the radio (13%), local media (6%), and the actions of family/friends and personal experience (13%). Media sources were more widely used in this group of evacuees.
The broader response of listening to the media, coupled with specific references over and over again to the Weather Channel suggests a greater independence in evacuation decision making. The Weather Channel became an independent source of information and confirmation of the storm's direction and severity. Several respondents also reported logging onto the Internet to get information; however, local emergency managers reported receiving phone calls asking for assistance in interpreting the statistics that residents found on the web.
The seeming difference between reliable sources of information and factors that convinced people to evacuate are important. Our respondents found media (of all kinds) an important source of information for hurricane characteristics, although in many cases it was orders from the governor and other considerations that convinced them to leave.
Of the 98 people who said simply "yes," the clear majority, 55%, explained their concerns over personal and family safety in general terms such as "I'm afraid," "better safe than sorry," or "be crazy not to [evacuate]." Others gave specific reasons, including past experience with hurricanes, living in a mobile home, and living near the water. The experience of "false alarms" did not seem to sway their perception of the risk. For 73% of this group, the plan to evacuate was consistent with their past evacuations for both Bertha and Fran.
Among the 70 who responded "no" were the few respondents (n=7 or 2% of the total population ) who indicated that they would not evacuate because the last few forecast hurricanes did not hit. Comments included, "Government leaders cried wolf too often" and "the last few didn't hit." The largest group, however, were those who believed their home was a safe place or at least safer than anywhere else (n=24 or about 7% of the total population). These respondents explained that their homes were sturdy, located on hills, or that they felt safer in their brick homes than out on the road. The second largest group were those whose job required that they stay (n=9 or about 3% of the total population). Most other responses, such as concern about being able to return to home, were shared by only a few individuals. The majority of these respondents (80%) did not evacuate for either Bertha or Fran.
The indirect effects of these "unnecessary evacuations" might be expected to be more apparent in the responses of the group (n=152 or 47% of the total population) that replied their actions would "depend" on the situation. The deciding factors that respondents gave in response to this question differed substantially from the factors that "convinced them to evacuate" for Bertha and Fran. While the actions of public officials are often reported as the most influential factors determining evacuation rates, these responses coupled with the reliance on the news sources including the National Weather Service, the Weather Channel, and the Internet for information on the hurricane suggest that individuals are relying on a variety of information sources to make personal risk evaluations.
The largest group within this subset, 16%, replied that their evacuation depended on the severity of the storm, but gave no indication of magnitude. Twelve percent responded that they would evacuate for storms classified as Category 4+. Others responded it would depend on a variety of storm conditions, such as path, probability of a hit, and potential for flooding. In total, the number of respondents planning to rely on evaluation of storm characteristics totaled 73%. In reporting the factors determining their willingness to evacuate, 10% of the group answered that their evacuation would depend on an order from the governor.
Within the total survey group, the advice of public officials, which convinced approximately 13% of respondents to leave for Fran, is accorded only a minor role in future decision making. While 4.6% of the total population would listen for the governor's order, there is no mention of actions by other officials (either in this group or among the "yes" respondents discussed earlier). While this factor did not decline in importance between Bertha and Fran, neither was it accorded a significant place among criteria for future events.
Personal assessment of the storm characteristics and its risks played a larger role. Individuals considered the quality of home construction and location, family safety and needs, data on storm tracks, strength and probabilities, relying on the media, and the Weather Channel in particular, to a larger degree than previously reported. Individual decisions about evacuation in these events was quite consistent. Those who evacuated for one storm and not the other (mostly for Fran and not Bertha) did so largely on the basis of storm characteristics as warnings were nearly identical. Differences in evacuation rates among communities were not accounted for by differences in level of risk, demographics, or community hurricane history. These differences suggest the need for greater investigation of community disaster culture and perceived relationships (or personal/professional relevance of other groups' concerns) among residents, authorities at different levels of government, and hurricane experts.
FEMA, 1996. Final Situation Report for Hurricane Bertha. http://www.fema.gov/hu96/besrp6.htm
Lawrence, M.B. 1997. National Hurricane Center, Preliminary Report Hurricane Bertha: 5-14 July, 1996. http://www.nhc.noaa.gov/bertha.html
Mayfield, M. 1997. National Hurricane Center, Preliminary Report Hurricane Fran: 23 August - 8 September, 1996. http://www.nhc.noaa.gov/fran.html.
Date:________________ Location:_____________ ID:__________________ HURRICANE EVACUATION RESPONSE QUESTIONNAIRE My name is ________________________ and I'm from the Geography Department at University of South Carolina in Columbia. We're undertaking a study of evacuation responses from residents in this area for both Hurricane Bertha and Hurricane Fran. Would you mind if I ask you a few questions? 1. Were you in the area when Hurricane Bertha or Hurricane Fran began to threaten the South Carolina coast? Yes(%) No(%) Bertha 92 8 Fran 98.5 1.5 If NO, then terminate the interview and thank respondent for his/her time. If YES, then continue. 2. Did you leave your home to go someplace safer Yes(%) No(%) before Hurricane Bertha hit? 36.5 63.5 before Hurricane Fran hit? 54.5 45.5 If YES, then continue. If NO, then go to Question #12. 3. Where did you go? (Check the category that is the closest to respondents answer. Be sure to ask question for each hurricane) Bertha Fran (%) (%) public shelter 1 2 motel/hotel 16 26 friend/relative 12 21 other (specify) 3 4 not applicable 63 46 4. Where (city or town, county) was that located? Bertha_______not summarized______________ Fran________ not summarized ______________ 5. How long did it take you to get there? (hours to nearest 1/2 hour) Bertha -- average 3.2 hours Fran -- average 3.0 hours 6. Did you experience any problems in trying to leave the area? Yes(%) No(%) Bertha 47 53 Fran 19 81 If yes, what were they?____________not summarized______________ 7. What convinced you to leave the area prior to landfall for Hurricane Bertha? For Hurricane Fran? (Ask as an open-ended question, place a 1 in the column for first reason, number 2 for second reason, and number 3 for third). First Reason Given (% of total respondents; N=323) Bertha Fran order from the Governor 7.7 11.1 advice from the Governor 1.2 0.9 order/advice from local public officials 3.4 1.9 order/advice from local/state police or firemen 0.6 1.2 advice from the National Weather Service 3.1 5.6 advice from the media (Weather Channel, local news) 6.2 8.0 advice/actions from friends/relatives 2.5 4.3 the severity of the storm 1.9 5.6 the probability of the storm hitting this area 1.5 3.4 other (list) 9 13 not applicable 62.8 44.9 Second Reason Given (% of total respondents; N=323) Bertha Fran order from the Governor 2.5 3.1 advice from the Governor 0.3 1.2 order/advice from local public officials 0.6 0.6 order/advice from local/state police or firemen 1.5 1.9 advice from the National Weather Service 0.9 2.8 advice from the media (Weather Channel, local news) 3.1 4.0 advice/actions from friends/relatives 0 0.9 the severity of the storm 4.0 6.5 the probability of the storm hitting this area 0.3 1.5 other (list) 4.6 7.4 not applicable or not answered 82.0 70 Third Reason Given (% of total respondents; N=323) Bertha Fran order from the Governor 0.3 0.6 advice from the Governor 0.3 0.3 order/advice from local public officials 0 0 order/advice from local/state police or firemen 0 0.6 advice from the National Weather Service 0.3 0.6 advice from the media (Weather Channel, local news) 0.6 1.2 advice/actions from friends/relatives 1.6 1.9 the severity of the storm 0.3 0.6 the probability of the storm hitting this area 0.6 0.6 other (list) 2.8 5.0 not applicable or not answered 93.2 88.5 Bertha_________________________________________ Fran___________________________________________ 7a. If there was a difference in responses between Bertha and Fran, ask respondent to explain. Record answer below. 8. What was the most reliable source of information that led to your decision to leave? Check the one that applies. (Interviewers - please follow the exact wording) % responses Bertha Fran Governor's order 2.8 2.8 Media - general 25.2 31.3 Media - local 3.4 5.0 Media - Weather Channel 6.8 10.9 Media - radio 4.0 5.3 Actions of friends/neighbors 0.9 1.6 Personal experience 1.2 1.6 other 1.2 1.9 not applicable 54.3 39.7 Bertha ___________________________________ Fran______________________________________ 8a. If there was a difference in responses between Bertha and Fran, ask respondent to explain. Record answer below. _________________________________________________________________ 9. What day/time did you leave the area? Use 24 hour clock and date/day Bertha - not summarized Fran - not summarized 10. When did you return home (day/time)? Use 24 hour clock and date/day Bertha - not summarized Fran - not summarized 11. In this area, where was the formal evacuation zone for Hurricane Bertha? not summarized Don't know _______ Hurricane Fran? not summarized Don't know________ 12. What made you decide not to go anyplace else (or not evacuate)? _________________________________________________________________ 13. If another hurricane approached the South Carolina coast, would you decide to leave the area and why? (Be sure to ask this as one question; Check Yes or No and fill in why) % responses Yes 30.7 No 21.6 Depends 47.6 Why _________________________________________________________________ Any additional comments here 14. How well do you feel the warning and evacuation process was handled? Check the category that applies. Bertha Fran % % extremely well 32.4 49.4 well 29.9 34.0 not very well 17.0 5.5 don't know 5.2 4.9 no answer 14.5 9.6 Any additional comments here 15. Do you feel you received enough information to make a good decision to evacuate or stay? % response Yes 91.3 No 4.6 No Answer 4.0 Any additional comments here 16. Do you feel you were told to evacuate . . . yes no no answer (%) (%) (%) Bertha - too early 19.7 57.1 22.5 Bertha - too late 5.5 71.3 22.5 Fran - too early 14.5 68.5 16.4 Fran - too late 2.5 81.2 16.4 We have just a few background questions we'd like to ask. 17. How many people live in your household? adults children (%) (%) 0 0 70.0 1 16.1 12.4 2 67.8 10.8 3+ 15.8 6.8 18. Is there anyone in your household with special needs? (Prompt: For instance, needing regular medical care) respondents (%) yes 12.1 no 87.9 19. How many vehicles do you have? # of cars 0 1 2 3 4+ % of respondents 2.5 31.7 46.6 13.7 5.5 20. How long have you lived here? Length of residence % of respondents Less than 1 yr 7.1 1-2 9.7 2-5 15.9 5-10 19.5 10-15 12.2 15-20 15.5 20+ 9.3 21. Do you own or rent? % of respondents own 77 rent 23 22. Is your home a fixed structure, a mobile home, or an RV? % of respondents fixed structure 84.5 mobile home 15.5 23. Have you experienced a hurricane before? % of respondents yes 74.9 no 25.1 24. Would you be willing to participate in a follow-up study? 233 (72%) people responded yes If yes, get name, address and phone number. Name___________________________________ Address_________________________________ Phone Number____________________________ Thank you very much for your time. Interviewers, please record the following information after respondent leaves. Demographics Race % respondents Anglo-American 90.6 African- American 8.2 Other 1.3 Sex % respondents Female 46 Male 54 Approximate Age % respondents 18-45 43.4 46-64 27.4 65 or older 29.2
Return to Hazards Center Home Page