1. THE CALLS ARE COMING FROM INSIDE THE STATE: Phone-banking after Hurricane Irma
Fletcher Durant, Preservation Librarian, UF Libraries
Good afternoon. Thank you to the Society of Florida Archivists for having me here today to talk about last hurricane season, even as we prepare to kick off the 2018 season in...22 days.

2. Harvey Irma Maria 2017 Hurricane Season
For this audience, I probably don’t need to spend too much time recapping what we all went through just a few shorts months ago. Please note that this talk will be focused on Irma and Florida, but other people likely have as much, if not more to say on the preparation, impact, and devastation of Harvey and Maria, as well as the impact of Irma beyond Florida.

3. Hurricane Irma September 10-11, 2017
But for those of you who missed it, Hurricane Irma made landfall at 9:10 AM on Sunday, September 10th on Cudjoe Key in Monroe County as a Category 4 storm and then again on Marco Island in Collier County at 3:35 PM as a Category 3 storm. Mandatory evacuations were ordered in 16 counties with 13 additional counties under voluntary evacuation orders. 6.5 million Floridians evacuated their homes. Hurricane force winds hit one-third of the state, with almost all of the state hit with tropical storm force winds. By 8 AM on Monday September 11th, Irma was passing by Gainesville as a tropical storm. By 2 PM the storm had passed out of Florida, heading north into Georgia.

4. $50 Billion in Recovery Costs (USA)
All 63 FL counties Declared
Federal Disaster Areas
1 in 3 Florida households applied for FEMA assistance
17 US Deaths
6.5 FL Million Evacuations
63% FL Households without power
All 63 counties were declared Federal Disaster Areas. Two-thirds of customers in Florida lost power at some point. And eventually, 1 in 3 households would apply for FEMA assistance.

5. UF Closed
9/8-9/12
Even though Gainesville did not bear the full brunt of the storm, UF closed September 8th through the 12th, with classes suspended through Wednesday the 13th. While the Libraries had some minor leaks and a power outage at an off-site facility, the collections were luckily untouched. We did however get to respond to several campus units that belatedly discovered that storing their permanent records on the floor in sub-grade rooms was not recommended protocol.
But even as we were assisting our own campus, there were questions being raised about what more we could do during the recovery and response phase for such a massive storm?

6. Formulating a Response
A week after Irma, I was contacted by Rebecca Elder, a fellow conservator and the coordinator for the National Heritage Responders, about initiating a project to contact heritage organizations that may have been impacted by Irma to collect data and to see if they needed any assistance from either the National Heritage Responders or the Heritage Emergency National Task Force. Rebecca had initiated a pilot phone-banking project with the University of Texas School of Information following Hurricane Harvey and was hoping to see the project expand to test its viability and usefulness.
The National Heritage Responders, which was formerly known as AIC-CERT, is a program run by the Foundation of the American Institute for Conservation to assist cultural institutions during emergencies. The number on the screen is their 24-hour hotline which will connect you directly to a conservator to provide advice on caring for collections during your emergency. And for major disasters, they send teams of conservators and heritage professionals to assist with recovery efforts, such as the Haiti Earthquake of 2010, North Dakota Floods of 2011, and Superstorm Sandy in 2012.
The Heritage Emergency National Task Force is joint effort of FEMA and the Smithsonian Institution to protect cultural heritage from the damaging effects of natural disasters and collection emergencies. Like NHR they help cultural organizations prepare for and respond to disasters both nationally and internationally through the coordination of 42 heritage organizations and Federal agencies. HENTF offers trainings and resources, as well as on the ground expertise as part of a coordinated response.

7. 📱 Who to contact? How to call them? What to ask?
Project Planning
Who to contact?
How to call them?
What to ask?
How to collect info/data? 📱
With the University of Texas sharing their experiences and documents, we weren’t starting from scratch, but there were still key questions that had to be tackled before we could begin our outreach.

8. Phone-Banking Basics Contact List Headsets or Phones Script/Survey
Volunteers
Google Voice Numbers
Snacks
Computers
Time
So like good information professionals we created a checklist and got to work. Some things were easy, it was the middle of the semester, so the computer training rooms in our Library weren’t booked solid. We could reserve the room and have student volunteers login to the computers there with their student IDs. Since we would have dedicated work-stations for the volunteers, we could borrow headsets from Library IT, that way student wouldn’t need to use their cellphones or bring their own laptops. That was a good thing, because it turns out that the room gets surprisingly spotty cell service.

9. What’s Florida Heritage?
NHR volunteers compiled list of museums, libraries, archives, historic sites, botanical and zoological gardens
The starting place of the project, and perhaps the biggest hurdle was figuring out what counts as Florida heritage, or if not defining heritage, at least compiling a list that could be used as the basis of the call sheet. This was a surprisingly big challenge because as far as the group of heritage professionals working on this project could tell, no central list of heritage organizations in Florida exists. NHR volunteers combed through existing public lists, membership rolls, web resources, and travel guides to come up with a list. Just developing the list took over a weeks worth of work, and no one involved thinks that it is actually comprehensive, even at this point.

10. But, but by September 25th, we had a draft list with contact information that eventually included almost 600 institutions in every part of the state. Some of these were actually multiple collections within a single institution, while the original list also included every county courthouse, which it was determined should not be called. And of course, as we learned during calling, having a list with general information phone numbers and actually getting in touch with someone who is willing and able to provide information on the status of an institution can take some time and some talking.

11. Organizations Per Region44 69 55
158 60 31 77 65
Organizations Per Region
The organizations on our list were coded by county and region. As might be expected, the number of institutions pretty closely tracks with the population of the region, with the Southeast and the Suncoast holding the majority of institutions.

12. We also attempted to categorize the institutions by type, using very broad breakdowns. I would not claim that this chart reflects the make-up of heritage organizations in the state, as the museum label is applied very, very, very broadly to places that have collections that are not primarily historic buildings, animals, or books or manuscripts. The “Nature” category broadly captures parks, wilderness areas, and botanical gardens.

13. With this list in mind, a natural partner for this project was UF’s Museum Studies Program. The graduate program had a pool of interested and engaged students to organize the phone-banking. Many of the students also had experience working with cultural organizations across the state, and at times, that experience helped make contacts with recalcitrant folks on the other end of the line.

14. We used Google Forms for our call scripts and data entry
We used Google Forms for our call scripts and data entry. The web forms were easy to develop and modify to meet our needs, and allowed multiple surveyors to enter data at any one time. Each call would get its own entry, with the script there on the page for the volunteer to read and work through.

15. We used Google Voice and Google Hang-Outs to place our calls
We used Google Voice and Google Hang-Outs to place our calls. We had set up a central Gmail account that we intended to use for communications and use the associated Google Voice number to place outgoing calls. Google Voice allows you to choose your location, so you are assigned a phone number with the appropriate area code. So all of the project calls could originate from a 352 area code, rather than from an assortment of cell phone numbers.
However, it turns out that Google doesn’t expect or allow five people to be calling on the same Google Voice number at once. Luckily, most of our volunteers had a Gmail account that they could link to a Google Voice number and place calls through that number. Our call script did include the central Google Voice number for the project for people to call back, and I checked those voic s daily to try to keep up on those messages.
Together, Google Voice and Google Forms made it very easy to bring in volunteers, give them 5 minutes of training and then have them off and calling with only minimal supervision.

16. September 20, 2017 September 26, 2017 October 3, 2017
Contacted by NHR and HENTF
Maria hits Puerto Rico
September 26, 2017
Nate forms in Southern Caribbean
Organized UF team lead by Museum Studies Graduate Students
October 3, 2017
Phone-banking begins
Organizing takes time, especially for a pilot project like this. Even with Texas’s templates it took us 2 weeks to get up and running. I would expect it to go more quickly during future attempts, but disasters are tricky things. Timing can be difficult. Even 3 to 4 weeks after Irma hit, several smaller organizations still had phone messages saying that they were closed because of damage during Irma. Were those just voic s that staff had forgotten to update or was there something more serious there?

17. By early October, we were ready to get to work and start making calls
By early October, we were ready to get to work and start making calls. We made 570 calls over 13 sessions, with 13 different volunteers, in trying to contact 511 organizations over a 5 week period. We first attempted to contact organizations on our list in Monroe, Collier, and Miami-Dade counties before working our way north through the state, tracking the path of the hurricane. By the end of month, I was mostly playing phone-tag with people who had left messages, although, I think that the final calls to Pensacola went out the first week of November.

18. Average length of call: 5 minutes 45 seconds
Project Findings
53.5 person hours of calls
Average length of call: 5 minutes 45 seconds
Average length of call recording damage: 10 minutes 10 seconds
Longest call: 32 minutes
And we made a lot of calls. The average length of each call was 5 minutes and 45 seconds. If there was damage that length stretched out to over ten minutes. The longest call was 32 minutes. And the most frustrating call was getting put on hold and being transferred around Miami-Dade College for 27 minutes before leaving a voic for someone who never called back. A lot of time was spent convincing volunteers and front desk staff to transfer us to someone, anyone in the organization that might be able to offer us answers and be a point of contact for NHR and HENTF staff.
And the calling was emotionally tough too. The best case scenario is that you reach someone who is fine, and never really needed us to go to the trouble of contacting them in the first place. The worse case scenario is that we actually reach the people who need help, people who are struggling with damage to their facilities or collections that they don’t know how to deal with. This is a project where the feel-good successes meant that we were sort of wasting everyone’s time.

19. When we made calls, we were able to speak with people 43% of the time
When we made calls, we were able to speak with people 43% of the time. 42% of the time we left a message. 12% of the time the phone just rang and rang. And 4% of the time someone hung up on us (often repeatedly). During our calling, we also found organizations that don’t have publicly listed phone numbers and places that had closed unrelated to Hurricane Irma.

20. We left a lot of messages, but we also talked to a lot of institutions
We left a lot of messages, but we also talked to a lot of institutions. You might be surprised to hear that some folks weren’t too happy to take a survey on the phone from someone saying that they are working with FEMA. We knew this was a possibility, but in the moment it was always a little jarring.

21. Of the 199 institutions that we eventually spoke with, 68 or 34% had sustained some sort of damage to their buildings. The damage types are not very surprising. A lot of roof damage and broken windows.

22. Of that self-reported damage, 65% was minor, 26% was major, and 1 organization reported that their facilities had been destroyed. We did not define major and minor damage for respondents, so obviously there is a gray area here for further investigation. The organization that reported their facilities as being destroyed was a nature preserve that had lost the boardwalk through the swamp that was their primary attraction. Obviously self-reporting damage can vary greatly depending on who is doing the reporting. For instance, even though some of UF’s library buildings had minor leaks, I didn’t report that as damage, knowing where the leaks are is just part of having historic buildings.

23. Only 18 institutions or 9% of contacted institutions reported any damage to their collection.

24. Two-thirds of collection damage was described as minor, with one-third reported as major. One garden reported their collections as destroyed.

25. As you can see, the damage by institution type actually tracks fairly closely with the overall sample. Museums and Libraries suffered damage rates of 9%. Nature Attractions and Zoos at 16%, and Historic Houses at 35%. I think this is fairly intuitive, as historic structures have a lot of vulnerabilities, but also that even minor damage to their building is seen as damage to the collections. There might also be a relationship to response rates, as historic houses that weren’t affiliated with the State Park system seemed to have a fairly high response rate in general; their volunteers were easier to get on the phone and more likely to answer questions than larger more bureaucratic institutions. We also see the elevated damage rate at the botanical gardens and zoos which are clearly susceptible to high winds in ways that traditional collections housed within buildings are not.

26. We can see from this heat map that the institutions from the full call list are distributed across the state with greater density around population centers..

27. Similarly the damaged institutions also are fairly evenly distributed across the state. Some areas like the Space Coast and Crystal River look like they may have escaped, although this may just be artifacts of response rates.

28. But if we look at the insurance claims across the state, we can see that the survey data actually lines up fairly well with the claims data.

29. % of Total Reported Damage 31% 8% 12.5% 17% 14% 2% 5% 11%SE SW CE C CW NE NC NW
Reported Damage 20 5 8 11 9 1 3 7
% of Total Reported Damage
31% 8%
12.5%
17%
14% 2% 5%
11%
Rate of Damage
38%
15%
28%
35%
30% 7%
18%
20%
Again, looking at the damage by region and the rate of damage, we can see that Irma impacted everyone, but particularly organizations in the southern and central regions. So, let’s see a show of hands, who here works for an institution that suffered some damage either to facilities or collections during Irma?

30. Florida Heritage Orgs Sustained Damage
1 out of 4
Florida Heritage Orgs Sustained Damage

31. 69% of orgs with damaged collections wanted an NHR call back🙋🙋
69% of orgs with damaged collections wanted an NHR call back ☏
61% with damaged facilities wanted a HENTF contact
When we spoke with them, 69% of organizations with damaged collections wanted a preservation specialist from the National Heritage Responders to call them to provide advice to aid them in the recovery. 61% of organizations were interested in hearing from a HENTF coordinator about possible FEMA funding that they may have been eligible for.

32. IT’S NEVER A GOOD TIME FOR A HURRICANE
Hurricane Maria
Mid-terms
Richard Spencer visit/campus shutdown
Planning and preparation make response faster, easier, and better.
As we look forward to the 2018 hurricane season, we know that there is never a good time for a hurricane to hit. University students may have lots of unstructured time on their hands, but October 2017 saw our students dealing with the fallout from Hurricane Maria on Puerto Rico, their mid-term exams, and closures and cancellations for 2 days to deal with a campus visit by white supremacist and Nazi sympathizer Richard Spencer. This upcoming year couldn’t be worse, right?

33. LESSONS LEARNED/FUTURE MOVES
Better Lists
Goldman and Tansey’s Repo Data Project, funded by SAA, seeks to id archival repositories in the USA. m
Tech Mis-Steps
Call center model means multiple Google Hangouts/Voice accounts needed. Can’t rely on volunteers to set accounts up.
Volunteers
Hard to maintain efforts. Seek to boost turnout in first 2 weeks when enthusiasm is highest.
Communication Keys
Use improved list to send out preliminary survey by . But list needs to include general addresses or be kept up-to-date.
Polish Script
Format script to provide earlier yes/no for damage to shorten calls. Rewrite text for voic to push org to call back or fill out web form.
Outreach
Let orgs know that outreach is coming and who we are. Most time on calls spent explaining project and building trust.
And what can we do next time to be more effective?
First, we can try to build better lists. We can use the list we have, with the updated contact information we collected to be better prepared and more efficient. The Society of American Archivists is also funding a project, led by Ben Goldman and Eira Tansey to identify and geolocate archival repositories in the US. They will be releasing a beta-version of their data at RBMS next month, but having been able to preview their list for Florida, I can say that 100 of the 244 repositories on their list were not on our call list for Irma.
Second, technology. Having run through Google Voice once, we can be prepared with multiple project accounts so that we don’t have to rely on volunteers to use the personal gmail accounts.
And speaking of volunteers, all of whom are great, it was clearly hard to maintain the effort and commitment to a project of this scale. We had a great turnout and made 248 calls on the first two days. It took us 11 more sessions to make the remaining 322 calls. Some of that is the nature of making return calls, but if we can boost turnout at the start of the project, we could get in touch with the impacted areas faster. Hopefully to provide those organizations in need with more time to prepare their claims for FEMA.
I would also consider using alternate forms of communication. I think that if we ed out a link to the survey form, we could get some data on respondents before we started calling. Even a yes/no on reporting damage and a phone number to call could facilitate the process. Of course, we would need up-to-date contacts for this to work
We could polish the script to help shorten the length of the average call. Spending 5 minutes to explain what we are doing just to hear that the organization was untouched or to leave a voic isn’t an effective use of our time. Just having a question right up at the top about whether the organization suffered damage could cut down on the time spent collecting good news about places that had weathered the storm safely. Though, more volunteers would also help resolve this bottle-neck.
Finally outreach. Hang-ups and people refusing to answer wasn’t a major problem, but working with state organizations like SFA, like FLA, like FAM, and the Humanities Council could get out the word to their members, so that when our calls or s do come in, our volunteers don’t have to spend as much time explaining themselves and the project.

34. WHERE DO WE GO? WHAT DO WE DO?
Can we use this data to target disaster response efforts? Was it helpful? Are there limits?
So what did we get out of this project? Obviously, I got to give this wonderful presentation, but was this helpful to anyone? Would we, should we do this again if the situation calls for it?
The most immediate benefit was to the 18 institutions that reported damage to their collections that National Heritage Responders were able to reach out to to offer advice. The cooperation with HENTF and FEMA was...less successful. While FEMA staff were able to use the data collected by the University of Texas’s efforts to contact institutions impacted by Harvey and assist them in navigating the Public Assistance process, the FEMA Liaisons in Region 4, which covers Florida, were not allowed to contact institutions from the list that UF generated. I don’t know the reasons for that, but it is what it is. Perhaps following future hurricanes that will change.
Was the data helpful? I would say that we don’t know. I don’t think that we can say that our data is statistically significant in terms of planning or preparing for the next storm. However, I think that there are some specific takeaways for future projects in future storm seasons:
First, spend the extra effort to contact historic homes and historic sites. It is worth reaching out until we can make contact, as these places appear more likely to have been damaged.
Second, trust and communicate with the neighbors. Always ask if the people know of any other local organizations that suffered damage. We made several additions to our list because people knew of other collections that had suffered worse than they had. The people who talked with us wanted to help out the other organizations in their communities that they knew had been hit by the storm.
Third, enforce clean data. Volunteers are great, but if you provide them with an option for “other” or a free text box, they will take it. And then I’m stuck cleaning up spreadsheets for weeks.
Finally, perhaps making a statewide effort truly a statewide effort. Gainesville escaped relatively unscathed in the scheme of things, but that may not always be the case. Perhaps a project like this should have multiple stakeholders across the state. Prepared to help out our fellow Floridians? Might it be easier for the community stakeholders to reach out to their local heritage organizations to check on them? Would they get a better response than calls from a random Gainesville phone number? I don’t know, but connected communities are resilient communities.
We know these storms are coming. We don’t know where, and we don’t know when, but everyone in this room will be impacted by a storm in the coming years. Preparations are important, so let me remind you all to update your disaster plans and phone trees if you didn’t already do so on May Day. But response is also a reality, and the more that we can do together, the more that we can ensure that our shared heritage survives.

35. Lourdes Santamaria-Wheeler
THANKS TO
Lauren O’Neill
Elizabeth Bouton
Lourdes Santamaria-Wheeler
Briley Rasmussen
Lori Foley
Jess Unger
Andrew Robb
Rebecca Elder
This was a big team effort to run the project and I want to thank everyone who helped make it function, but especially Lauren O’Neill and Liz Bouton, two great Museum Studies students who took the project and ran with it.

36. Any questions? You can find me at @fletcherdurant & fdurant@ufl.edu
I think that we are saving questions for the end, but when the time comes please feel free to ask away. Thanks.