Journal of Emergency Management

same access to information or ability to self-evacuate they face similar operational challenges related to
because their liberty is restricted. disaster planning, interorganizational ­collaboration,
inmate care, legal concerns, accountability, staffing,
The Government Accountability Office (GAO) and training. It follows then that cross-organizational
has identified hospital inpatients, residents in nurs- isomorphic learning from within the same industry is
ing homes, and the transportation-disadvantaged a pragmatic way to avoid repeating the past mistakes
as vulnerable groups.5 In its post-Katrina report to of others.
Congress, the GAO identified the transportation-dis-
advantaged as “the elderly, persons with disabilities, Eighth amendment issues
and people who, by choice or circumstances, do not The Eighth Amendment to the United States (US)
own or have access to a personal vehicle.”5 The GAO
stressed the responsibility, incumbent on all levels Constitution entitles inmates to freedom from cruel
of government, to effectively plan for and facilitate and unusual punishment. Although this prohibition
hurricane evacuations for disadvantaged populations. at one time meant freedom from cruel and unusual
Moreover, a lack of preparation on the part of nursing sentences, in 1976 the US Supreme Court (SCOTUS)
home facility administrators was highlighted. This also included the living conditions within correctional
is a salient point given that local, state, and federal facilities.15 This idea was further expanded in 1994
correctional facilities have a similar organizational when the SCOTUS declared that jail and prison
hierarchy as well as a custodial relationship with the administrators have a duty to take reasonable meas-
populations in their care. ures for the protection and safety of those in their
care.16 This clearly implies a responsibility for inmate
Organizational learning from past events impact- wellbeing during disasters.
ing similar institutions is a fundamental element in
creating better preparedness.25 The similarities found In 2016, the Bureau of Justice Statistics’ (BJS)
between disaster planning for nursing homes and cor- most recent study estimated the total population of
rectional institutions grow exponentially when com- adult inmates in the US to be in excess of 1.9 million.29
paring correctional facilities to each other because As a point of comparison, this is roughly the equiva-
they are so similar. Although the specifics of each lent to the population of Houston, Texas30 and is the
disaster are often unique,26 the common nature of largest incarceration rate in the world.31 With such a
hurricanes with a genesis in the Caribbean and large population to care for, it is therefore incumbent
South Atlantic provides many opportunities for iso- on politicians, emergency planners, and jail or prison
morphic learning to those entities exposed to their administrators to appropriately prepare correctional
hazards. Accordingly, this study draws on Toft and facilities for foreseeable natural and human-induced
Reynolds theory of isomorphic learning—“lessons hazards.14 In the context of hurricanes, this includes
from one setting can be relevant to settings that either a timely evacuation before landfall, or ensuring
are similar”28—which extends the earlier work on inmate safety and humane treatment when shelter-
institutional isomorphism by DiMaggio and Powell.27 ing in place. The two selected case-studies, Katrina
Isomorphism is used as a framework for compar- and Maria, demonstrate that adequate safety meas-
ing the OPP in Katrina with the MDC in Maria. ures are not being implemented. These examples
Specifically, cross-organizational isomorphism “which illustrate a failure on the part of the corrections sys-
applies where the organizations belong to different tem, emergency planners, politicians, and society, but
sets of owners, are managed and staffed by different it does not have to be so.13,14,32 Best practices and real-
people, but operate in the same industry.”28p72 The world examples of evacuation planning and facilita-
cases of the OPP and the MDC are both within the tion for transportation-disadvantaged groups already
US correctional facility industry and therefore can exist. This provides a base of knowledge and expe-
be considered similar in that they both provide the rience to inform correctional facility preparedness
same service—i­ncarceration of inmates. For example, for the disasters of the future.5,33-35 Accordingly, the

Special Issue on Puerto Rico

Journal of Emergency Management 99
Vol. 19, No. 8

research design herein adopts a case-study method Maria.5,13,14,17,32 The only scholarly works directly
to answer the question “what preparedness lessons focused on disaster related correctional facility evacu-
can be learned from comparing the experiences of ations in the US were legal papers written for law
correctional facilities during Hurricanes Katrina and reviews by law students remunerating the Eighth
Maria?” Amendment rights of inmates13,14,17,32 and a report
from the GAO regarding the evacuation of vulner-
METHODOLOGY able populations, which referenced inmates.5 The
American Civil Liberties Union (ACLU) did publish
The interpretivist paradigm, an inductive a large report based on its investigation of the OPP
approach, and a cross-case synthesis method were after Katrina; however, this paper was not sub-
chosen to enable a critical evaluation of the impacts of jected to peer review.38 One international publica-
Katrina and Maria on selected correctional facilities tion focused on prison disasters in New Zealand was
within the impacted areas.36,37 The primary aim of uncovered but because of significant political and
the cross-case synthesis approach is to maintain the procedural differences with the US it was considered
integrity of the entire case; then, to synthesize emer- not applicable.39 A Google search uncovered journalist
gent patterns across the selected cases.36 Determining and prisoner advocacy groups who reported on facility
the patterns of response and lessons learned from conditions and administrative actions in both Katrina
the OPP Katrina and the MDC Maria experiences and Maria using various media outlets. Government
requires an argumentative interpretation rather than publications from the US Bureau of Prisons (BOP),
quantitative analysis.36 The aforementioned research the GAO, and the BJS were used to verify demo-
question seeks to identify what happened and what graphic information, accountability relationships, and
can be done to enhance the preparedness of correc- the existence of evacuation or shelter-in-place docu-
tional facilities for disasters in the future which pro- ments related to disasters.5,21,29,40-42
vides the motivation and rationale for the selection of
a cross-case synthesis method. This study addresses the noted gap in the research
and literature regarding the preparedness of correc-
Several criteria informed the selection of the tional facilities for disaster using an emergency man-
OPP and MDC case studies. The criteria included agement (EM) lens. While the legal considerations of
population size and legal status, regular exposure humane treatment are important to prisoner rights,
to a natural hazard, adequate warning systems, and they add little to the real-world implementation of a
plentiful media accounts of their response. In order workable hurricane preparedness framework for cor-
to apply the lessons of isomorphism it was important rectional facilities.
that the selected cases be representative of potential
disaster situations.25,28 The examination of these two Analysis
correctional facilities is especially relevant given the The corpus of secondary qualitative data collated
frequency of hazards impacting similar institutions
throughout the US correctional system. via an electronic literature survey was subjected to
content analysis to determine the key themes related
The ProQuest, Wiley, Scopus, Science Direct, and to the evacuation and humane treatment of inmates
Google Scholar databases were used to establish a during times of disaster. Preliminary analysis identi-
corpus of qualitative data focused on the evacuation fied three types of literature, legal writings, media
and humane treatment of inmates. The key words accounts, and government publications which affords
utilized were: prison and jail evacuations; prison a degree of validity through triangulation.36 These
or jail disasters; vulnerable population evacuations; three corpora of literature were coded for both of the
and hurricane evacuation planning. This review of selected case studies. Media publications were the
the literature identified five scholarly works refer- primary source of data regarding what happened in
encing the conditions of the OPP during Katrina, these two correctional facilities. Moreover, the ACLU
and zero focused on the MDC in Puerto Rico during

Special Issue on Puerto Rico

100 Journal of Emergency Management
Vol. 19, No. 8

report was also a valuable source regarding the OPP Gusman, implemented a shelter-in-place order at the
experience.38 OPP.40 Moreover, guards, their families, and other
civilians were allowed to shelter in the jail further
Content analysis resulted in 15 codes. Code complicating an already challenging situation. Prior
clustering informed six categories that were further to landfall, the OPP turned off prisoner phones and
refined using thematic development resulting in televisions. This prevented inmates from accessing
three emergent themes.43,44 Table 1 shows this pro- outside information about the approaching storm or
gression from codes to themes. loved ones in its path. Perhaps urging or questions
from an inmate who knew the storm was approaching
These three themes establish the foundation would have spurred the Sheriff’s Office into action,
underpinning the posited recommendations for although an inmates ability to communicate with
improvements to disaster preparedness in correctional the higher reaches of the office is questionable. These
facilities. Note: the law-related literature offered occurrences demonstrate the intrinsic vulnerability of
recommendations for legal accountability reform; inmates when facing disaster.
however, the recommendations herein adopt a more
pragmatic, operational approach to the implementa- When the levees broke in New Orleans on August
tion of a workable hurricane preparedness framework 29, electricity, medicine, food, and water quickly dis-
for correctional facilities. appeared at the OPP while sanitation deteriorated
rapidly.40-42 Many guards either abandoned the jail
FINDINGS or could not get to work because of flood conditions.
Those guards and civilians who remained at the OPP
The findings drawn from the cross-case synthe- did not have enough food and water for themselves
sis of the qualitative corpus of data are presented let alone the inmates. Forced to remain inside the
herein in the form of two case study narratives. These jail with starving, thirsty, and increasingly desperate
accounts provide a synopsis of the OPP experience inmates, the guards and civilians observed firsthand
during Katrina, and the MDC experience during what conditions behind the bars were like.42 Inmates
Maria, which inform the posited recommendations. remained locked in their cells while flood waters rose
with only a few guards and other persons to provide
Katrina and the OPP, Louisiana, USA— the necessary care. Moreover, while cells filled with
August 23-31, 2005 water, generators failed because of lack of fuel and the
sewer system backed-up. Mixed-in with the returning
On August 25, 2005, Katrina made landfall in water were oil, gas, and human waste which flooded
Florida as a category 1 hurricane, intensifying to the cells where inmates were still confined. Unable
become a category 5 on August 28, dominating the Gulf to leave their cells or flush the toilets, inmates were
of Mexico and making landfall in Louisiana as a cat- forced to relieve themselves in the water then remain
egory 3 hurricane with winds of 120mph on August 29. confined with the waste. Interviews and photos which
1,833 people died; so, Katrina is rightly considered one emerged through the media and advocacy groups in
of the worst natural disasters ever to impact the US.45 the aftermath revealed that inmates were desperate
Much has been written about the survivors who did to escape these conditions so began breaking through
not or could not evacuate, particularly in New Orleans. walls and knocking down doors in an attempt to sim-
However, the experiences of almost 8,000 inmates who ply avoid the life-threatening conditions.38,40-42
were left behind in the OPP are often excluded from
the survivor narratives. It is important to note that Following their evacuation on the evening of
the majority of the OPP population at the time were Thursday September 1 by Louisiana Department of
pretrial detainees.13,14 Pretrial detainees are those who Corrections (DOC) officers from Angola, both inmates
have not been convicted and by law are presumed inno- and guards blamed Sheriff Gusman for the inhu-
cent while retaining all preconviction rights. mane conditions that ensued.38,41,42 The OPP had an

Despite a mandatory evacuation order for the
city of New Orleans, Orleans Parish Sheriff, Marlin

Special Issue on Puerto Rico

Journal of Emergency Management 101
Vol. 19, No. 8

Table 1. Theme development

Codes Categories Themes

Interagency coordination Planning and preparation action Predisaster emergency planning was
Intra-agency coordination Decisions made pre-onset piecemeal at best
Plan or planning documents
Hurricane prep actions

Shelter in place decision indication
Evacuation decision indication
Inmate demographic information

Hurricane warnings from government Advance notice of hurricane onset Few exogenous checks on either institu-
Hurricane warnings from media/others Accountability problems tion existed for ensuring sound disaster
mitigation, preparation, and response
Prison oversight
Hierarchy in the prison

Guard behavior after onset Actions of prison staff during response Inmate living conditions after d­ isaster
onset were inhumane in violation
Prison administration behavior/decision were inappropriate of ­established eighth amendment
p­ rotections
Exposure to hazards Living conditions were poor after onset
Access to life essentials

evacuation plan which was in place before Katrina been left in terrible conditions during and after hur-
made landfall; however, the Sheriff opted not to ricanes.13,14 A recent and visceral example of one such
evacuate and failed to adequately plan for the sus- situation can be seen by examining the response of
tainment of both inmates and the guards.40 Fuel, food, the MDC in Puerto Rico to Hurricane Maria.
water, and medicine were not stockpiled in any rea-
sonable amount and no public explanation as to why Maria and the MDC, Guaynabo, Puerto Rico—
has yet been given. Never before had the OPP, or New September 16-October 2, 2017
Orleans, experienced a storm like Katrina. Although
the Federal Emergency Management Agency (FEMA) On September 17, 2017 Hurricane Maria formed
led the Hurricane Pam Exercise—July 16-24, 2004— as a tropical storm in the Lesser Antilles, becoming
which simulated the effects on Louisiana of a slow- a category 5 hurricane by September 18, making
moving category 3 hurricane.46 Thus, it is surpris- landfall in Puerto Rico as a category 3 hurricane on
ing that politicians and administrators were wholly September 20.47 Puerto Rico, and by extension of the
unprepared to deal with Katrina’s landfall and after MDC, is frequently impacted by hurricanes due to its
effects. The GAO, among others, documented numer- location in the Caribbean. The MDC in Guaynabo is
ous lessons learned which influenced the passage of on the coast near to San Juan and, unlike the OPP, is
the Post-Katrina Emergency Management Reform not a local jail. It is a Federal Detention Center oper-
Act of 2006.5 It would seem then that the nation’s ated by the US Bureau of Prisons and is a part of the
correctional system and politicians had a what not to Department of Justice.48 As it is a federal facility, the
do guide for future hurricanes impacting correctional MDC must adhere to BOP emergency protocols and
facilities, but this has not been the case.17 Following policies which include evacuation procedures. Similar
Katrina, similar instances of cross-organizational iso- to the OPP during Katrina, at the time Maria made
morphic failures have occurred where inmates have landfall the MDC housed mostly persons awaiting
trial or sentencing.49

Special Issue on Puerto Rico

102 Journal of Emergency Management
Vol. 19, No. 8

In 2005, shortly before Hurricane Katrina, the the MDC were not considered a threat to the public,49
National Institute of Corrections (NIC), a component no plan was made to evacuate them prior to Maria
of the BOP, published its Guide for Preparing for and making landfall. The shelter-in-place order persisted
Responding to Prison Emergencies, which has not even after it became well known that Maria would
been updated in the 15 years since.50 As the MDC is a be the strongest hurricane to hit the island since
BOP facility, it should have been operating under this Hurricane George in 1998. Unlike the OPP and
guide, but it was not. Savilonis13 noted that the NIC Sheriff Gusman’s decision not to evacuate, none of the
checklist for evacuation and disaster preparedness database or internet searches yielded a formal source
is not broadly known and rarely used, even by those or publication, scholarly or otherwise, that revealed
who are aware of its existence. This indicates a failure which BOP authority made the decision to have the
in cross-organizational isomorphic learning and that MDC shelter-in-place.
few within the BOP view natural hazards prepared-
ness as a salient issue, even those in significantly Puerto Rico sustained extensive infrastructure
hazardous zones. damage which has been well documented.47,53,54 The
lack of power on the island contributed to the squalid
As the frequency and strength of hurricanes conditions experienced by the MDC inmates. After
have grown,51 evacuation planning for the inmates seven days of being locked in their cells with no power,
at the MDC in Guaynabo has not kept pace. Maria’s no working toilet, and very little water, they were
approach to the island as a category 5 storm was finally evacuated to a federal prison in Yazoo City,
well documented and warnings were given days in Mississippi on September 27.55 Upon arrival, they
advance as it eased to a category 3. In spite of the were given a bottle of water and a sandwich which
GAO’s documented lessons learned from Katrina and was the first food they had consumed in more than
the guide created by the NIC, the inmates held at 24 hours.49
the MDC were not evacuated until several days after
Maria made landfall.49,52 Although Puerto Rico is Cross-synthesis
located in a known hurricane hazard zone, the neces- There are several key similarities between the
sary food, water, medicine, and fuel caches at the cor-
rectional facility were inadequate. Hurricane season two case studies. Both the OPP and the MDC are
in the Caribbean occurs roughly the same time every exposed to hurricanes originating in the Caribbean
year and Maria had been monitored for several days. and South Atlantic which highlights the need for
Moreover, the experiences and lessons of the OPP both facilities to engage in similar planning and
during Katrina had been available to both the correc- preparation for these events. The number and types
tional industry and the public for over a decade. There of inmates were similar allowing some decision alter-
was a lengthy opportunity for cross-organizational natives, especially in the housing of pretrial detain-
isomorphic learning to occur. This raises the question ees.38 Ample decision time existed prior to landfall for
of why the facility was not evacuated in a timely man- both facilities. Unlike sudden onset disasters such as
ner or, at the very least, adequate supplies procured earthquakes, early hurricane warnings were availa-
to ensure humane conditions during a shelter-in-place ble through the National Hurricane Center. Published
order? accounts of inmate living conditions were also similar
providing clues about what operational actions may
Instead of being evacuated or adequately supplied be best suited for future disaster preparation and
for sheltering in place, the inmates at the MDC were response. Evacuation of both institutions did finally
forced to endure conditions and treatment similar to occur, but only after days of inmate lockdown in
those incarcerated at the OPP during Katrina.14,49 squalid conditions. Finally, the OPP and the MDC are
Akin to the OPP experience, conditions included cell representative of other institutions within the US cor-
flooding, water shortages, inability to flush the toilets, rectional industry which is charged with the contain-
and no electricity.52 Although most of the inmates at ment and care of inmates. Large county and city jails

Special Issue on Puerto Rico

Journal of Emergency Management 103
Vol. 19, No. 8

exist in every state56 and there are two other MDCs Maria case, was this preparedness guide not followed
plus an additional 109 other US federal correctional for what was clearly going to be a major hurricane?
facilities.57 Therefore, isomorphic learning across all
these institutions, within the same industry, should The level of neglect displayed by politicians and
be occurring. correctional facility administrators during Katrina
and Maria is not unique. The federal government is
Several differences also exist between the two considered a progressive leader in establishing best
cases. The level of government responsible for each practices for correctional facility conditions and treat-
differs. The OPP is a county jail while the MDC is a ment. However, politicians and administrators at all
federal level detention center. Available holding facili- levels have ignored the road to preparedness and
ties for potential evacuees is another difference which smoother collaboration laid out for them by the NIC.
may impact planning. Because the MDC is part of a Interestingly, the FY 2019 performance budget for the
larger network of federal institutions, collaboration BOP indicates that each correctional facility has an
between sister institutions is overseen by the BOP emergency plan developed in conjunction with local
while the OPP is a standalone operation. governments.23 Although, historically they have been
unreliable in the oversight and execution of these
DISCUSSION AND RECOMMENDATIONS plans.14

Content analysis of the findings, as seen through Recommendations
the lens of isomorphic learning, resulted in seven Seven recommendations for an improved cor-
posited recommendations for enhancing the prepar-
edness of correctional facilities: these represent the rectional facility disaster preparedness framework
theoretical contributions of this study. resulted from this study. These recommendations are
forwarded in the hope of preventing a repeat of condi-
Discussion tions inmates were subjected to during Katrina and
Society tends to marginalize inmates and correc- Maria as a nation is only a nation if it can protect all
its citizens: not just the law-abiding ones.
tional facility evacuation considerations have proven
to be no different.14 Mandatory evacuations ordered FEMA should assist the BOP to update preparedness
by officials in response to disasters should apply to and evacuation plans, procedures and practice. Inter-
jails and prisons as much as they do to hospitals vention from either FEMA, the GAO, or both is needed.
and nursing homes. Clear standards for federal Clearly, the BOP has not policed itself appropriately
prison evacuation and shelter-in-place scenarios were and FEMA is a disaster planning, coordination, and
developed by the NIC in 2005.50 This extensive guide support organization.58 Through its grant programs
covers many disaster planning issues which admin- and gatekeeper status for Presidential Declarations,59
istrators need to consider. For example, conducting FEMA wields considerable power in achieving compli-
an annual vulnerability assessment and risk analy- ance with mitigation and preparedness planning.60 At
sis, clear evacuation and relocation plans including the federal level, this could include revisions to the
alternative sites with transportation agreements, NIC checklist which would stimulate the creation of
updating or creating mutual-aid agreements with Mutual Aid Agreements with agencies external to the
other organizations to sustain logistics and supply, prison but integral to its disaster plan.
developing adequate response plans for a shelter-in-
place when an evacuation is not possible, planning Officers of the GAO should evaluate the emergency and
for an adequate cache of supplies and equipment, continuity planning and broader preparedness of the
and many more. Virtually, all the issues encountered BOP. As a working agency for Congress, the GAO’s
by correctional facilities impacted by hurricanes and stated mission is to “support Congress in meeting its
other disasters since Katrina are addressed in this constitutional responsibilities, and help improve the
one planning document. Why then, especially in the

Special Issue on Puerto Rico

104 Journal of Emergency Management
Vol. 19, No. 8

performance and ensure the accountability of the fed- planning efficacy should come from state level emer-
eral government.”61 Although not intended as a watch- gency management agencies.
dog for correctional facilities, the GAO does provide
oversight of the DOJ and, by extension, the BOP. Jails and prisons should be included on vulnerability
Holding the BOP accountable to its own disaster poli- maps. A Social Vulnerability Index, known as SoVI, is
cies, and to Congress, would be a step-change. often used during mitigation planning. It is typically
generated via Geographical Information Software
A recent example of the BOP lacking disaster using Census Bureau data, and does not routinely
preparedness was an electrical outage on January list jails as a vulnerable hotspot.64 This is another key
27, 2019 at the Federal MDC—Brooklyn, New York. omission, given the concentrated population typified
Similar to the MDC in Puerto Rico, Brooklyn inmates by the characteristics of social vulnerability incarcer-
went without light and heat for a week.22 While tour- ated within correctional facilities.
ing the Brooklyn MDC following the power outage,
New York City Council member Jumaane Williams Disaster planning at the state, county, and municipal
said, “There seemed to be no emergency plan, and level should also include jail administrators in emer-
there seemed to be no plan to create a plan, just a gency exercises. Everything from tabletop drills to full-
matter of waiting.”22 This demonstrates that emer- scale multi-agency mobilizations should include a jail
gency preparedness is still lacking at correctional representative of sufficient authority to make collabora-
facilities across the US despite the development of tive decisions. While these exercises cannot anticipate
the NIC checklist in 2005. As a result of the Brooklyn every variable, the planning process itself does facilitate
incident and similar reports from other facilities, relationship building between organizational stake-
in March 2019, US Senators Tammy Duckworth holders which is critical during disaster response.65,66
and Margaret Hassan, a ranking member of the
Subcommittee on Federal Spending Oversight and Multi-agency planning should include response con-
Emergency Management, wrote a letter to the GAO tingencies for shelter-in-place inmates. These contin-
Comptroller General, Gene Dodaro. This letter spe- gencies should include, at a minimum, how to get
cifically requested an official inquiry into the BOP’s staff to work or keep them there, sustenance of supply
preparation for and response to both natural and needs (food, water, medicine), sanitation and waste
human-induced disasters.23 management, and power generators with the means
to keep them running.67 The Department of Homeland
FEMA could tie funding to adequate planning and Security (DHS) guidance for residential emergency
preparation for state and local correctional facilities, sustenance is 72 hours.68 Correctional facilities should
as it currently does with municipal and state mitiga- be held to the same standard.
tion plans.62 Mitigation grants for prison infrastruc-
ture upgrades could be initiated. However, a state CONCLUSIONS
would only receive grant money once it demonstrated
that its prisons and jails had complied with emer- This cross-case synthesis yielded seven recom-
gency and continuity planning, training exercises, and mendations across all levels of government intended
validation requirements. to prevent, or at least mitigate, the all too frequent
inhumane conditions for inmates caused by disasters.
County and city jails should be monitored for emer- Three are recommended accountability and support
gency plan preparedness by the state as part of county measures for the federal BOP and state correctional
and municipal EM planning. City or county mitiga- facilities achieved through greater engagement with
tion plans rarely, if ever, include local jails.63 This is a FEMA and the GAO. While, the other four outline
considerable omission given the large vulnerable pop- recommendations for improvements in EM planning
ulation who are in custody there. Oversight of local at county and municipal levels.

Special Issue on Puerto Rico

Journal of Emergency Management 105
Vol. 19, No. 8

The Eighth Amendment guarantees inmates free- improve disaster preparedness on the part of US
dom from cruel and unusual punishment. The condi- correctional facilities. Furthermore, this approach
tions endured at both the OPP in New Orleans and could be expanded to the comparative analysis of
the MDC in Puerto Rico were inhumane, unnecessary, international facilities, of which little scholarly
and preventable. Consequently, inmate suffering is work has been done. Benefits from a qualitative
perpetuated although a clear path to reducing it, in survey of all correctional institutions within a given
the form of the NIC checklist, exists. Looking ahead, geographical location, eg, a single state or county,
the findings of this study demonstrate a clear need could yield insight into preparedness behaviors of
for all levels of government, and its EM components, multiple facilities overseen by a central authority. By
to learn from past mistakes (isomorphism) regard- extension, a survey of all federal institutions could
ing correctional facility response to natural hazards. also yield informative findings. The theoretical and
Learning from disasters such as Katrina and Maria practical gains from this and subsequent research
will engender greater preparedness and help to could make significant contributions to the social
ensure the wellbeing of the inherently vulnerable vulnerability, prison management, and EM planning
inmates for whom they have a duty of care. literatures.

Limitations of the study Carl Dement, BA (Hons), MA, School of Criminal Justice, University of
A limitation of this study is acknowledged in Central Oklahoma, Edmond, Oklahoma.

the comparison of disaster planning and response Tony McAleavy, BA (Hons), MSc, PhD, Fire & Emergency Management
between a county jail (the OPP) and a federal prison Program, Oklahoma State University, Stillwater, Oklahoma.
(the MDC). However, this limitation is mitigated
by the similarities of both facilities in inmate legal REFERENCES
status, duty of care, population size, and disaster 1. Lindell MK, Prater CS: A hurricane evacuation management
exposure. Additionally, both institutions, regardless decision support system (EMDSS). Nat Hazards. 2006; 40(3): 627-
of government level, coordinated disaster preparation 634. DOI: 10.1007/s11069-006-9013-1.
and response with the officials, and other stakehold- 2. Prater CS, Lindell MK: Politics of hazard reduction. Nat Hazards
ers, in their locale. Therefore, the recommendations Rev. 2000; 1: 73-82.
posited by this study are transferable to other cor- 3. Waugh WL: Shelter From the Storm: Repairing the National
rectional facilities.34 Emergency Managment System After Hurricane Katrina. Thousand
Oaks, CA: Sage; 2006.
This cross-case synthesis study was limited to two 4. Fernandez M, Fausset R: Houston mayor’s no-win dilemma:
correctional facilities in the US system rather than a Whether to tell residents to stay or go. The New York Times.
larger number of case studies. However, the two cases 2017:11.
presented here occurred during large, highly publi- 5. Bascetta C: Disaster Preparedness: Preliminare Observations
cized disaster events which contributed to greater on the Evacuation of Vulnerable Populations Due to Hurricanes
validity because of the triangulation available from and Other Disasters. Washington, DC: United States Government
multiple sources of data regarding them.36 As noted Accountability Office; 2006.
earlier, the situations which arose in the OPP and the 6. Cutter SL, Boruff BJ, Shirley WL: Social vulnerability to environ-
MDC are not outliers in the response of correctional mental hazards. Soc Sci Q. 2003; 84(2): 242-261.
facilities to disaster. 7. Cutter SL: The vulnerability of science and the science of vulner-
ability. Ann Assoc Am Geogr. 2003; 93(1): 1-12. DOI: 10.1007/978-
Future research 90-481-2350-6.
The cross-case synthesis method is a valu- 8. Finch C, Emrich CT, Cutter SL: Disaster disparities and differen-
tial recovery in New Orleans. Popul Environ. 2010; 31(4): 179-202.
able tool for pattern recognition across two or more DOI: 10.1007/s11111-009-0099-8.
cases.36 Future studies involving a broader range 9. Comfort L, Wisner B, Cutter S, et al.: Reframing disaster policy:
of cases could identify further opportunities to The global evolution of vulnerable communities. Environ Hazards.
1999; 1(1): 39-44. DOI: 10.3763/ehaz.1999.0105.
10. Fothergill A, Peek LA: Poverty and disasters in the United
States: A review of recent sociological findings. Nat Hazards. 2004;
32: 89-110. DOI: 10.1023/B:NHAZ.0000026792.76181.d9.
11. Fussell E: The long term recovery of New Orleans’ population
after Hurricane Katrina. Am Behav Sci. 2015; 59(10): 1231-1245.
DOI: 10.1177/0002764215591181.

Special Issue on Puerto Rico

106 Journal of Emergency Management
Vol. 19, No. 8

12. Hoffman S: Preparing for disaster: Protecting the most vulner- 34. Gray BH, Hebert K: Hospitals in hurricane Katrina: Challenges
able in emergencies. UC Davis Law Rev. 2008; 42: 1491-1547. facing custodial institutions in a disaster. J Health Care Poor
13. Savilonis M: Prisons and disasters. 2013. Available at http://hdl. Underserved. 2007; 18(2): 283-298. DOI: 10.1353/hpu.2007.0031.
handle.net/2047/d20004879. Accessed February 1, 2020. 35. Glick R, Bish DR, Agca E: Optimization-based decision support
14. Omorogieva W: Prison preparedness and legal obligations to to assist in logistics planning for hospital evacuations. J Emerg
protect prisoners during natural disasters. 2018. Available at http:// Manag. 2013; 11(4): 261-270.
www.columbiaclimatelaw.com. Accessed February 1, 2020. 36. Yin R: Case Study Research and Applications: Design and
15. Estelle v. Gamble, 429 US 97 (1976). Methods. 6th ed. Thousand Oaks, CA: Sage; 2018.
16. Farmer v. Brennan, 511 US 825 (1995). 37. Lewis-Beck M, Bryman A, Liao T, eds.: The Sage Encyclopedia of
17. Vumback M: Evacuating the incarcerated: The intricacies of Social Science Research Methods. Thousand Oaks, CA: Sage; 2004.
keeping prisoner’s Eighth Amendment rights intact during natural 38. Balaban E, Jawetz T: Abandoned and Abused: Complete Report.
disasters. Law J Soc Justice. 2019; 11: 1-17. 2006. Available at https://www.aclu.org/report/abandoned-abused-
18. US Department of Justice, Bureau of Prisons: Organization complete-report. Accessed February 1, 2020.
and Functions Manual. 2020. Available at https://www.justice.gov/ 39. Le Dé L, Gaillard JC: Disaster risk reduction and emergency
jm/organization-and-functions-manual-8-bureau-prisons. Accessed management in prison: A scoping study from New Zealand. J
February 1, 2020. Contingencies Cris Manag. 2017; 25(4): 376-381. DOI: 10.1111/1468-
19. Blaikie P, Cannon T, Davis I, et al.: At Risk: Natural Hazards, 5973.12165.
People’s Vulnerability, and Disasters. Abingdon, United Kingdom: 40. Williams B: Reflections on Katrina’s first year: The story of
Routledge; 1994. chaos and continuing abuse in one of America’s worst justice
20. Morrow BH: Identifying and mapping community vulnerability. systems. Prison Legal News. 2007. Available at https://www.prison-
Disasters. 1999; 23(1): 1-18. legalnews.org/news/2007/apr/15/reflections-on-katrinas-first-year-
21. Inmate Statistics: Race; 2018. Availble at https://www.bop.gov/ the-story-of-chaos-and-continuing-abuse-in-one-of-americas-worst-
about/statistics/statistics_inmate_race.jsp. Accessed February 1, justice-systems/. Accessed February 1, 2020.
2020. 41. Webster R: New Orleans prison’s future in question. New
22. Nicholas J: Just leave them to die. The crisis at Brooklyn’s Orleans CityBusiness. 2005. Available at https://neworleanscity-
federal jail reveals how jails and prisons are not prepared for a business.com/. Accessed February 1, 2020.
disaster. The Appeal. 2019. Available at https://theappeal.org/mdc- 42. Webster R. Orleans Parish Criminal Sheriff’s Office employ-
brooklyn-jail-prison-emergency-preparedness/. Accessed February ees describe chaos in prison following hurricane. New Orleans
1, 2020. CityBusiness. https://neworleanscitybusiness.com/. Published March
23. Duckworth T, Hassan M: GAO BOP final letter. 2019. Available 27, 2006. Accessed February 1, 2020.
at https://www.duckworth.senate.gov/imo/media/doc/03.29.19-GAO 43. Harding J: Qualitative Data Analysis: From Start to Finish.
Bureau of Prisons Final Letter.pdf. Accessed February 1, 2020. Thousand Oaks, CA: Sage; 2013.
24. Ferris T, Moreno-centeno E, El-sherif M, et al.: Studying the 44. Erlingsson C, Brysiewicz P: A hands-on guide to doing content
usage of social media and mobile technology during extreme events analysis. African J Emerg Med. 2017; 7(3): 93-99. DOI: 10.1016/j.
and their implications for evacuation decisions: A case study of afjem.2017.08.001.
Hurricane Sandy. Int J Mass Emerg Disasters. 2016; 34(2): 204-230. 45. Waugh WL: Katrina and the governors. Public Organ Rev. 2009;
25. Buchanan DA: Reflections: Good practice, not rocket science— 9(4): 343-351. DOI: 10.1007/s11115-009-0092-9.
understanding failures to change after extreme events. J Chang 46. FEMA: Hurricane Pam Exercise Concludes. 2004. Available
Manag. 2011; 11(3): 273-288. DOI: 10.1080/14697017.2011.597571. at https://www.fema.gov/news-release/2004/07/23/hurricane-pam-
26. Quarantelli EL: Converting disaster scholarship into effective exercise-concludes. Accessed February 1, 2020.
disaster planning and managing: Possibilities and limitations. Int 47. Pasch R, Penny A, Berg R: National Hurricane Center Tropical
J Mass Emerg Disasters. 1993; 11(1): 15-39. Cyclone Report: Hurricane Maria; 2018. Available at https://www.nhc.
27. DiMaggio PJ, Powell WW: The iron cage revisited: Institutional noaa.gov/data/tcr/AL152017_Maria.pdf. Accessed February 1, 2020.
isomorphism and collective rationality in organizational fields. Am 48. Southeast Region Locations: Federal Bureau of Prisons. 2019.
Sociol Rev. 1983; 48(2): 147-160. Available at https://www.bop.gov/locations/map.jsp?region=SER.
28. Toft B, Reynolds S: Learning from Disasters: A Management Accessed October 22, 2019.
Approach. 3rd ed. London, UK: Perpetuity Press; 2005. DOI: 49. Chrastil N: A Puerto Rican federal inmate’s horrifying account
10.1007/978-1-349-27902-9. of what the prison did after Hurricane Maria. ThinkProgress. 2018.
29. Kaeble D, Cowhig M: Correctional Populations in the Unites Available at https://thinkprogress.org/puerto-rico-federal-deten
States, 2016; 2018. Available at https://www.bjs.gov/content/pub/pdf/ tion-center-allegedly-abused-inmates-hurricane-maria-494c2a8
cpus16.pdf. Accessed February 1, 2020. 306e8/. Accessed February 1, 2020.
30. US Department of Commerce, US Census Bureau: QuickFacts 50. Schwartz J, Barry C: A Guide to Preparing for and Responding
Houston City, Texas. 2020. https://www.census.gov/quickfacts/hou to Prison Emergencies. 2005. Available at http://static.nicic.gov/
stoncitytexas. Accessed February 1, 2020. Library/020293.pdf. Accessed February 1, 2020.
31. Walmsley R: World Prison Population List; 2013. Available 51. Chandra A, Moen S, Sellers C: What Role Does the Private Sector
at https://www.prisonstudies.org/sites/default/files/resources/down- Have in Supporting Disaster Recovery, and What Challenges Does
loads/wppl_10.pdf. Accessed February 1, 2020. It Face in Doing So? Santa Monica, CA: Rand Corp; 2016. DOI:
32. Robbins IP: Lessons from Hurricane Katrina: Prison emergency 10.7249/pe187.
preparedness as a constitutional imperative. Univ Michigan J Law 52. Martinez Y, Flagg A, Caballero A: Prisons and the del-
Reform. 2008; 42(1): 199-206. uge: For prisoners in a hurricane, there’s nowhere to run. The
33. Kiefer JJ, Montjoy RS: Incrementalism before the storm: Marshall Project. 2017. Available at https://www.themarshallpro
Network performance for the evacuation of New Orleans. Public ject.org/2017/10/20/prisons-and-the-deluge. Accessed February 1,
Adm Rev. 2006; 66(1): 122-130. 2020.

Special Issue on Puerto Rico

Journal of Emergency Management 107
Vol. 19, No. 8

53. Response and Recovery after Maria: Lessons for Disaster Law 62. Federal Emergency Management Agency: The Hazard
and Policy. UC Berkeley. Available at https://escholarship.org/uc/ Mitigation Assistance Grant Programs; 2015. Available at https://
item/936195d5. Accessed February 1, 2020. www.fema.gov/media-library-data/1441133724295-0933f57e7ad
54. National Hurricane Center: Costliest US Tropical Cyclones 4618d89debd1ddc6562d3/FEMA_HMA_Grants_4pg_2015_508.
Tables Updated.; 2018. Available at https://www.nhc.noaa.gov/news/ pdf. Accessed February 1, 2020.
UpdatedCostliest.pdf. Accessed February 1, 2020. 63. Masterson J, Peacock WG, Van Zandt S, et al.: Planning for
55. Gates J: State is housing 1,200 federal prisoners from Puerto Community Resilience: A Handbook for Reducing Vulnerability to
Rico. Mississippi Clarion Ledger. 2017. Available at https://www. Disasters. Washington, DC: Island Press; 2014.
clarionledger.com/story/news/2017/10/18/state-housing-1-200-fed 64. Cutter SL, Mitchell JT, Scott MS: Revealing the vulnerability of
eral-prisoners-puerto-rico/777228001/. Accessed February 1, 2020. people and places: A case study of Georgetown county, South caro-
56. Zeng Z: Jail Inmates in 2018. 2020. Available at https://www.bjs. lina. Hazards, Vulnerability Environ Justice. 2012; 90(4): 83-114.
gov/content/pub/pdf/ji18.pdf. Accessed February 1, 2020. DOI: 10.4324/9781849771542.
57. Bureau of Prisons: Federal Bureau of Prisons—Our Locations. 65. McConnell A, Drennan L: Mission impossible? Planning and
2020. Available at https://www.bop.gov/locations/list.jsp. Accessed preparing for crisis. J Contingencies Cris Manag. 2006; 14(2): 59-70.
February 1, 2020. DOI: 10.1111/j.1468-5973.2006.00482.x.
58. Sylves R: Disaster Policy and Politics: Emergency Managment 66. Clarke L: Some functions of planning. In: Mission Improbable:
and Homeland Security. 3rd ed. Thousand Oaks, CA: Sage; 2020. Using Fantasy Documents to Tame Disaster. Chicago, IL: University
59. Barzilai-Nahon K: Toward a theory of network gatekeeping: of Chicago Press; 1999: 1-15.
A framework for exploring information control. J Am Soc Inf Sci 67. Johnson M: How to prepare for emergencies in your correc-
Technol. 2008; 59(9): 1493-1513. DOI: 10.1002/asi. tional facility ahead of time. CorrectionsOne. 2018. Avaailable
60. Berke P, Smith G, Lyles W: Planning for resiliency: Evaluation at https://www.correctionsone.com/products/communications/emer
of state hazard mitigation plans under the Disaster Mitigation gency-response/articles/how-to-prepare-for-emergencies-in-your-cor
Act. Nat Hazards Rev. 2012; 13(2): 139-149. DOI: 10.1061/(ASCE) rectional-facility-ahead-of-time-xosiXTUrk6gNLwRJ/. Accessed
NH.1527-6996.0000063. February 1, 2020.
61. GAO: GAO at a glance. 2019. Available at https://www.gao.gov/pdfs/ 68. US Department of Homeland Security: Build a Kit. 2020.
about/gao_at_a_glance_2019_english.pdf. Accessed February 1, 2020. Available at https://www.ready.gov/kit. Accessed February 1, 2020.

Special Issue on Puerto Rico

108 Journal of Emergency Management
Vol. 19, No. 8

JEM Understanding recovery and resilience from natural
disasters in hospitality organizations

Kristin A. Horan, PhD
Blake Scott, MPH
Ahlam Farzan, MD

Marian L. Ortíz-Aponte, MS
Alejandra Rivera-García, MS

Jennifer Marshall, PhD
Anthony J. Masys, PhD

Mindy Shoss, PhD
Adriana Campos, MPH
Lida Orta-Anés, PhD

ABSTRACT INTRODUCTION

Extreme weather events, such as Hurricane María, Hurricane María was a Category Five hurricane
shed light on the importance of understanding the fac- that made landfall in Puerto Rico in September 2017.
tors that promote resilience, defined as bouncing back The storm has been noted for its intense immediate
after adverse events. The current study took a quali- damage and mortality rate1 and prolonged a­ ftermath–
tative focus group approach toward understanding as evidenced in a recent article discussing sustained
resilience in employees of the hospitality industry after out-migration due to long-term employment impact.2
Hurricane María. The hospitality industry plays an Public health scholars note that the extraordinary
important role in disaster responses due to its role in impact of Hurricane María on the island can be
supporting local and national economies, job stability explained by geographical factors, pre-existing finan-
for vulnerable employees, and in supporting response cial and structural vulnerabilities, and the delayed
efforts, eg, housing aid workers. Through a series of in- response in aid.3 The hurricane impact was particu-
depth focus groups, employees from San Juan, Puerto larly challenging due to the embeddedness of tour-
Rico and surrounding areas shared their Hurricane ism activities in the island’s economy and the global
María experiences in terms of themselves as indi- downturn in demand for travel to this region follow-
viduals, their workplaces, and their community. Major ing the extreme weather events of 2017.4 As a result,
themes focused on the importance of the preparedness climate-sensitive industries depending on tourist
phase of the disaster response timeline, measuring demands, such as the hospitality industry, were espe-
impact and response on multiple levels in interde- cially vulnerable to this catastrophic immediate and
pendent systems, creating and promoting awareness long-term impact.
of resources, and acknowledging both struggle and
strength when conceptualizing transformative poten- At the same time, hospitality organizations are
tial. These findings can inform actionable strategies noted for their interdependency with their host com-
for individuals, organizations, and communities. munity, which plays an important role in crisis man-
agement for hospitality organizations by promoting
Key words: resilience, weather events, hospitality coordinated rather than disparate response efforts.5
industry, Hurricane María Brown et al.6 describe the need to foster resilience

DOI:10.5055/jem.0643 Special Issue on Puerto Rico
Journal of Emergency Management
Vol. 19, No. 8 109

among hospitality organizations, citing reasons such to the functioning or survival of the organization),
as the physical, economic, and social impacts from dis- although the two concepts are inherently intertwined
asters, their support of the local and national economy, as individual resilience contributes to the resilience
high rates of employment of vulnerable populations, capacity of systems.12
and their pivotal role in disaster response efforts. Yet,
scholars have noted that conceptual and operational Integrative reviews of empirical resilience litera-
features of existing resilience literatures may not ture at both the individual13 and organizational level14
adequately account for individual, group, and commu- suggest that resilience can be thought of as a capabil-
nity interdependencies and typically do not result in ity or trait, a process, and an outcome. That is, enti-
actionable information for these stakeholders.7 ties, eg, individuals, organizations, and communities,
can be classified as resilient based on existing char-
Thus, the aim of this paper is to use qualitative acteristics or capabilities that are protective against
focus groups with hospitality employees to identify adverse events–values, tolerance for ambiguity, per-
actionable information to increase resilience among sonnel, or infrastructure–actions taken in response to
employees of hospitality organizations toward disas- adverse event–communication, ­decision-making, and
ters. In particular, we utilize qualitative focus groups goverance processes–or achieving financial or opera-
to understand the characteristics of hospitality work- tional performance outcomes following an adverse
ers, hospitality organizations, and their host com- event (refer to Gibson and Tarrant15 for a conceptual
munity that contribute to resilience post-Hurricane model that integrates these schools of thought). In
María. Our research is in line with a growing scholarly light of these differing conceptualizations and in line
and practical interest in the phenomenon of “bouncing with Gibson and Tarrant,15 we also take a broad con-
back” after adversity or resilience.8 Vulnerability to ceptualization to resilience, viewing it as the capabili-
adverse events and resilience are inherently intercon- ties and processes of individuals and organizing enti-
nected concepts,9 as research can be used to under- ties that are aimed at limiting disruption following
stand and promote resilience among populations with negative events. It is important to note that whether
differential patterns of exposure to shocks and stress- conceptualizing resilience as a capacity, process, or
ors. Thus, we anticipate that our findings will not outcome of organizations, multiple studies support
only advance the scholarly literature on resilience but the benefits of resilience, as resilient organizations
also provide concrete recommendations for hospitality reap positive outcomes such as a general sense of
workers, hospitality organizations, and communities individual and organizational wellness during or
to promote resilience in the future. after disaster or disruptive events.16

Resilience from extreme weather events Although the definition of resilience can cover
The literature on resilience focuses on recovery many challenging conditions,11 such as scandals and
discrete errors in the case of organizations, of par-
and positive functioning in the face of adverse events,10 ticular interest to disaster management literature
both at the individual and system level. Broadly, resil- is positive functioning of an organization following a
ience is defined as “the maintenance of positive func- natural disaster such as an extreme weather event.
tioning under challenging conditions.”11 In this paper, Linnenluecke et al.17 discuss extreme weather events
we focus on the interplay of multiple levels of an in the context of organizational resilience, positing
ecological system, which we conceptualize broadly as that a greater understanding of organizational resil-
capturing resilience in workers, workplaces, and the ience can help workplaces avoid threats to survival
community in which both entities exist. Specifically, and the rippling effects to the industry and economy
we focus on the role of individual, organizational, and as a whole. They argue for the importance of a focus
community factors in fostering individual resilience. on extreme weather events in the disaster manage-
We assess individual resilience within the context ment and organizational resilience literatures for
of their organizations and communities (as opposed several reasons. First, extreme weather events modify

Special Issue on Puerto Rico

110 Journal of Emergency Management
Vol. 19, No. 8

patterns of demand for several industries crucial to for several reasons. First, the hospitality industry
the functioning of the local and national economy. In features climate-sensitive inputs and demand-based
industries with “climate-sensitive inputs and loca- recovery patterns discussed by Linnenlueke et al.,17
tions,”17 organizations are more vulnerable to long- meaning that global demand for the services offered
term financial disruption on top of immediate physi- by the hospitality industry respond to changes in
cal property damage. Second, Linnenluecke et al.17 climate and that changes in demand following an
argue it is imperative to understand organizational extreme weather event can delay recovery of the
resilience in response to extreme weather events industry.4 Although hotel room demand can increase
because such events are and will likely continue to in the short term to house disaster recovery workers
occur at an increased frequency. Scientists argue and insurance adjusters,21 the long-term impact of
human-induced climate change makes more frequent weather events is severe for hospitality organiza-
and intense weather events more likely.18,19 It is inter- tions and the local economy. For example, a report
esting to note that resilience can occur in response by Tourism Economics prepared for Visit Florida in
to both a large-scale event, described as a “shock,” or 2018 found declines in multiple indicators of hotel
cumulative exposure to repeated adverse events.11 demand, tourism performance, and the state economy
More frequent and intense weather events or natural in the year following Hurricane Irma. Although some
disasters can be thought of as both a large-scale shock markets do recover to pre-event levels, others strug-
and a cumulative strain.20 gle to recover quickly. For example, the Florida Keys
had yet to recover prehurricane levels of demand and
Despite what is known about resilience, scholars revenue-per-available room as of 6 months post-Hur-
note several “unknowns” that should be addressed in ricane Irma.22 This indicates that it may be especially
future resilience research.13 Difficulties in advancing important to understand resilience to weather events
the resilience literature involve both conceptual and or other natural disasters among industries with
operational limitations.7 First, on a conceptual level, climate-sensitive inputs, given that the interaction
identifying actionable steps to cultivate resilience is of physical property damage and patterns in demand
difficult. Although it is relatively easy to quantify hav- may delay recovery.
ing “bounced back,” there is less agreement on what
promotes bouncing back. Second, on an operational Second, the hospitality industry is also noted for
level, resilience occurs in interdependent systems, cumulative strain from repeated exposure to weather
with high levels of interaction among individuals, the events, a consequence of the increase in the frequency
organization they belong to, and the community in of extreme weather events.17 For example, a profile
which the organization exists. Most resilience studies of Gran Meliá Coco, a hotel in Puerto Rico, discusses
do not adequately account for complexities and inter- immediate damage from Hurricane María in 2017,
dependencies at all levels.7 their road to recovery and reopening on December
20, 2019, and the impact on travel plans and cancela-
To address these gaps, we focused on the hospi- tions of large meetings caused by the earthquakes
tality industry. The hospitality industry not only is that started on December 28, 2019 and progressed
central to recovery from hurricane disasters but also into January 2020.23 Although previous experience
provides an important setting by which to examine in a natural disaster can be useful for hotel leaders,24
the interrelatedness of individual, workplace, and cumulative exposure to repeated weather events may
community factors in promoting resilience. The fol- deplete the physical, human, social, and economic
lowing sections describe the hospitality context and capital that would ready an organization for future
introduce the questions that guided this research. adverse events.6

The hospitality context In addition to these factors, the hospitality
It is especially important to understand the resil- industry and individual and community function-
ing following a disaster event are also inherently
ience to weather events in the hospitality industry6

Special Issue on Puerto Rico

Journal of Emergency Management 111
Vol. 19, No. 8

interconnected. For example, the hospitality industry employee and organizational resilience, and hospi-
is often leaned upon during times of crisis. The hos- tality literatures.
pitality industry plays a pivotal role in disaster pre-
paredness and response,25 particularly in providing In keeping with this need, this study aimed to
shelter for response teams. It is imperative to support identify factors that contribute to the resilience of
the positive functioning of the hospitality industry individuals employed in hospitality following an
during times of crises, so that they are able to provide extreme weather event. We used qualitative methods
essential services for those delivering crucial aid to to assess individual, organizational, and community
civilians. factors contributing to resilience and recovery fol-
lowing Hurricane María among hotels in San Juan,
Finally, the hospitality industry is a major sup- Puerto Rico, and surrounding areas to answer the
porter of local and national economies,26 with the research question: What individual, organizational,
accommodations subsector alone supporting over two and community factors will hotel employees identify
million jobs in the United States in 2017 and stimu- that contributed to recovery and resilience following
lating $300 billion dollars in consumer spending. The Hurricane María?
hospitality industry is not immune to downturns, with
times of financial crisis in this industry negatively METHODS
affecting stability of jobs and consumer spending. The
consequences of poor financial performance follow- Procedure
ing a weather event can be particularly negative for All research activities were approved by the
employees within the hospitality industry, given that
the industry major employer of workers who belong Institutional Review Boards of each university that
to vulnerable populations, including low skill work- participated in this study. In November 2018, members
ers, low wage workers, and job insecure workers27 of the research team performed an initial visit to scope
who are more likely to lose their jobs when workers hurricane-related damage in San Juan, Puerto Rico,
are laid off.28 Promoting the positive functioning of and to identify hotels as data collection sites. After the
individuals and organizations in this industry during completion of that visit, a list of nine potential partner-
and after adverse events would support stability and ing hotels was created. These hotels were contacted
income security for vulnerable employees and local and with a request to support data collection on their
national economies. property. Four of these hotels were big chain 4-5 stars
hotels and five hotels were 2-3 stars and boutiques
The current research hotels. The selection criteria for the hotels was that
Despite the importance of developing resilience they stayed open with some operations after Hurricane
María. Only four of the smaller, locally owned hotels
to weather events in hospitality organizations, agreed to participate in the project.
hospitality-focused disaster resilience research suf-
fers from the same “unknowns” as the larger body The focus groups were moderated by two faculty
of resilience literature. Although the number of researchers and six graduate students with expertise in
studies focusing on hospitality resilience to disas- Occupational Health and Safety and Industrial Hygiene
ters has increased,6 hotel stakeholders still report or Industrial Organizational Psychology. Moderators
significant gaps in knowledge, capital, and commu- were trained at a planning workshop in February 2019.
nity-based partnerships that promote resilience to Specifically, they were provided with a moderator guide
natural disasters.24 Thus, there is an opportunity that contained information on avoiding verbal and
to address the need for actionable strategies to nonverbal bias in focus group moderation, encouraging
foster resilience that acknowledge the complex participation, and responding to potentially emotional
interdependencies of individuals, organizations, and content using skills from motivational interviewing
communities7 within the disaster management, (open-ended questions and affirming, reflective, and
summarizing statements). The moderators participated
in practice focus group exercises.

Special Issue on Puerto Rico

112 Journal of Emergency Management
Vol. 19, No. 8

In May 2019, the research team conducted series on factors contributing to resilience at multiple levels,
of 1-hour focus groups at partnering hotel properties. referencing each level depicted in Figure 1:
Four hotels participated, with properties ranging from
2-star to 3-star class and represented different styles 1.  Thinking about yourself as a person,
of hotels, including budget traveler hotels, bed and what personal characteristics have con-
breakfasts, and boutique hotels. Partnering facilities tributed to the ways that you are doing
ranged from 10 to 30 employees and 10 to 34 rooms. well since Hurricane María? What per-
They were located in a variety of areas, with two sonal characteristics have contributed to
located in Historic Old San Juan, one being located the ways in which you are still struggling
near the University of Puerto Rico, and one venue from Hurricane María?
located in Rio Grande. Focus groups were conducted
primarily in Spanish. English or a mix of Spanish and 2.  Thinking about yourself as an employee,
English was also used when employees were comfort- what aspects of your work life have con-
able with that language combination. tributed to the ways that you are doing
well since Hurricane María? What aspects
The focus group began with obtaining informed con- of your work life have contributed to the
sent, introductions for the moderators and participants, ways that you are still struggling from
and a description of rules for participation that empha- Hurricane María?
sized confidentiality and respectful communication. The
first question asked participants to describe the ways 3.  Thinking about yourself as a member of
in which Hurricane María impacted them as an indi- the community, what aspects of your com-
vidual, an employee, and a member of the community. munity have contributed to the ways in
The following three questions served as prompts for the which you are doing well since Hurricane
primary qualitative analyses and gathered information

Figure 1.  Conceptual model and a priori coding categories. Asterisk indicates that code was added after initial
review of transcripts upon author agreement.

Special Issue on Puerto Rico

Journal of Emergency Management 113
Vol. 19, No. 8

María? What aspects of your community translation by at least one research assistant fluent
have contributed to the ways in which you in Spanish. Translated notes and transcriptions were
are still struggling after Hurricane María? reviewed and coded according to a hybrid inductive
and deductive approach.29 Specifically, a code book
4.  Picture yourself, your organization, and was deductively developed with potential codes based
your community some time from now. on the following process. First, the authors reviewed
Imagine that you are able to say that you existing organizational and community resilience
have recovered completely and you, your toolkits such as the City Resilience Profiling Tool by
organization, and community are doing the United Nations Human Settlements Program-
well. What does that mean to you? Headquarters (UN-Habitat), the Organizational
Resilience Index by the British Standards Institution,
Following the focus group, participants were pre- and Community Resilience Indicators by the Federal
sented with a debriefing form that presented free or Emergency Management Agency.
sliding-scale mental health resources, in the event
that talking about their Hurricane María experiences Second, the authors reviewed the academic lit-
caused distress. Participants were also compensated erature for frameworks and models of organizational
with a $35 gift card as a participation incentive. resilience.30-32 Third, the authors reviewed an over-
view of community resilience models and toolkits,33
Participants which compares and contrasts six community resil-
Across all focus groups, 19 employees partici- ience models according to their conceptual lenses,
definition of resilience, scale of focus, dimensions
pated. The average number of participants in each or domains in the model, assessment and planning
focus group was five participants. The sample was steps, resilience tools, and strengths. Across each of
55 percent female and the most common age range these steps, the authors noted categories of resilience-
reported was 45-54 years of age (32.1 percent), fol- promoting factors common across multiple sources
lowed by 25-34 (26.6 percent) and 35-44 (22.9 per- and achieved consensus on the coding categories in
cent). Most participants worked in housekeeping (25 Figure 1 at a planning workshop in February 2019.
percent) or front desk (25 percent) roles, followed by A final coding category of “Other Major Themes” was
maintenance/repair (15 percent) or office and admin- added to the code book to allow themes to inductively
istrative (15 percent). The remainder of the sample emerge from the data. Based on the preliminary feed-
filled supervisory roles (10 percent) or owned the back from coders, a category of “Hurricane Impact”
hotel property (10 percent).* Most of the sample had was added and agreed upon by all study authors.
been working with their current hotel organization
for 5-10 years (30 percent), followed by 2-5 years (35 Coders were provided with a structured coding
percent) and less than 2 years (20 percent). worksheet containing a title and definition of all cod-
ing categories and were instructed to record a descrip-
Coding and analysis tion of major themes from the data consistent with
At least one focus group moderator took typed that category, the frequency of agreement with that
theme, and an illustrative quote or timestamp for that
notes during the focus group sessions. Focus group ses- major theme. Resulting codes were interpreted using
sions were audio recorded and transcribed. Notes and MAXQDA® analytic software34 (VERBI Software,
transcriptions were translated into English by a pro- 2019). An inter-rater agreement of 90 percent was
fessional transcription company. All transcripts were achieved between the two coders.
reviewed with audio for accuracy of transcription and
RESULTS
*Owners were only permitted to participate when they fulfilled daily
work tasks similar to employees. Four major themes were identified, some with
multiple subthemes related to the a priori deductive

Special Issue on Puerto Rico

114 Journal of Emergency Management
Vol. 19, No. 8

categories occurring within each theme. Actionable some geographical areas, shortages lasted for more
strategies associated with each deductive category than 6 months. According to the respondents, power,
are summarized in Table 1. For each major theme, water, diesel fuel, gas, and communication resources
illustrative quotes are provided in italics. were scarce. The post-storm timeline of the hurricane
resulted in a near-total shutdown of the economy that
Theme 1: Disaster response timeline: lasted for months, resulting in many people without
Preparedness, response, and recovery financial resources and worrying that it will become
even harder to survive. Some respondents reported
When discussing the impact of the hurricane, that higher prices of goods exacerbated the problem
participants discussed experiences related to prepar- and increased stress. Furthermore, others reported
edness, response, and recovery. These observations that they were affected by losing their job(s) and
were usually framed within the contexts of their per- ability to pay for basic necessities. This theme is
sonal strengths, challenges, or gaps in response and most closely linked to the deductive coding category
the associated impact. Overall, little was described in of Economic and Disaster Management Planning,
terms of the preparation phase, though many partici- indicating that an actionable strategy for individuals,
pants observed that they are more prepared for future organizations, and communities would be to regularly
storms. As such, the majority of the observations engage in planning and preparedness efforts.
focused on the response and recovery timeline.
Bueno, para mi por lo menos, me recupere
Lack of resources, infrastructure, and housing porque no era nada fácil. ¡Usted . . . hasta
built to withstand hurricane winds increased com- hambre pasé! ¡Hasta agua! Porque por lo
munity vulnerability. Some respondents linked their menos en mi casa duramos hartísimo sin
fears and stress to the lack of resources (power, water, agua. [I recovered [but] it wasn’t easy at
food, and electricity shortages), which were an ongo- all. I experienced even hunger. Hunger!
ing and consistent problem for the communities. This
was especially true in the first few months, and in

Table 1. Strategies to promote resilience

Deductive codes Associated strategy

Economic and disaster management Engage in planned, frequent disaster preparedness efforts; encourage disaster
planning planning efforts that integrate individual, organizational, and community levels

Individual attitudes and behaviors Offer workplace or community sponsored resilience training that teaches individual
coping skills

Experience/learning/training Offer workplace trainings for disaster response

Culture of caring, social networks, and Participate in community-level social capital building; encourage individual and
leadership workplace disaster plans that leverage social networks; ensure strong community
leader response to disaster

Awareness and use of resources/ Create workplace and community awareness campaigns for disaster recovery
infrastructure resources so that individuals are aware of resources; encourage resource utilization
plans for individuals, workplaces, and communities

Transformative potential Invest in critical infrastructure that will allow individuals, workplaces, and
communities to return to functioning quickly, ie, electrical grid, roads, and
hospitals, create policies that ensure equitable distribution of disaster management
resources within organizations and communities; utilize community strengths in
recovery plans

Special Issue on Puerto Rico

Journal of Emergency Management 115
Vol. 19, No. 8

Need of water! Because in my house, we looking at starry nights since there was less light pol-
had to last a long time without water.]— lution due to power outages).
Participant from budget traveler hotel
near University of Puerto Rico. Participant: “I had never been through any-
thing so harsh. So in all aspects of my
The constant feeling that the fear is pre- life, I was just very thankful.” Facilitator:
sent that something like that could hap- “So that just made you kind of take a step
pen again and I have to be prepared. But back and appreciate what you still have left.”
in a way, I feel that we don’t have enough Participant:“Yes.”—Participant from boutique
resources around us to actually be pre- hotel in Historic San Juan, spoken in English.
pared for that kind of thing.—Participant
from boutique hotel in Historic San Juan, On the organizational level, respondents reported
spoken in English. being grateful for their employment with the hotels
because they were some of the few still operating.
. . . . María ayudo a muchas personas a Many noted that their employer allowed them to
pensar, ‘Okay, tengo que estar preparado’ stay on property post-storm, as well as use amenities
[María helped a lot of people to think,“Okay, like washing machines, generators, or electricity to
I have to be prepared.”]—Participant from charge electronics, and take ice back to their home.
boutique hotel in Historic San Juan. Employees noted that they worked long hours doing
jobs that were not routine to them post-storm to clear
Theme II: Impact in multiple contexts: Individual, debris. Their properties also housed many of the gov-
organizational, and community experience ernment responders who came to the island. Some felt
a sense of normalcy post-storm by going to work, while
Participants discussed experiences that impacted others found it difficult to separate the trauma of the
them in multiple contexts, meaning as an individual, storm from work activities. Respondents reported
an employee in a hotel, and a member of a com- that the experience of going through Hurricane María
munity. The discussion of impact on multiple levels helped them to better prepare the hotel for a disaster
connects to a number of deductive coding catego- through lessons learned, and that the sense of com-
ries: Individual Attitudes and Behaviors (individual radery that developed among the staff taught them
level), Experience/Learning/Training (organizational personal skills that were useful to foster social cohe-
level), and Culture of Caring, Social Networks, and sion and resilience in their community.
Leadership (community level). Actionable strategies
based on this major theme could include offering Tu no puedes separar el laboral a lo per-
training related to individual coping skills, planning sonal aquí, después de esta catástrofe. Eso
organization training related to disaster response, todo este entrelazado. ¿Por qué? Porque
and helping communities to identify and strengthen tienes personas que vienen a trabajar a
partnerships. un trabajo marizadas. Por lo tanto, a
nosotros sí nos afecto mucho económica-
Regarding individual experiences, participants mente. [You can’t separate working situa-
referenced fear during the storm and traumatic tion from personal situation. They impact
impact from the visual destruction in the aftermath of each other. Why? Because the people who
the storm. Many experienced difficulties in commut- came to work who are still traumatized.
ing and communicating. To cope with such difficulties, So, this is everything. It didn’t affect just
some relocated for an extended period of time until economically.]—Participant from bed and
electricity returned. Others coped through a positive b­ reakfast/wedding venue in Río Grande.
attitude and gratitude, faith, and finding positive
outcomes post-hurricane (such as being able to enjoy

Special Issue on Puerto Rico

116 Journal of Emergency Management
Vol. 19, No. 8

.  .  . No dejo este trabajo porque yo soy una los arboles, le metían mano entre todos
persona que soy agradecida. Y aquí, aquí he y a la misma vez nos divertíamos y nos
aprendido a ser responsable. Aquí he apren- echábamos chistes. Porque lo teníamos que
dido a lidiar con diferentes caracteres, difer- hacer y sin paga . . . era para que circularan
entes nacionalidades. Entonces me ha dado los vehículos, porque, sino ¿que carro iba a
una gran experiencia y eso ha sido uno de los entrar aquí a San Juan? Los que podían
triumfos por lo cual yo no dejo este trabajo venían . . . y los demás no podían. Y eso fue
porque tu aprendes algo nuevo todos los días. lo que .  .  . lo que nos unimos todos. Toda
Y no solamente lo tienes que aplicar aquí, esta parte de la comunidad, y empezamos
lo puedes aplicar en la sociedad, lo puedes a picar los arboles, a limpiar un poco mas
aplicar en todos tus trabajos y lo puedes la calle. Fue algo divertido y hubo mucha
aplicar de diferentes formas. [I don’t leave unión entre los vecinos. [There was a lot of
this job because I am a person who knows unity among community members. If we
how to be grateful. And here I learned to be had to remove trees, we all lend a hand and
responsible. Here I learned to get along with at the same time we would have fun and
different kinds of personalities. I understand joke around. We had to do it and without
different nationalities. It’s a great experi- pay. It was so vehicles could circulate a
ence and this has been one reason at this job little. Otherwise, what vehicle would enter
I learn something new every day that you San Juan? The ones that could come and
don’t have to apply just here. You can apply pass or else they couldn’t. That was what -
it in society. You can apply it at any job. we all joined hands and we began to work,
And you can apply it in different ways.]— to clean a little, and there was a lot of unity
Participant from budget traveler hotel near among neighbors.]—Participant from con-
University of Puerto Rico. temporary hotel in Historic Old San Juan.

At the community level, respondents talked about Entonces, a nivel de comunidad, yo
a sense of pride in being Puerto Rican and the gener- entiendo no solo en épocas de huracán
ous acts they experienced in their communities, such as que la universidad debe unirse un poquito
the sharing of food and electricity, and clearing debris mas a la comunidad y darnos un poquito
together. Some reported that their hotel provided hot mas de ayuda a la comunidad, tanto de
meals to the community post-storm. One respondent seguridad. [So, at the community level, I
mentioned that they felt like they could not freely speak understand that it’s not only the hurricane
their political opinions to government employees stay- and that the university should get a little
ing at the hotel, and others mentioned a general lack of closer to the community and provide a lit-
trust in the government, both local and federal. Some tle more help to the community as far as
respondents noted that the local university is a great safety.]—Participant from budget traveler
resource that could do more to engage with the local hotel near University of Puerto Rico.
community and possibly serve as an emergency shelter
in the future. Many local agencies were hindered post- Theme III: Resources build capacity:
storm, such as police who could not get to emergency Resource loss and gain
calls because of road conditions, and sanitation crews
unable to pick up trash which led to disease concerns. When discussing resources, many participants
described a loss of personal support or necessity
Pero hubo mucha unión aquí en la comu- resources post-hurricane. Most participants reported
nidad .  .  . Di ‘okay’, si tenias que cortar the loss of at least one of the following resources:
power, employment, housing, transportation, water,

Special Issue on Puerto Rico

Journal of Emergency Management 117
Vol. 19, No. 8

food, and communication resources. In some cases, Theme IV: Strength with struggle:
participants reported being without such resources Acknowledging challenges and triumphs
for a great length of time. The struggle to maintain
resources influenced stress, fear, and flashbacks. When discussing Hurricane María, it became clear
However, participants did report leveraging their that although participants endured great struggle,
connection with their employer or their community to they also identified several strengths that influenced
either secure resources or cope with loss of resources. both resources and response. In terms of struggle, par-
For example, the majority of hotel workers shared ticipants described a breakdown of crucial infrastruc-
a positive experience in terms of how their fellow ture, wide-spread damage to buildings and roadways,
employees and community members supported each lack of access to food and water, inequities in aid dis-
other through recovery efforts. Many were able to live tribution, and fear of theft and assault. The associated
in the hotel with their families for free while working impact of this struggle is undeniable, as participants
long hours or being able to use hotel amenities such reported anxiety, depression, trauma, and PTSD.
as showers and electricity. Some employees assumed
new duties related to offering resources and sup- However, in terms of strength, participants shared
port, such as cooking for other employees and hotel stories of coming together for the greater good. For
residents. According to those employees, they gained example, some hotels provided hot meals to the com-
better experience in how to respond to disaster events munity after the storm. The respondents worked well
and share responsibilities with other employees. together to repair the hotel after Hurricane María
and were flexible with what duties they took on.
Together, this theme suggests that resources con- Many noted that this positive functioning was made
sist of individual, workplace, and community capaci- possible by resources provided by the hotel, such as
ties, but that both damage and resilience impact a place to stay, electricity, ice, and salary continuity.
the availability of these resources. Lessons learned
within this theme most closely relate to the Hurricane The content of this theme most closely aligns with
Impacts and Awareness and Use of Resources/ the Transformative Potential deductive coding cat-
Infrastructure deductive coding category. Actionable egory. Participants directly offered many actionable
strategies emerging from this theme include a greater strategies within this theme, which include investing
attention to existing infrastructure within organiza- in critical infrastructure such as the electrical grid,
tions and communities, as well as efforts to promote increasing preparedness in essential services such as
awareness and effective utilization of resources. hospitals, creating policies that promote equitable aid
distribution, forming and strengthening community
Era como sentir que entre la tempestad partnerships, and framing recovery efforts on the
todo estaba bien hotel veníamos a trabajar. existing strengths of the community, ie, a focus on
Nunca nos quedamos sin trabajo. Veníamos tourism to boost the local economy.
a limpiar el hotel.  .  . lo manejamos súper
bien .  .  . Y teníamos un horario, o sea, no And, again, we were given the option of
querían dejar el hotel . . . no había muchos taking a shower here, we could stay even if
recursos. [I was like feeling with the storm it was just one night. You would feel nor-
everything was okay because we came mal. You had air conditioning for just one
to work. We were never without work. night, or charge all your stuff and wash
We came to clean and we worked super your clothes, which was so difficult at that
well. We had schedules. They did not want time.—Participant from boutique hotel in
to leave the hotel. There were not many Historic San Juan, spoken in English.
resources.]—Participant from bed and
breakfast/wedding venue in Rio Grande. Nuestro primer evento con una pareja de
locales, la boda fue un éxito. Pero fue un
éxito. ¿Por qué? Por el esfuerzo de todos

Special Issue on Puerto Rico

118 Journal of Emergency Management
Vol. 19, No. 8

aquí. [Our first event with a couple, the prevalent in previous resilience literature across
wedding was a success, but it was a success a number of disciplines. Most notably, this study
because of everyone’s effort.]—Participant acknowledges and directly assesses interdependen-
from bed and breakfast/wedding venue in cies at multiple levels within an individual’s ecologi-
Rio Grande. cal system and results in actionable strategies to help
foster resilience. Major themes from this qualitative
DISCUSSION research may support future intervention work that
aims to promote a more concerted disaster planning
The literature on resilience stemming from mul- effort in organizations and communities, individual
tiple disciplines provides a useful lens through which attitudes and behaviors supportive of resilience,
to examine impact and recovery from weather events greater investment in infrastructure and aware-
such as Hurricane María. The unique vulnerabilities ness and use of resources, meaningful collaborations
and critical economic and humanitarian importance between organizational and community entities, and
of the hospitality industry highlight the need to study critical analysis of previous experience and collective
resilience from weather events in this particular sec- identity that supports transformative potential.
tor. Data from qualitative focus groups conducted
among hotel employees in San Juan, Puerto Rico, There are several limitations that should be
and surrounding areas revealed a number of themes addressed. The partnering hotels were smaller,
related to resilience and associated actionable strate- independently owned hotels, and future research
gies to help individuals, organizations, and communi- should replicate these findings in larger chain hotels.
ties prepare for future adverse weather events. Although geographic regions assessed were repre-
sentative of San Juan and the surrounding areas,
Participants reported a general disaster timeline future research should also replicate these findings in
of preparedness, response, and recovery, while noting other geographic areas beyond Puerto Rico that also
that the preparedness phase was generally lacking for feature climate-sensitive industries and exposure to
Hurricane María. On a more positive note, many partic- weather events. Future research could also examine
ipants reported more awareness of preparedness steps the cumulative effects of multiple types of adverse
for future weather events. Both impact and response events or disasters. For example, Brown et al.6 offer
occurred on multiple levels within participants’ eco- both extreme weather events and public health crises
logical system, including individual, organizational, as examples of disasters that could impact the hospi-
and community levels. Participants shared stories of tality industry. Although the timing of data collection
severe and prolonged resource loss, with some being did not allow for an examination of resilience toward
able to secure vital resources through connections with both types of disasters, hospitality organizations
their work organization and community. Finally, a currently find themselves impacted by both hurri-
crucial theme emerged in the coexistence of struggle cane season and the COVID-19 pandemic.35 Future
and strength. While the negative impacts of Hurricane research should ensure that resilience toward mul-
María are undeniable, participants shared stories of tiple adverse events is well understood and incorpo-
collective strength that helped individuals, organiza- rated into theoretical models and intervention efforts.
tions, and communities recover. This theme highlights Although the creation of the coding scheme integrated
the transformative potential and capacity for resilience many existing theoretical models and tool kits, it is
that is inherent in the connections between interde- possible that important antecedents of resilience may
pendent systems. That is, messages of “I am strong” not have been reflected in the coding scheme. As such,
were often outweighed by messages of “we are strong.” we employed a balance of inductive and deductive
methods that would allow important unaddressed
Strengths, limitations, and future research directions themes to emerge. As a final future research direction,
The features of this qualitative study address we encourage researchers to replicate the findings in

a number of conceptual and operational limitations

Special Issue on Puerto Rico

Journal of Emergency Management 119
Vol. 19, No. 8

other industries that may be particularly vulnerable recover and thrive following weather events. Hotel
to extreme weather events and play an important role employees discussed their experiences as individu-
in disaster recovery. Doing so may help promote a als, employees, and community members related to
balanced community-level response to disasters that Hurricane María in focus groups conducted in San
effectively balances the input of multiple subsystems Juan, Puerto Rico, and surrounding areas. Findings
such as industries. reveal that future efforts to promote resilience against
weather events in hospitality organizations must
Implications for research and practice emphasize preparedness and planning, organizational
The emergent themes associated with this and community partnerships, resource availability,
and the transformative potential of communities.
research bear a number of implications for research
and practice. The interdependencies among indi- ACKNOWLEDGMENTS
vidual, organizational, and community impacts and The Sunshine Education and Research Center is funded in
resources support the increased use of ecological part by the Centers for Disease Control and Prevention, National
models and multilevel analytical techniques in the Institute for Occupational Health and Safety (CDC/NIOSH Training
resilience literature. The endorsement of actionable grant #T42-OH008438).
predictors of recovery and resilience demonstrates We would like to acknowledge the assistance of the follow-
that resilience researchers in multiple disciplines ing individuals with data collection and/or analysis: Katherine
should more often measure predictors of resilience Ciarlante, Liz A. Nieves-Santiago, Alyssa Perez, Zoé Rodríguez, and
in addition to measuring “having bounced back.” In Miriam F. Escobar. We are grateful to Dr. Steve Jex for his friendly
practice, the results support a collaborative and coop- review of this manuscript.
erative approach. Interdisciplinary practice teams
are likely needed to develop individual-level training Kristin A. Horan, PhD, Psychology Department, University of Central
programs, organizational planning and preparedness Florida, Orlando, Florida. ORCID: https://orcid.org/0000-0001-9318-
efforts, and advocacy and partnerships at the com- 044X.
munity level. The findings suggest that traditionally
siloed approaches focusing on individuals, work- Blake Scott, MPH, University of South Florida, Tampa, Florida.
places, or communities are limited. Indeed, although
the actionable strategies may seem too simple, focus Ahlam Farzan, MD, University of South Florida, Tampa, Florida.
group responses indicated that a lack of coordination
across these levels prevented planning, awareness, Marian L. Ortíz-Aponte, MS, University of Puerto Rico, San Juan, Puerto
or effective implementation of disaster response. A Rico.
comprehensive approach that integrates individuals,
workplaces, and communities is needed to maximize Alejandra Rivera-García, MS, University of Puerto Rico, San Juan,
the reach and impact of actionable strategies such as Puerto Rico.
disaster planning or training.
Jennifer Marshall, PhD, University of South Florida, Tampa, Florida.
CONCLUSION
Anthony J. Masys, PhD, University of South Florida, Tampa, Florida.
Weather events have and will likely continue
to increase, creating challenges for industries with Mindy Shoss, PhD, University of Central Florida, Orlando, Florida.
climate sensitive inputs and demand patterns, cumu-
lative strain from repeated disruptive events, and Adriana Campos, MPH, University of South Florida, Tampa, Florida.
vulnerable populations of workers, such as those
employed in the hospitality industry. Insights from Lida Orta-Anés, PhD, University of Puerto Rico, San Juan, Puerto Rico.
multiple disciplines examining resilience can help
hospitality employees and hospitality organizations REFERENCES
1. Kishore N, Marqués D, Mahmud A, et al.: Mortality in Puerto
Rico after Hurricane María. N Engl J Med. 2018; 379(2): 162-170.
2. Santos-Lozada AR, Kaneshiro M, McCarter C, et al.: Puerto
Rico exodus: Long-term economic headwinds prove stronger than
Hurricane María. Popul Environ. 2020; 42(1): 43-56.
3. Rodríguez-Díaz CE: María in Puerto Rico: Natural disaster in a
colonial archipelago. Am J Public Health. 2018; 108(1): 30-32.
4. Rosselló J, Becken S, Santana-Gallego M: The effects of natural
disasters on international tourism: A global analysis. Tourism
Manag. 2020; 79: 104080.

Special Issue on Puerto Rico

120 Journal of Emergency Management
Vol. 19, No. 8

5. Racherla P, Hu C: A framework for knowledge-based crisis man- 20. AlBattat AR, Mat SA: Emergency preparedness for disas-
agement in the hospitality and tourism industry. Cornell Hosp Q. ters and crises in the hotel industry. Sage Open. 2010; 3(3):
2009; 50(4): 561-577. 2158244013505604.
6. Brown NA, Orchiston C, Rovins JE, et al.: An integrative frame- 21. Tan A: The impact of hurricanes on supply and demand: What
work for investigating disaster resilience within the hotel sector. owners and investors of Florida hotels should know. Consulting
J Hosp Tourism Manag. 2018; 36: 67-75. report by HVS International. 2006. Available at https://www.hospi
7. Rose A: Defining and measuring economic resilience to disas- talitynet.org/file/152002742.pdf. Accessed July 9, 2021.
ters. Disaster prevention and management. Int J. 2004; 13(4): 22. Tourism Economics: The impact of Hurricane Irma on the
307-314. Florida Tourism Economy. A report prepared for Visit Florida. 2018.
8. Hillmann J: Disciplines of organizational resilience: Contributions, Available at https://www.visitflorida.org/media/46680/fl-hurricane-
critiques, and future research avenues. Rev Manag Sci. 15(4): impacts.pdf. Accessed July 9, 2021.
879-936. 23. Palmer B: Resilient Puerto Rico is open for visitors. Professional
9. Miller F, Osbahr H, Boyd E, et al.: Resilience and vulnerability: Convention Management Association. 2020. Available at https://
Complementary or conflicting concepts? Ecol Soc. 2010; 15(3). www.pcma.org/puerto-rico-earthquake-recovery-business-events/.
Available at https://www.jstor.org/stable/26268184?seq=1#metadata_ Accessed July 9, 2021.
info_tab_contents. 24. Ivkov M, Blešic´ I, Janic´evic´ S, et al.: Natural disasters vs hotel
10. Southwick SM, Bonanno GA, Masten AS, et al.: Resilience defi- industry resilience: An exploratory study among hotel managers
nitions, theory, and challenges: Interdisciplinary perspectives. Eur from Europe. Open Geosci. 2019; 11(1): 378-390.
J Psychotraumatol. 2014; 5(1): 25338. 25. Lagiewski R: How the hospitality industry plays a vital role
11. Vogus TJ, Sutcliffe KM: Organizational resilience: Towards a in relief efforts when disaster strikes. GreenBiz. September 18,
theory and research agenda. In 2007 IEEE International Conference 2019. Available at https://www.greenbiz.com/article/how-hospital
on Systems, Man and Cybernetics. Ieee, 2007; 3418-3422. ity-industry-plays-vital-role-relief-efforts-when-disaster-strikes.
12. Lengnick-Hall CA, Beck TE, Lengnick-Hall ML: Developing Accessed July 9, 2021.
a capacity for organizational resilience through strategic human 26. International Trade Association: Travel, tourism, & hospitality
resource management. Hum Resour Manag Rev. 2011; 21(3): spotlight: The travel, tourism, and hospitality industry within the
243-255. United States. Available at https://www.selectusa.gov/travel-tour
13. Hartmann S, Weiss M, Newman A, et al.: Resilience in the ism-and-hospitality-industry-united-states. Accessed July 9, 2021.
workplace: A multilevel review and synthesis. Appl Psychol. 2020; 27. Iverson RD, Deery M: Turnover culture in the hospitality indus-
69(3): 913-959. try. Hum Res Manag J. 1997; 7(4): 71-82.
14. Ruíz-Martin C, López-Paredes A, Wainer G: What we know and 28. Desmond M, Gershenson C: Housing and employment insecu-
do not know about organizational resilience. Int J Prod Manag Eng. rity among the working poor. Soc Probl. 2016; 63(1): 46-67.
2018; 6(1): 11-28. 29. Fereday J, Muir-Cochrane E: Demonstrating rigor using the-
15. Gibson CA, Tarrant M: A ‘conceptual models’ approach to organ- matic analysis: A hybrid approach of inductive and deductive cod-
isational resilience. Aust J Emerg Manag. 2010; 25(2): 6. ing and theme development. Int J Qual Methods. 2006; 5(1): 80-92.
16. Norris FH, Stevens SP, Pfefferbaum B, et al.: Community 30. Allen R, Toder F: A model of organizational recovery. J Emerg
resilience as a metaphor, theory, set of capacities, and strategy for Manag. 2004; 2(1): 41.
disaster readiness. Am J Commun Psychol. 2008; 41(1-2): 127-150. 31. Bruneau M, Chang SE, Eguchi RT, et al.: A framework to quan-
17. Linnenluecke MK, Griffiths A, Winn M: Extreme weather titatively assess and enhance the seismic resilience of communities.
events and the critical importance of anticipatory adaptation Earthq Spectra. 2003; 19(4): 733-752.
and organizational resilience in responding to impacts. Bus Strat 32. Patriarca R, DiGravio G, Costantino F, et al.: An analytic frame-
Environ. 2012; 21(1): 17-32. work to assess organizational resilience. Saf Health Work. 2018;
18. Trenberth KE, Jones PD, Ambenje P, et al.: Observations: 9(3): 265-276.
Surface and atmospheric climate change. In Solomon S, Qin D, 33. Colussi M: Overview of community resilience models and toolkits.
Manning M, et al. (eds.): Climate Change 2007: The Physical Transition United States. Available at http://www.tranzsitionus.org/
Science Basis: Contribution of Working Group I to the Fourth sites/default/files/Overview%20of%20Community%20Resilience%20
Assessment Report of the Intergovernmental Panel on Climate Models%20and%20Toolkits%20(1).pdf. Accessed July 9, 2021.
Change. Cambridge: Cambridge University Press, 2007; 235-336. 34. VERBI Software: MAXQDA 2020 online manual. Available at
19. Wilbanks TJ, Romero LP, Bao M, et al.: Industry settlement maxqda.com/help-max20/welcome. Accessed July 9, 2021.
and society. In Parry ML, Canziani OF, Palutikof JP, et al. (eds.): 35. Mamane D, Dietzel L, Wolf M, et al.: Businesses already
Climate Change 2007: Impacts Adaptation and Vulnerability: strained by COVID-19 should prepare for hurricane season. The
Contribution of Working Group II to the Fourth Assessment Report Real Economy Blog. July 22, 2020. Available at https://realecon
of the Intergovernmental Panel on Climate Change. Cambridge: omy.rsmus.com/businesses-already-strained-by-covid-19-should-pre
Cambridge University Press, 2007; 357-390. pare-for-hurricane-season/. Accessed July 9, 2021.

Special Issue on Puerto Rico

Journal of Emergency Management 121
Vol. 19, No. 8

Notes

Special Issue on Puerto Rico
122 Journal of Emergency Management

Vol. 19, No. 8

JEM Displacement after disaster: Challenges and
opportunities responding to Puerto Rican evacuees

in Central Florida after Hurricane Maria

Tanya Buhler Corbin, PhD

ABSTRACT INTRODUCTION

Major disasters often displace populations, Background
requiring coordinated response efforts from govern- Hurricane Maria hit Puerto Rico on September
mental and voluntary organizations to support and
assist an influx of evacuees. Despite the frequency 20, 2017. The Category 4 storm brought widespread
and significance of this occurrence, this aspect of flooding, wind speeds of 155 mph, and devastated the
disaster management has received limited scholarly island.1 As part of the most active hurricane season the
attention, with research predominantly focusing on United States experienced since 2005, Hurricane Maria
response and recovery at the impacted disaster sites. was the third major hurricane to hit the United States
This study investigates disaster management and in just a few weeks, following closely behind Hurricanes
support offered to Puerto Rican evacuees arriving in Harvey and Irma. Hurricane Harvey impacted an area
Central Florida after Hurricane Maria. A multia- that included Houston as a Category 4 storm, dump-
gency resource center (MARC) was established at ing heavy rains lasting for several days and causing
the Orlando Airport to support evacuees, facilitat- massive flooding. Less than 2 weeks later, Hurricane
ing a network of agencies to coordinate response Irma made landfall across several Caribbean islands,
efforts. The analysis uses data obtained through including Puerto Rico, before hitting Cudjoe Key in Key
in-depth, semistructured interviews from members West, Florida as a Category 4 hurricane on September
of the Orlando metropolitan area disaster response 6, 2017, then continuing north through the middle of
community who worked at the MARC center, and the state of Florida. By the time Hurricane Maria made
content analysis of state and local government docu- landfall in Puerto Rico, resources were stretched beyond
ments. The results identify challenges and successes capacity, particularly from the Federal Emergency
and common themes in disaster support for displaced Management Association (FEMA), which was simulta-
evacuees after disasters. Results confirm some con- neously responding to a succession of major storms and
ventional wisdom about disaster response, such as severe wildfires in the western states.2
persistent significant challenges related to shelter
and housing, and offer new insights about how to Further exacerbating the challenges in respond-
successfully support displaced evacuees, particularly ing to Hurricane Maria were additional factors spe-
using MARCs. This study provides foundational cific to the island. FEMA’s 2017 hurricane season
knowledge to inform future research and contributes after action report noted that planning assumptions
to the development of successful support for those severely underestimated the impacts of a major
who are displaced after disaster. hurricane landfall in Puerto Rico, so they were
underprepared to respond.2 Puerto Rico was already
Key words: disaster response, Hurricane Maria, struggling with infrastructure challenges and social
displacement, Florida, disaster management vulnerabilities, including a weak power grid, an
economy in crisis, the historical effects of colonialism,

DOI:10.5055/jem.0624 Special Issue on Puerto Rico
Journal of Emergency Management
Vol. 19, No. 8 123

and dysfunctional politics before the hurricane.3,4 In The Hurricane Katrina diaspora offers important
the face of many challenges, Puerto Rico experienced insights for understanding other events that cause
an out-migration from the island to the mainland for internal displacement in the United States, and
a decade preceding Hurricane Maria.5 According to how to best respond to these disasters. In the edited
the Puerto Rico Research Hub: “The state of Florida collection by Weber and Peek,13 scholars provide
is now home to more than one million Puerto Ricans. a thorough exploration of the experiences of Gulf
One out of every five Latinos in Florida is Puerto Coast residents who were displaced after Hurricane
Rican, and the Orlando metropolitan area is one of Katrina. They include major United States host cit-
the most concentrated regions with Puerto Rican ies: Denver, Atlanta, Louisiana (Baker, Baton Rouge,
residents [in Florida]. The estimates of the post- Lafayette, and New Orleans), Mississippi (Jackson),
Hurricane Maria migration indicate Florida as the Missouri (Columbia), South Carolina (Columbia/West
most likely destination for Puerto Ricans leaving the Columbia), and Texas (Austin, Dallas, Houston, and
island.”6 Huntsville).13 Among the important insights from this
work was a more thorough understanding of the ways
Prior research: Displacement after disasters in which pre-existing vulnerabilities exacerbated the
After disasters, residents in affected areas are challenges the evacuees faced in the new communi-
ties. After disasters, affected communities already
often forced to leave their homes and communities. face daunting challenges of rebuilding and recover-
Depending on the scope and severity of the event, ing. When displaced populations navigate establish-
among other factors, this evacuation can be shorter ing essential lifelines such as housing, healthcare,
term or much longer term, spanning from weeks, to employment, transportation, schooling, and childcare
months, and in some cases, years. The term “displace- in a new place without social networks and amid
ment” has typically been used in reference to inter- trauma, challenges are significantly increased.
national conflict regions, and countries experiencing
political instability that results in forced displace- The need to secure temporary and permanent
ment of residents.7 However, in recent decades, disas- housing for those displaced by disasters has received
ters have brought displacement concerns more promi- more attention from scholars in recent years.14,15
nently to the forefront in politically stable nations, Securing housing after disaster is one of the most
including in the United States.8,9 prominent challenges facing those affected by dis-
asters.15 Establishing stable housing is essential for
Increasingly, developed nations are contending other dimensions of recovery, as it serves as an anchor
with widespread displacement, which may be tempo- for re-establishing stability and consistency in other
rary or become longer lasting or permanent, depend- recovery areas.16 Hurricane Katrina revealed failures
ing on a variety of push and pull factors. Push factors in planning for shelter and housing needs when many
are conditions in the place of origin that affect the evacuees are displaced. As a result, disaster housing
likelihood of displacement, such as economic oppor- policy changes have been enacted, including the con-
tunities, political intolerance, or other hardships. Pull gressional directive to develop a National Housing
factors are the opportunities present in the place of Strategy, to meet short- and long-term housing needs
displacement that can improve a person’s life, such for people affected by disasters.17 These housing pol-
as job opportunities, attractive amenities, access to icy changes have unfortunately had limited success.18
good schools, and affect the likelihood that evacuees Current planning practices have paid little atten-
will remain displaced.10 The most prominent exam- tion to housing large numbers of displaced people
ple of internal displacement in the modern era in beyond sheltering and short-term accommodations
the United States occurred after Hurricane Katrina, after disaster.
where nearly 1.5 million people were displaced
across the Gulf Coast region (Louisiana, Mississippi, Displaced populations facing temporary shel-
Alabama, Florida, and Georgia).11,12 ter and permanent housing challenges are also

Special Issue on Puerto Rico

124 Journal of Emergency Management
Vol. 19, No. 8

in need of intensive resources and support. The Research question
receiving communities must be prepared with the Despite the displacement that is common when
organizational capacity and resources to support
their arrival. This includes coordination and coop- an area is hit by large-scale, high-impact disasters,
eration across multiple jurisdictions as well as the scholars know little about how communities sup-
need to use improvization and innovation during port the displaced population during a disaster.
the response.19 These substantial challenges are Scholars are beginning to seek to address the gap in
crucially important to understand, as evacuees will our knowledge, calling for research to examine the
continue to need host communities and provide “space between short-term, temporary (and often
resources in future disasters. recurrent) evacuations and long-term, permanent
relocations” after disasters.23 Communities receiving
Multiagency resource center a large influx of evacuees will navigate the challenges
Multiagency resource centers (MARCs) have of resource allocation and coordination, among oth-
ers. Given the importance of Florida, and particularly
recently been developed and used as part of a the Orlando region, as a central location for evacuee
national mass care strategy in responding to dis- displacement after Hurricane Maria, this study inves-
asters and aiding those who are displaced. The tigates the following research question: how did the
MARC Planning Resource document was writ- disaster response organizations in Central Florida
ten by a team that includes the American Red respond to support evacuees from Puerto Rico after
Cross, Catholic Charities, The Salvation Army, and Hurricane Maria?
FEMA.20 According to the planning document (p. 7),
“MARCs are designed to: Expedite individual, fam- METHODS
ily and community recovery following a disaster;
Provide efficient, effective assistance to individu- Case study and case selection
als and families affected by a disaster in a single, A case study approach is used for this research
‘one stop shop’ location; Minimize the time and
travel distance needed for affected individuals and because the primary goal is to understand the pro-
families to obtain assistance after a disaster; Aid cesses involved in disaster response and support for
reunification of family and friends; Maximize the those displaced by disaster and requires in-depth
use of collective resources and expedite the ability explanation by those managing this response.24 The
of organizations to deliver services by: 1) eliminat- Orlando metropolitan area is selected for this case
ing the need for participating organizations to set study, as this region received the largest displace-
up individual assistance centers, and 2) facilitating ment of evacuees from Puerto Rico after Hurricane
coordination and information sharing between par- Maria. Estimates indicate that between 114,000 and
ticipating organizations; Enable effective coordina- 213,000 people (2-4 percent) of the total population
tion with government and non-government agencies; of Puerto Rico left in the year after Hurricane Maria,
Facilitate the transition to long-term recovery.”21 with Florida receiving the highest numbers of Puerto
A MARC center was established at the Orlando Rican displaced residents.25,26
Airport immediately after Hurricane Maria to sup-
port evacuees arriving from Puerto Rico. The center Interviews
operated at the Orlando Airport until it was moved In-depth interviews were conducted with emer-
from the airport into the community on January 2,
2018 and continued to serve evacuees in the com- gency management, nonprofits, and faith-based
munity until March 18, 2018.22 Nearly 35,000 people leaders active in the Orlando metropolitan area
were processed in about 5 months at Orlando MARC disaster response to Hurricanes Irma and Maria in
(Interviewee 1, Interview, May 16, 2018). May 2018. Purposive sampling was used to identify
initial potential interviewees from the agencies
that responded to Hurricanes Irma and Maria, par-
ticularly the agencies participating in the Orlando

Special Issue on Puerto Rico

Journal of Emergency Management 125
Vol. 19, No. 8

MARC at the Orlando Airport. Additional partici- Although the interview subjects consented to
pants were identified through snowball sampling. allowing their names to be used, because this geo-
The purposive sampling method is commonly used graphic location and the sample size are a smaller
in qualitative disaster research, as it allows for the than anticipated subset of the original research design,
identification of interview subjects who are most the decision was made to preserve the confidentiality
relevant to the study. Supplementing this with of the interviewees by omitting potentially identify-
snowball sampling facilitates the identification of ing information about each individual to avoid the
additional interviewees who are not readily identifi- possibility of inadvertent harms to them. Therefore,
able to the researcher a priori but can add impor- descriptions of the interviewees’ organization type and
tant insights.27 response roles are intentionally general (Table 1).

Interviewees were experienced in disaster man- After receiving IRB approval, potential inter-
agement in Florida and worked together at the MARC view subjects received an initial email inviting their
Orlando Airport center assisting evacuees from Puerto involvement to participate in the research study and
Rico when they arrived in Orlando. FEMA is missing be interviewed, which included written consent infor-
from the sample, as they did not respond to a request mation. Interviews were conducted in-person at the
for an interview. Their absence does not detract signif- interviewees’ places of business, with one interview
icantly from the insights from this research, though conducted over the phone because of scheduling chal-
it would be preferable to include them. This study lenges when the interviewer was in the region con-
seeks to understand how a local community (Orlando) ducting interviews. Five interviewees were identified
coordinated within their existing disaster response in the initial sample frame, with the remaining three
structures and how they engaged in improvisation identified from the snowball sampling.
to accommodate and assist displaced evacuees; thus,
the local perspective is the most informative for the Audio recordings of interviews were transcribed
research question. by a professional company. Interviews were semis-
tructured, with each interview length of time varying
The interviews analyzed in this study are part of from 21 minutes to 1.5 hours. One interviewee was
a larger research project evaluating disaster response not recorded, and thus, handwritten notes were used
to Hurricanes Irma and Maria. For this research in place of audio file and transcription. The interview
question, eight interviewees were relevant. There questions were designed to understand organizational
is no established minimum number of interviewees planning and capacity, operations, and coordination of
required for qualitative research, but a recent study the disaster response. Questions focused on cooperative
has documented purposive samples reaching satura- efforts between other disaster response organizations,
tion within the first 12 interviews, with metathemes what went well and what was challenging or did not
present with six interviews.28 There were an esti- go well, as well as policy recommendations for changes
mated 40 organizations and agencies responding to for the future. Interview transcripts were analyzed
the displaced evacuees at the MARC Orlando airport using open coding, from the grounded theory approach
center, and eight interviewees represent both a rea- developed by Corbin and Strauss.30 Common themes
sonable percentage and a representative sample of related to the challenges, successes, and general man-
the types of organizations involved in the response. agement of the evacuee support after Hurricane Maria
In this case, saturation was reached with a smaller were identified from the coding. (A summary of the
sample size, where the population of interest is interview methods is shown in Table 2.)
small and required coordination at higher levels of
analysis. The responders involved in these efforts Content analysis of government
can answer interview questions with similar mental documents and websites
models; thus, there is high information power and
saturation.29 Interview data were supplemented by a content
analysis of the Florida state government documents

Special Issue on Puerto Rico

126 Journal of Emergency Management
Vol. 19, No. 8

Table 1. Interviewee information*

Interviewee Organization type Response role*

Interviewee 1 Emergency management Coordinate and lead the response with public safety officials, elected
officials, nonprofit organizations, and government agencies

Interviewee 2 Local government agency Provide evacuees with information, resources, services, and connect
them with governmental agencies, community organizations

Interviewee 3 Nonprofit organization Provide and coordinate food and meals for disaster victims

Interviewee 4 Nonprofit organization Respond to evacuee needs as identified through the Disaster Distress
Helpline, connect them with support services

Interviewee 5 Nonprofit organization Coordinate evacuee services, provide resources, services, and connect
them with governmental agencies, community organizations

Interviewee 6 Nonprofit organization Provide evacuees with information, resources, services, and connect
them with governmental agencies, community organizations

Interviewee 7 Faith-based organization Executive Director: Responsible for overall operations, coordinated
with national and regional offices EOCs and other VOADs (Voluntary
Organizations Active in Disasters). Disaster response is part of the larger
mission of charitable work

Interviewee 8 Faith based organization Area leader for Florida counties: religious leader is the primary work,
with disaster response a part of service, along with housing, and support
services for people in need. Coordinated with agencies and offered
spiritual and emotional care along with clothing and shelter

*All of the interviewees worked together with the displaced Puerto Rican evacuees through the MARC center in Orlando.

and website information related to Hurricane Maria. information about services for Puerto Ricans who
The relevant written documents and website content relocated to the area and was the basis for the con-
were identified using the search term “Hurricane tent analysis.
Maria” at the Florida Division of Emergency
Management (FDEM) website (https://www.flori RESULTS
dadisaster.org/info/maria/), City of Orlando website
(http://www.cityoforlando.net), and by searching for The coding of the interview data and the content
the Executive Orders issued by (then) Governor Scott analysis of the governmental documents and websites
related to Hurricane Maria, from the Governor’s web- revealed five overarching common themes from the
site (at https://www.flgov.com/all-executive-orders/). interview data about how to manage and support
A content analysis of the FDEM website was con- a large displaced population after a major disaster
ducted, where the statewide response to Hurricane event. The emergent themes are related, but distinct
Maria, including news reports and details about enough to treat separately in the results. Table 3 sum-
the resource centers in Orlando and Miami, as well marizes the topics discussed by each interviewee, and
as links and information about VOADs and non- the emergent themes are discussed in detail in the
profit organizations. The database searches from the sections that follow.
Governor’s website yielded eight Executive Orders
related to Hurricane Maria issued by Governor Scott. Previous experience with disasters
The City of Orlando website likewise offered detailed Florida is experienced with hurricanes and

storms and works statewide for effective disaster

Special Issue on Puerto Rico

Journal of Emergency Management 127
Vol. 19, No. 8

Table 2. Interview methods

Interviewee Status Source Length Recording Format
Semistructured
Interviewee 1 In person, May 16, 2018 Snowball sample 1 hour Audio recording Semistructured
Semistructured
Interviewee 2 In person, May 16, 2018 Sample frame 55 minutes Written notes Semistructured
Semistructured
Interviewee 3 In person, May 16, 2018 Sample frame 1 hour 30 minutes Audio recording Semistructured
Semistructured
Interviewee 4 In person, May 17, 2018 Snowball sample 40 minutes Audio recording Semistructured

Interviewee 5 In person, May 17, 2018 Sample frame 1 hour 20 minutes Audio recording

Interviewee 6 In person, May 17, 2018 Snowball sample 1 hour 20 minutes Audio recording

Interviewee 7 Telephone, June 7, 2018 Sample frame 21 minutes Audio recording

Interviewee 8 In person, May 17, 2018 Sample frame 37 minutes Audio recording

management. Particularly in Central Florida, which they are often not faced with the same bureaucratic
tends to be more protected from the most severe hur- hurdles and limitations that governmental agencies
ricane impacts, the emergency management network have to navigate and, thus, are able to be more flex-
has routinely served as a staging and coordination ible and adaptable to changing conditions and needs
site to support other parts of the state that experience (Interviewees 4, 5, and 6, Interviews, May 17, 2018).
more direct storm impacts (Interviewee 1, Interview, Experience with the same type of disaster (hurricane)
May 16, 2018). Interviewee 3 described this experi- and having established protocols and plans in place
ence, noting that their facilities are used to pre-­ before the event made it easier to support evacuees
position supplies about a week prior to landfall, as arriving from the island after Hurricane Maria.
well as to use as a staging ground to respond after a
storm. This experience helps streamline the response: Previous experience with disasters also facilitated
“The state operation, emergency operation center is the improvisation and formation of new partnerships
right here in Orlando- their warehouse where they in real time during response. Interviewee 4 offered
keep all the MREs and the water, and all of that emer- the example of establishing a new partnership with
gency stuff, so, we start making phone calls ahead of nearby hotels, which provided airport shuttles, and
time, and say, ‘Hey let’s put five trailer loads of water using Uber as creative solutions for transportation
and five trailer loads of MREs in this building so that needs from the airport. It was described as chaotic,
the morning after we can immediately begin distrib- but innovation was possible because of the previous
uting and not be filling out FEMA forms, reports, all experience with disasters (Interviewee 4, Interview,
that’.” (Interviewee 3, Interview, May 16, 2018). May 17, 2018). Another interviewee, leading a faith-
based organization with extensive disaster response
Interviewees indicated that previous experience experience, noted that they had recently responded
with disaster management was a crucial factor in to Hurricanes Harvey and Irma and were mobile
effectively managing the influx of displaced Puerto and ready to respond to Hurricane Maria evacuees
Ricans (Interviewees 2, 3, 4, and 8). They noted that as well because of their experience (Interviewee 8,
they had learned from previous experiences with dis- Interview, May 17, 2018). Previous disaster experi-
asters and adjusted their approaches, which resulted ence was important partially because it provided an
in more effective coordination across agencies and geo- opportunity to establish and develop strong relation-
graphical areas when responding to Hurricane Maria ships across the disaster community, both regionally
evacuees. The nonprofit organizations mentioned that and nationally.

Special Issue on Puerto Rico

128 Journal of Emergency Management
Vol. 19, No. 8

Table 3. Interviewee topic themes across the state. All the interviewees said that hav-
and discussion frequency* ing a MARC at the airport was an incredibly success-
ful interagency and intergovernmental coordination
Previous disaster experience is important (14) strategy, as it made it faster and easier to coordinate
and innovate as needed. For example, having FEMA
Established relationships and partnerships are important; on site allowed for FEMA claims to be created and
coordination efforts with partners, agencies, and submitted and updated rapidly when evacuees first
organizations; working together informally and formally arrived at the airport (Interviewee 1, Interview, May
in the community (33) 16, 2018).

Importance of effective government leadership (6) Nonprofit organizations were able to successfully
coordinate with their local and broader networks to
Shelter and housing challenges (short and long term; respond quickly and efficiently to the evacuees, while
FEMA policies; building codes) (21) adapting to unexpected circumstances. For exam-
ple, disaster responders realized quickly that they
Vulnerabilities affected evacuee resources needed (acute needed Spanish speaking personnel at the airport
services such as medical care, basic necessities like food center to communicate with evacuees. There was a
and clothing; Spanish speaking responders, elderly and need to quickly find people to help with translations
sick evacuees) (16) (Interviewee 4, Interview, May 17, 2018). Another
nonprofit leader described the collaboration that the
Intergovernmental coordination challenges across CEO initiated between multiple agencies by using
agencies and nonprofits, eg, FEMA/state/local (18) previously established relationships to quickly coor-
dinate response at the airport; they were able to set
Importance of advanced preparation and planning (13) everything up, including transporting supplies, in
only 2 days. Each agency was functioning autono-
Improvisation was a key feature in response, eg, using mously while still working collaboratively with other
hotels, Uber, and other unplanned resources (4) agencies (Interviewee 5, Interview, May 17, 2018).
The challenge of coordination across various agencies
Providing mental health services/spiritual counsel was and jurisdictions was made easier by the established
important in response, eg, evacuees feeling suicidal and relationships across a national nonprofit network,
anxiety (3) coordination with local and state level emergency
management agencies, and the state level agencies
Nonprofit organizations have greater flexibility than (Interviewee 3, Interview, May 16, 2018).
official agencies in response (fewer regulatory ­
constraints) (2) Likewise, faith-based organization leaders ech-
oed these sentiments, noting a highly coordinated
Schools strained by influx of students (2) partnership with other faith-based partners and
nonprofits across the state and nation (Interviewee
*Frequency of mentions by interviewees is denoted 7, Interview, June 7, 2018). They described the impor-
in parentheses after each topic. Some topics were tance of interconnected networks and collaboration as
mentioned fewer times but were identified as high effective because of the pre-established relationships
importance and involved a lengthy treatment, which is that are formalized into the disaster response plans.
discussed in the results section. Several organizations have a designated, clearly
defined role within the emergency operations plan.
Established relationships and intergovernmental They coordinate with the American Red Cross, a for-
and interagency coordination-MARC mal FEMA partner responsible for providing shelter
and serving food, and faith-based organizations who
Experience with disasters was a necessary but
not sufficient condition for the effective manage-
ment of Hurricane Maria evacuees’ needs. Disaster
experience provided an opportunity to establish and
develop strong relationships across the disaster
community, both regionally and nationally, which
interviewees consistently mentioned as the most
critical factors for success and as challenges during
the response. Many organizations had previous expe-
rience coordinating their efforts in the region and

Special Issue on Puerto Rico

Journal of Emergency Management 129
Vol. 19, No. 8

serve food and offer spiritual and emotional support. Effective government leadership is essential
The planned collaboration with the city and county Florida’s state and local governmental leaders
government is a great asset (Interviewee 8, Interview,
May 17, 2018). “When it comes to disasters, we don’t supported displaced evacuees from the island with
have to get approval to go into action; we can do it strong leadership, resource mobilization, and coordi-
immediately. And then there’s coordination region- nation. According to the interviewees and confirmed
ally or nationally or internationally, but the local unit by the content analysis from the government web-
can go into action immediately without any further sites, the local emergency management agencies in
authority. And that’s one thing that I like about [our Orlando, Mayor Buddy Dyer, and Governor Rick Scott
organization] because we’re all given the authority to worked together along with Puerto Rico’s Governor
do that immediately. While others may be still trying Ricardo Rosello to lead and fund efforts to help
to organize people, but our local people are allowed evacuees arriving to Florida from the island after
to do that.” (Interviewee 8, Interview, May 17, 2018). Hurricane Maria hit (Interviewee 1, Interview, May
16, 2018).
Assessments about working with FEMA were
varied across interviewee accounts, with most of The City of Orlando has a long-standing relation-
them finding coordination with FEMA challenging. ship with Puerto Rico. The Hispanic Office for Local
One nonprofit leader said that while collaborat- Assistance (HOLA) has provided information and
ing with FEMA presented obstacles initially, they referrals to Spanish-speaking residents and newcom-
developed partnerships to help support the families ers to Central Florida since 2004. This office connects
during the response and early in the recovery pro- residents with over 100 government and community
cess (Interviewee 6, Interview, May 17, 2018). In organizations that offer services such as employment,
contrast, a faith-based organization noted that there housing, healthcare, and other needs.31 When visiting
was considerable confusion in communication from the HOLA outreach center, families and individuals
FEMA about deadlines for short-term housing sup- meet with staff members who connect them with
port (Interviewee 7, Interview, June 7, 2018). local resources. They have access to information,
computers, and bilingual staff to assist them as
Another nonprofit leader described frustration needed. With a large Puerto Rican population already
coordinating with FEMA: “They’ve got a tremen- present in Florida, as well as many native Spanish
dous amount of money, and a tremendous amount of speakers, Florida was a likely state to be a choice for
will to do something. They typically spend way too evacuees. The region has family support networks
much money doing what they’re doing and then they and established resources and support for Puerto
really don’t have a good distribution network. So, in Ricans, but the community still faces challenges in
the times we could get synced up with them and we discrimination and job opportunities.32 Interviewees
could get them to send their food through here, it mentioned HOLA and the important role they played
worked out well. But you’re still dealing with FEMA in connecting the evacuees with agency resources
who wants to send 10 trucks at a time. Nobody’s got and coordinating with the 40 other agencies at the
an appointment. They don’t know where they are. MARC, and that they had relationships with and
They don’t know what they’ve got on there. It’s just experience working with them during past disasters
a big, convoluted mess…The metaphor I often use is (Interviewees 2, 4, 5, and 6).
a monkey with a shotgun.” (Interviewee 3, Interview,
May 16, 2018). Overall, nonprofits and faith-based Governor (now US Senator) Rick Scott established
organizations had stronger relationships with local a strong partnership with former Governor Ricardo
and state partners, while governmental agencies at Rosello in Puerto Rico, working together to prepare
the state and local levels worked closely with FEMA, in advance of the storm and continuing to coordinate
as expected. The long-term relationships between and during the response and recovery to offer support
across organizations improved response efforts. for Puerto Ricans displaced in Florida. A review of
executive orders issued related to Hurricane Maria

Special Issue on Puerto Rico

130 Journal of Emergency Management
Vol. 19, No. 8

by Governor Scott indicated that he declared a state local leaderships in managing intergovernmental
of emergency on October 2, 2017.33 Governor Scott relations with FEMA.
traveled to Puerto Rico in advance of the storm and vis-
ited the island 11 times after Hurricane Maria struck, Shelter and housing for evacuees
continuing to advocate for displaced Puerto Ricans.34 A consistent theme emerging from the interviews

The content analysis of the government websites was that helping evacuees find temporary shelter and
indicated a proactive partnership between the gov- longer-term housing options was the most significant
ernors of Puerto Rico and Florida. On September 26, challenge of the response. Interviewees all discussed
2017, Governor Scott visited Puerto Rico, and upon housing challenges when asked what the greatest
returning, he mobilized resources to support Puerto challenges were in supporting the evacuees. Housing
Rico’s evacuees. State level resources were made avail- challenges were not only mentioned with the greatest
able to support the evacuation and evacuees coming frequency by the interviewees, with 21 separate men-
in from Puerto Rico. Initially, three MARC locations tions (Table 3), but were discussed at great length by
were established in Florida, at Orlando International the interviewees. Others expressed some frustration
Airport, Miami International Airport, and the Sea with FEMA and the way the political leadership coor-
Port at Fort Lauderdale. Interviewees (especially 1, 5, dinated with them. For example, interviewees indi-
and 6) discussed the importance of the MARC centers. cated that it was challenging for them to coordinate
For example, Interviewee 1 described the center: “We with FEMA with housing plans, primarily because of
established what we call a multiagency resource center the decision to continue transitional sheltering assis-
at the Orlando International Airport, and Miami, and tance (TSA) funding rather than shifting to a longer-
at the seaport. Ours [Orlando Airport] was the biggest term option of Emergency Housing Assistance from
one. We had over 40 agencies, governmental, and non- FEMA. One interviewee discussed the possibility that
profit, faith-based that were there. We initially started the TSA program remained in place because the gov-
operating 12, 14, 16-hour days… And the intent of the ernor of Puerto Rico was concerned about people not
state was to provide a one-stop shop for the evacuees returning, and speculated that was why they made
coming in from Puerto Rico where they can get refer- the decision to authorize TSA approvals for 15 days at
rals or information, or services. The governor estab- a time. That process was stressful for evacuees, who
lished a host state agreement with the government would often be waiting until the last day of their hous-
of Puerto Rico, and we coordinated and supported the ing support to find out whether they could remain in
state in that effort. Us, being the local government, their current housing situation.
we were supporting the state effort.” (Interviewee 1,
Interview, May 16, 2018). Likewise, interviewee 6 men- Local emergency management and local agencies
tioned that the center with all the support agencies recommended through the Office of the Governor
together in one location was funded by the governor in Florida that a request should be made to use
(Interviewee 6, Interview, May 17, 2018). the FEMA Emergency Housing Assistance Program
instead of the TSA program. The rationale for this
As indicated by the website document analysis described the following: “What this program does, is if
and confirmed by interviewees, Governor Scott and there’s something more long term, it allows for fam-
Mayor Buddy Dyer worked together and with multiple ily members to get assistance from FEMA to rent out
agencies and emergency management offices to coor- apartments or existing houses at the fair market rate.
dinate relief and support efforts. Their coordination So, what they would do is they would pay for a fam-
with the Puerto Rico Federal Affairs Administration ily to stay at an apartment or a house to anywhere
and United for Puerto Rico, among other organiza- between 12 to 18 months, which is a longer window.
tions, was an important part of their successful Then the person can be re-adjusted, reallocated and
response (Interviewee 2, Interview, May 16, 2018). have a stable platform for the family unit, where they
Interviewees highlighted the importance of state and can bring it to the school, get a job and move around

Special Issue on Puerto Rico

Journal of Emergency Management 131
Vol. 19, No. 8

in a more secure environment. TSA doesn’t afford I remember this one gentleman came in, dialysis,
that. TSA is a very quick and dirty get you to a hotel missed like three or four sessions, and he was as
and then we’ll figure out what needs to be done. It’s y­ ellow as this paper; that’s how he was. He looked like
not, it doesn’t really provide a long-term solution. a Simpson character. That’s how yellow he was. And we
But that program was not requested.” (Interviewee 1, knew he was bad because he wasn’t a happy camper.
Interview, May 16, 2018). He was in very bad condition, but we were able to
get him to a hospital.” (Interviewee 1, Interview, May
Vulnerabilities affect evacuee resources needed 16, 2018). These are examples of the unique vulner-
Second only to housing related challenges in abilities the MARC center helped displaced evacuees
contend with after they evacuated the island.
importance, interviewees discussed the urgency to
address the needs arising from vulnerabilities of the DISCUSSION
evacuees arriving from Puerto Rico. Interviewees
mentioned vulnerabilities 16 times, nearly as often The displacement of Puerto Rican residents after
as housing challenges (Table 3). Vulnerabilities were Hurricane Maria provides an important opportunity
primarily related to a lack of financial and other to increase our broader understanding of disaster
resources, medical needs, trauma, and mental health response for displaced populations. The Orlando met-
needs. Recognition and familiarity with the pre- ropolitan region was an ideal site selection because
existing vulnerabilities the evacuees were contending of the MARC center, as well as the prominence of
with and understanding the new and overwhelming Orlando as a destination for evacuees. Although
challenges they were navigating were crucially impor- Puerto Rican residents were displaced throughout the
tant to effectively supporting them. Interviewees indi- country, Florida received the most Puerto Rican resi-
cated that the MARC center triaged evacuee arrivals dents, with the Orlando metropolitan area being the
according to the most urgent needs, while thinking top metro area for FEMA assistance claims outside of
about how to support them as they transitioned to Puerto Rico after Hurricane Maria.35
longer-term needs. Nonprofit organizations supplied
essential necessities such as clothing and other goods, This study confirms the conventional wisdom
and housing vouchers to evacuees: “Needs ranged about the importance of coordination and collabora-
from the youngest demographic to the oldest. Having tion, and planning for effective disaster response, and
no electricity puts the medically vulnerable at risk, offers new insights about the ways in which effective
such as those with cancer. That’s what we were doing. response can be implemented at centralized evacu-
So, we had United Way, Red Cross, and Salvation ation sites to support displacement after disasters.
Army. So just imagine, people were coming in that Intergovernmental coordination in disaster response
lost everything, they flooded out, they only had what is frequently identified by scholars as challenging and
they had on. So, United Way, Goodwill were provid- political in a federalist system.36 This study confirms
ing vouchers for clothing. We had moms that were that this remains an ongoing challenge, particularly
coming in that needed baby formula for their kids.” coordinating with FEMA. However, important new
(Interviewee 1, Interview, May 16, 2018). insight emerged from the interviewees about what
worked well and mitigated the challenges of the inter-
There were serious and urgent health vulner- governmental coordination. The pre-existing relation-
abilities: “We had a bunch of folks that came in, that ships between the Governors in Florida and Puerto
because of the situation in Puerto Rico, they couldn’t Rico helped develop support for evacuees from the
get chemotherapy. People were getting on the plane island before the hurricane made landfall, and for
in Puerto Rico and landing already in need of dialy- many months after the event. Likewise, the City of
sis, were diabetic, and needed ambulances to hospi- Orlando and the metropolitan area worked together to
tals. The Department of Health was at the resource coordinate the response, which was particularly effec-
center to screen people and route them to hospitals. tive through the establishment of the MARC center,

Special Issue on Puerto Rico

132 Journal of Emergency Management
Vol. 19, No. 8

which included a coalition of more than 40 agencies to The interviewees emphasized the need to be adapt-
meet evacuees at the airport and assist them. able to new and changing needs, which are always fac-
tors in responding to disasters. Interviewees described
An important finding from this research is that situations where they had to be creative and engage in
the MARC centers are successful and important improvisation, using hotel booking websites for secur-
for responding to a large influx of displaced evacu- ing temporary shelter and Uber to offer transportation
ees after a disaster. The MARC model has been for evacuees. From helping the family who needs a
implemented more frequently in recent years after car seat for their baby to getting someone who needs
Hurricane Maria, and important updates to the plan- urgent medical attention to the hospital, disaster
ning documents have been integrated based on feed- response requires creative thinking. This is consistent
back from Florida and others who have experience with prior research on the importance of improvisation,
with them.37 Interviewees recommended that other which is essential for effective disaster response.40
jurisdictions consider developing a plan for a central-
ized disaster center like the MARC established after This case study provides a foundational point for
Hurricane Maria in Orlando and Miami. This was further research. While a case study offers in-depth and
a clear success in providing a centralized resource detailed accounts of the coordination and response after
center for support for evacuees, making it easier for Hurricane Maria, it lacks generalizability. Future studies
the agencies to coordinate with each other, being in should consider other displaced evacuee locations where
proximity. Since responding to Hurricane Maria, the MARC centers are established, and employ mixed meth-
disaster management community has developed a ods research designs, perhaps adding survey data and
family assistance plan to activate in case of another additional interviews, among other methods and data to
emergency or disaster in the Orlando and Central contribute to this area of research. Disaster events will
Florida regions (Interviewee 1, May 16, 2018). This continue to increase in frequency and scope, and mutu-
type of disaster response innovation is an example ally supportive relationships across governmental and
of a way other regions and agencies might consider nonprofit agencies will remain centrally important. This
planning for an influx of evacuees after disasters. research offers a starting point and insights into success-
ful approaches to supporting evacuees and increasing
The most universal and prominent challenges cooperation and effectiveness in response, particularly
mentioned by interviewees in responding to the with improvisation and establishing MARC centers, as
evacuees were related to finding temporary, short-, well as identifying ongoing challenges such as housing
and longer-term housing options for evacuees. The and intergovernmental coordination.
Orlando metro region was facing an affordable hous-
ing shortage before Hurricane Maria hit and was Tanya Buhler Corbin, PhD, Department Chair and Associate Professor,
ranked second among metropolitan areas with the Security and Emergency Services, Embry-Riddle Aeronautical
least affordable housing units in the United States University Worldwide, Daytona Beach, Florida. ORCID: https://orcid.
in 2017.38 Securing financial support, vouchers, and org/0000-0003-1308-3637.
physical housing units for the evacuees was difficult,
and interviewees expressed great concern that future REFERENCES
planning efforts should address this challenge. This
is consistent with disaster housing research that 1. Coto D: Maria destroys homes, triggers flooding in Puerto
notes the need for more research in this area, as well Rico. AP News. September 21, 2017. Available at https://apnews.
as highlighting the difficulties in finding housing for com.
the displaced.39 Future research should continue to 2. FEMA: Hurricane season FEMA after-action report. FEMA.
expand knowledge about the ongoing importance of 2017. Available at https://www.fema.gov/sites/default/files/2020-
housing and shelter challenges after a disaster, and 08/fema_hurricane-season-after-action-report_2017.pdf. Accessed
the prominence of these concerns when large popula- February 1, 2021.
tions have been displaced. 3. Young R: Blackout in Puerto Rico [Video File]. Boston, MA: PBS,
2018.
4. Cortez J: Puerto Rico: Hurricane maria and the promise of dis-
posability. Capital Nat Social. 2018; 29(3): 1-8.

Special Issue on Puerto Rico

Journal of Emergency Management 133
Vol. 19, No. 8

5. Abel JR, Deitz R: The causes and consequences of Puerto Rico’s 23. McAdam J: Displacing evacuations: A blind spot in disaster
declining population. Fed Reserve Bank New York Curr Issues Econ displacement research. Refugee Surv Q. 2020; 39(2): 583-590, 585.
Financ. 2014; 4: 1-8. 24. Yin R: Case Study Research and Applications, 6th ed. Thousand
6. Puerto Rico Research Hub: Community. Available at https://sci- Oaks, CA: Sage Publications.
ences.ucf.edu/puerto-rico-hub. Accessed June 9, 2021. 25. Meléndez E, Hinojosa J: Estimates of Post-Hurricane Maria
7. Esnard A, Sapat AE: Displaced by Disaster. New York: Taylor & Exodus from Puerto Rico. Centro Research Brief. New York:
Francis, 2014. CUNY, 2017: 1.
8. Fordham M: Disaster and development research and practice: 26. Hinojosa J, Meléndez E: Puerto Rican Exodus: One Year Since
A necessary eclecticism? In Rodriguez H, Quarantelli EL, Dynes Hurricane Maria. Centro Research Brief. New York: CUNY, 2018:
R (eds.): Handbook of Disaster Research. New York, NY: Springer, 5. Available at https://centropr.hunter.cuny.edu/sites/default/files/
2007: 335-346. RB2018-05_SEPT2018%20%281%29.pdf. Accessed February 1,
9. Smith G, Wenger D: Sustainable disaster recovery: 2021.
Operationalizing an existing agenda. In Rodriguez H, Quarantelli 27. Phillips B: Qualitative Disaster Research: Understanding
EL, Dynes R (eds.): Handbook of Disaster Research. New York, NY: Qualitative Research. New York, NY: Oxford University Press, 2014.
Springer, 2007: 234-257. 28. Guest G, Bunce A, Johnson L: How many interviews are
10. Belcher JC, Bates FL: Aftermath of natural disasters: Coping enough? An experiment with data saturation and variability. Field
through residential mobility. Disasters. 1983; 7(2): 118-128. Methods. 2006; 18(1): 59-82. DOI: 10.1177/1525822X05279903.
11. Hori M, Schafer MJ, Bowman DJ: Displacement dynamics in 29. Malterud K, Siersma VD, Guassora AD: Sample size in
Southern Louisiana after hurricanes Katrina and Rita. Popul Res qualitative interview studies: Guided by information power. Qual
Policy Rev. 2009; 28(1): 45-65. Health Res. 2016; 26(13): 1753-1760. DOI: 10.1177/104973231561
12. US Senate: Ad Hoc Subcommittee on Disaster Recovery. Far 7444.
from Home: Deficiencies in Federal Disaster Housing Assistance 30. Corbin J, Strauss A: Basics of qualitative research. In
after Hurricanes Katrina and Rita and Recommendations for Techniques and Procedures for Developing Grounded Theory,
Improvement. Washington, DC: Committee on Homeland Security 3rd ed. Thousand Oaks, CA: SAGE Publications, 2008. DOI:
and Governmental Affairs, US Senate, 2009: Senate Print 111-117. 10.4135/9781452230153.
13. Weber L, Peek L (eds.): Displaced: Life in the Katrina Diaspora. 31. Hispanic Office for Local Assistance: Available at http://www.
Austin, TX: University of Texas Press, 2012: 1-20. cityoforlando.net/hola. Accessed February 1, 2021.
14. Sapat A, Esnard A (eds.): Coming Home after Disaster: Multiple 32. Aranda E, Rivera FI: Puerto Rican families in Central Florida:
Dimensions of Housing Recovery. Boca Raton, FL: CRC Press, 2017; Prejudice, discrimination, and their implications for successful
Peacock WG, Dash N, Zhang Y: Post-disaster sheltering temporary integration. Women Gender Families Color. 2016; 4(1): 57-85.
housing and permanent housing recovery. In Rodríguez H, Donner 33. Office of the Governor: Emergency management-Hurricane
W, Trainor JE (eds.): Handbook of Disaster Research, 2nd ed. Maria. Executive Order Number 17-259, Issued October 2, 2017.
Madison, WI: Springer, 2018: 569-594. 34. Derby K: Rick Scott visits Puerto Rico to see how island is recov-
15. Peacock WG, Dash N, Zhang Y, et al.: Post-disaster sheltering, ering from earthquakes. Florida Daily. January 13, 2020. Available
temporary housing and permanent housing recovery. In Rodríguez at https://www.floridadaily.com. Accessed January 31, 2020.
H, Donner W, Trainor JE (eds.): Handbook of Disaster Research, 2nd 35. Sutter J, Hernandez S: ‘Exodus’ from Puerto Rico: A visual
ed. Madison, WI: Springer, 2018: 569-594. guide. Cnn. February 21, 2018. Available at https://www.cnn.com.
16. Kroll-Smith S, Baxter V, Jenkins P: Left to Chance: Hurricane Accessed January 25, 2021.
Katrina and the Tale of Two New Orleans Neighborhoods. Austin, 36. Kapucu N, Garayev V: Collaborative decision-making in emer-
TX: University of Texas Press, 2015. gency and disaster management. Int J Public Adm. 2011; 34(6):
17. Corbin TB: Lessons learned after disaster? Emergency shel- 366-375. Kapucu N, Arslan T, Collins ML: Examining intergovern-
ter and housing policy after Hurricane Katrina. In Disaster mental and interorganizational response to catastrophic disasters:
Management and Human Health Risk V. Johannesburg: WIT Toward a network-centered approach. Adm Soc. 2010; 42(2): 222-
University Press, 2017: 1-8. 247. Schneider S: Who’s to blame? (mis) perceptions of the intergov-
18. Smith G: Pre- and post-disaster conditions, their implications, ernmental response to disasters. Publius. 2008; 38(fall 4): 715-738.
and the role of planning for housing recovery. In Sapat A, Esnard DOI: 10.1093/publius/pjn019; Gerber BJ: Disaster management in
A (eds.): Coming Home after Disaster: Multiple Dimensions of the United States: Examining key political and policy challenges.
Housing Recovery. Boca Raton, FL: CRC Press, 2017: 277-292. Policy Stud J. 2007; 35(2): 227-238.
19. Kendra JM, Wachtendorf T: Community innovation and disas- 37. National Mass Care Strategy: MARC planning resource.
ters. In Rodriguez H, Quarantelli E, Dynes R (eds.): Handbook of Available at https://nationalmasscarestrategy.org/multi-agency-
Disaster Research. New York: Springer, 2006: 316-334. resource-center-marc-planning-resource/. Accessed June 10,
20. Multi-Agency Resource Planning Center: Available at https:// 2021.
nationalmasscarestrategy.org/multiagency_resource_center_ 38. National Low Income Housing Coalition: The gap: A short-
planning_resource_jt_v-1-0_2017_06_12/. Accessed June 1, 2021. age of affordable homes. 2018. Available at http://nlihc.org/sites/
21. Multi-Agency Resource Planning Center: Available at https:// default/files/gap/Gap-Report_2018.pdf. Accessed June 5, 2021.
nationalmasscarestrategy.org/multiagency_resource_center_ plan- 39. See, for example, the edited volume by Sapat A, Esnard A (eds.):
ning_resource_jt_v-1-0_2017_06_12/, p. 7. Accessed June 1, 2021. Coming Home After Disaster: Multiple Dimensions of Housing
22. Orlando multi-agency resource center to serve those impacted Recovery. Boca Raton, FL: CRC Press, 2017.
by hurricane maria through end of the month. Available at https:// 40. Kendra JM, Wachtendorf T: Improvisation, creativity, and the
www.floridadisaster.org/news-media/news/orlando-multi-agency- art of emergency management. In Durmaz H, Sevinc B, Yayla AS
resource-center-to-serve-those-impacted-by-hurricane-maria/. (eds.): Understanding and Responding to Terrorism. Washington,
Accessed June 2, 2021. DC: IOS Press, 2007: 324-335.

Special Issue on Puerto Rico

134 Journal of Emergency Management
Vol. 19, No. 8

JEM Hurricane effects, mitigation, and preparedness in the
Caribbean: Perspectives on high importance–low prevalence

practices from agricultural advisors

Nora L. Álvarez-Berríos, PhD
Sarah S. Wiener, MS

Kathleen A. McGinley, PhD
Angela B. Lindsey, PhD
William A. Gould, PhD

ABSTRACT practices help mitigate hurricane effects, improved
interagency coordination for hurricane response and
Frequent hurricanes affect agricultural produc- preparedness, and integrated educational campaigns
tivity, food security, economic security, and human with advisors and land managers.
wellbeing in the Caribbean islands. We assessed recent
hurricane effects on the agricultural sector, and the Key words: disasters, risks, tropical agriculture,
challenges faced by farmers, foresters, and advisors emergency management, hurricane resilience
related to hurricane preparedness and recovery in
Puerto Rico and US Virgin Islands (USVI). We used INTRODUCTION
qualitative and quantitative survey methods to com-
pile perspectives from agricultural advisors related Agricultural economies and food security in the
to hurricane effects on farmlands, preparedness and Caribbean islands are vulnerable to extreme weather
recovery measures taken by land managers, and the events due to their location in the Atlantic Hurricane
needs regarding preparing for and responding to Belt, geographic isolation from continental-based
future hurricanes. Survey responses from over 200 resources, limited geographic and economic scale,
advisors at eight institutions provided insight into the and dependence on imported goods. The Caribbean
most devastating hurricane effects across farmlands, basin is exposed to an average of six hurricanes per
including issues related to power outages, commu- year,1 and the Eastern Caribbean is ranked as one
nication, road access, and fallen trees. Our results of the world’s most disaster-prone regions due to its
highlight strategies considered critical for hurricane high incidence of hurricanes.2 Furthermore, climate
preparedness and recovery but not prevalent in appli- models predict increases in hurricane intensity in
cation among land managers. Advisors’ perceptions the region with rising global temperatures. Although
suggested that farmers and ranchers apply essential models indicate that the total number of hurricanes
recovery practices, but critical short-term prepared- will remain similar to past patterns, significant
ness practices are limited, and long-term preparedness increases in the severity of hurricanes and associ-
practices are uncommon. Advisors also indicated the ated rainfall are projected for the Caribbean basin.3
need for more training and educational resources to An important mechanism to reduce vulnerability is
improve hurricane recovery response. We conclude that to learn from past experiences and incorporate those
better planning to minimize the vulnerability to future lessons into planning, preparation, and supporting
hurricanes can be achieved through an increased mitigation actions at different scales.
understanding of how preparedness and recovery
Recurrent hurricanes have shaped Caribbean
social and ecological landscapes. Repeated damages

DOI:10.5055/jem.0585 Special Issue on Puerto Rico
Journal of Emergency Management
Vol. 19, No. 8 135

to crops, livestock, forests, and infrastructure have led alleviate the effects of hurricanes on agricultural
to losses in the immediate aftermath of a hurricane, systems in the region. Exceptions include studies on
with persistent effects in some sectors.4,5 The sever- farmers’ experiences and challenges after hurricane
ity of short- and long-term effects of hurricanes is María in Puerto Rico by Rodriguez and Niles,10 the
related to vulnerabilities to the effects of high winds response of coffee farmers to Hurricane María in
and heavy rains,6 and also to constraints in executing Puerto Rico by Perfecto et al.,11 and on the oppor-
effective planning, response, and recovery intended to tunities and challenges for hurricane resilience on
reduce the secondary effects of hurricanes. Secondary agricultural and forest land in the US Southeast
effects include extended losses of power, transporta- and Caribbean by Wiener et al.12 Each of these stud-
tion infrastructure, energy, food distribution infra- ies demonstrate the importance of systematically
structure, and emergency medical support.5 Social learning from extreme climate events in order to bet-
and economic factors associated with hurricane vul- ter inform mitigation and adaptation practices and
nerability in the agriculture and forestry sectors in policies.
Puerto Rico and the US Virgin Islands (USVI) include
high dependence on government services and public Agricultural advisors represent a group of profes-
goods such as electricity, water, and transportation. sionals who, given their direct connection with land-
Reducing risk therefore requires an understanding owners, act as an information intermediary between
of vulnerabilities to primary and secondary effects the scientific, public-policy, and user communities.12,13
of hurricanes that can be addressed in the planning, Observations provided by agricultural advisors serve
response, and recovery phases. as a reflection of the landowners’ perception of climate
challenges, and their testimonies are deemed relevant
The 2017 hurricane season represented extraor- to facilitate means for climate adaptation.14,15 A major
dinary challenges for the US Caribbean. Puerto Rico strength of advisors comes from their on-the-ground
and USVI were hit by hurricane Irma (Category 5 on relationships with the farmers they serve and a close-
the Saffir-Simpson Hurricane Wind Scale (SSWS)) knit understanding of social contexts. Additionally,
on September 6 and hurricane María (Category 5 advisors are part of the agencies they work for and
SSWS for USVI and Category 4 SSWS for Puerto bring with them the capacity and limitations of their
Rico) on September 20. The social, ecological, and eco- agencies in terms of hurricane preparation, response,
nomic impacts were unprecedented in recent history. and recovery. Agricultural advisors, however, also may
Economic losses surpassed $43 billion in Puerto Rico be affected personally by the particular challenges
and $7.5 billion in USVI.7,8 In Puerto Rico, losses in that hurricanes bring, which can make it difficult for
agricultural production and agricultural infrastruc- them to serve their clients in the immediate after-
ture surpassed $2 billion.3 No official estimates of math of a hurricane or other disaster. Understanding
losses in USVI agricultural production were reported, these challenges is a first step toward supporting
although damages were similar in scope to those of agricultural advisors as they participate in hurricane
Puerto Rico.3 The lack of agricultural production and recovery, and better prepare for hurricane events.
ability to bring products to markets following hur-
ricanes Irma and María, combined with restrictions In this study, we analyze the challenges posed
on food imports and logistical challenges at ports by hurricanes Irma and María on the US Caribbean
and distribution centers, exacerbated food insecurity agriculture sector through a systematic assessment of
across the islands.9 the experiences and perceptions of agricultural advi-
sors in this region. We surveyed agricultural advisors
Despite the enormous impact of hurricanes on of eight public boundary organizations in Puerto Rico
agriculture and forestry in the US Caribbean, there and USVI to gather information on their perceptions
has been relatively limited development of quantita- of hurricane effects on the sectors they serve, the rela-
tive or qualitative information on the prevalence and tive importance of existing strategies for hurricane
value of mitigation and adaptation strategies that preparedness and recovery, and the adoption of these

Special Issue on Puerto Rico

136 Journal of Emergency Management
Vol. 19, No. 8

strategies by farmers. Moreover, this research evalu- a total of 2,380 ha of farmlands, from 219 farms, with
ates the prevalence or gaps in the implementation farms being an average of 11 ha.16 Approximately, 22
of critical practices to mitigate hurricane effects on percent of Puerto Rico’s and 7 percent of USVI’s land
farmlands. areas are occupied by farms (croplands, pastures, and
grazing lands).16 New initiatives by governments,
Background NGOs, private operations, and farmers organizations
aim to expand the agrarian sector by developing new
Physical setting and agriculture. The US Caribbean products, markets, and improving supply chains.19
includes six large islands making up Puerto Rico and
USVI, and over 800 mostly uninhabited smaller islands Hurricane effects on agriculture in the US Caribbean.
and cays. The two archipelagos cover 9,450 km2 with a The US Caribbean is exposed to tropical cyclones,
population of 3.4 million in Puerto Rico and 104,000 in ie, hurricanes, storms, and depressions, between
USVI. Steep gradients in rainfall and high diversity of June and November (hurricane season). From 1867
soils allow for a great variety of agricultural products. to 2017, 94 storms and hurricanes passed over the
Hay and pasture cover the greatest area, dairy gener- US Caribbean.20 From 1950 to 2018, 32 hurricanes
ates the most economic activity, and aside from grass- passed within 322 km (~200 miles) of the main
lands, coffee represents the largest total area of agri- islands, including six hurricanes passing directly over
cultural land (5,221 ha). Other principal crops include Puerto Rico and 11 hurricanes passing directly over
plantains, bananas, other fruits and vegetables, hay for or 80 km (50 miles) from the islands of St. Croix, St.
livestock, ornamentals and nurseries, and root crops Thomas, or St. John.21 Hurricanes have been respon-
for local consumption.16 Agriculture contributes to sible for major economic losses to the agricultural
about 1 percent of the GDP in Puerto Rico17 and about sector of the US Caribbean (Table 1). For example,
2 percent in USVI.18 Agriculture is important to local Hurricane Georges (Category 3 SSWS) in 1998 caused
livelihoods and represents a prevalent land use across a loss of $308 million to the agricultural sector of
the territories. Puerto Rico contains a total of 191,695 Puerto Rico ($665.6 in 2020),22 representing 7 percent
ha of farmlands from 8,230 farms, with farms being an of the total economic losses to the commonwealth.23
average of 23 ha, while USVI contains approximately Losses were mostly in coffee ($97.7 million), plantains

Table 1. Economic losses in the agricultural sector in a sample of hurricanes
that affected Puerto Rico (USD $, not adjusted for inflation). Data obtained from the

Puerto Rico Planning Board and the Puerto Rico Department of Agriculture7,29,30

Hurricane name Estimated Economic losses Percentage Percentage
(Category Saffir of losses in of losses in
Simpson Hurri- Date economic losses in the agricul- Percentage agricultural agricultural
cane Wind Scale) i­ nfrastructure
(gross, USD, not tural sector (USD, of total products

adjusted) not adjusted)

Luis (4) September 4-6, 1995 147,529,045 12,015,415 8 98.7 1.4

Marilyn (2) September 15-16, 1995 59,047,640 8,043,729 14 NA 0.9

Berta (1) July 8, 1996 33,110,668 6,006,140 18 NA 2.2

Hortense (1) September 9-10, 1996 489,547,129 128,390,000 26 94.7 5.3

Georges (3) September 21, 1998 4,287,000,000 307,900,000 7 NA 25.0

Irma (5) September 6, 2017 1,656,900,000 45,820,907 3 98.7 1.3

María (4) September 20, 2017 40,661,600,000 2,011,365,815 5 9.3 90.7

Special Issue on Puerto Rico

Journal of Emergency Management 137
Vol. 19, No. 8

($42.8 m), and bananas ($7.7 m). The coffee sector and internet communications were cut off throughout
lost 64 percent of its crops causing the abandonment the island a month after these hurricanes struck.27,28
of coffee production by some farmers and a decrease Radio became the only means to reach the public and
in coffee farms in the following years.24 Furthermore, transmit information about relief efforts, as well as
Borkhataria et al.24 report that hurricanes were con- the principal means of communication between agen-
sidered by farmers to be the most important obstacle cies and citizens.28 Landslides, fallen debris, damaged
to coffee production in Puerto Rico followed by lack of roads and bridges, electrical posts and trees downed
capital, unavailability of workers, and erosion, among by winds, and overflowing rivers caused widespread
others. disruption in transportation in the aftermath of the
2017 hurricanes. Particularly in rural areas, many
Hurricanes Irma and María effects on agriculture, for- roads were blocked or otherwise could not be used
ests, and infrastructure. Hurricanes Irma and María for an extended period exceeding 6 months in some
(2017) caused catastrophic damages to crops and infra- places, hindering the distribution of relief resources
structure across farmlands in Puerto Rico and USVI. (food, medicines, and tarps), and delaying the resto-
Their combined effects likely constitute the greatest ration of utilities, eg, power lines.27 In light of these
effect to the agricultural economy in recent years challenges, and the potential for climate change to
(Table 1). In Puerto Rico, crop insurance indemnities exacerbate many of those challenges, research on
paid to farmers to mitigate crop losses arising from the prevalence and effectiveness of hurricane-related
hurricanes Irma and María in 2017 encompassed 94 strategies for preparation, adaptation, and recovery
percent of the total indemnities paid in Puerto Rico in is important to continued economic and food security
the 2010-2019 period (~34 million).25 Losses in crops, in the region.
ornamentals, livestock, and animal products sur-
passed $2 billion. In USVI, farms, ranches, and infra- METHODS
structure including government agricultural offices
experienced sizable damages; however, no official We developed and administered a survey instru-
estimates of the economic losses caused by the storms ment to better understand the perspectives of public
have been published to date. Both Puerto Rico and sector land management advisors regarding hurri-
USVI experienced widespread defoliation, branch loss, cane preparedness and recovery in Puerto Rico and
and mortality of trees by hurricane winds.6,26 Millions USVI. The survey was initially created for advisors
of downed trees affected infrastructure, power lines, working for the Natural Resources Conservation
roads, and trails, filling right-of-ways with vegetated Service (NRCS), Cooperative Extension, and state
debris. Vegetative debris management became a costly forestry agencies in 2017, as part of a regional
and challenging issue due to prohibitions on burning, analysis of hurricane effects across nine states in
limited landfill capacity, and public concern about lack the Southeast US, Puerto Rico, and USVI.12 In 2018,
of appropriate use of wood resources. Potentially valu- we expanded the survey to include advisors work-
able downed trees were not utilized as wood products ing in selected agencies that provide agricultural
due to the lack of planning, an underdeveloped market support and are involved in hurricane preparedness
for salvage logs, and a poorly coordinated system to and recovery in Puerto Rico and USVI (Table 2). The
utilize salvage logs. survey was pretested by six agricultural advisors
from NRCS, Cooperative Extension, and the Puerto
Hurricanes Irma and María also caused the Rico Department of Agriculture (PRDA) and modified
failure of critical electric infrastructure and the loss to improve clarity based on their input. The survey
of power throughout the region, including severe was available in English and Spanish. Translation to
damage to telecommunications infrastructure.3 In Spanish was conducted by an accredited translator
Puerto Rico, approximately 80 percent of all cell and confirmed by two native Spanish speakers from
phone towers were still out of service, and television the US Forest Service.

Special Issue on Puerto Rico

138 Journal of Emergency Management
Vol. 19, No. 8

Table 2. Description of organizations in PR and USVI providing assistance
to the agricultural sector after hurricanes that participated in this study

Institution Description Role in hurricane assistance

Cooperative Extension: Cooperative Extension System (CES) is hosted by
University of Puerto
Rico-Mayagüez and Land Grant Universities. Extensionists provide tech- Provides resources and training related to
University of the Virgin nical assistance, and training incorporating the latest disaster preparedness and recovery
Islands scientific research on agriculture, natural resources,

health, nutrition, and more

Puerto Rico Depart- DAPR facilitates and promotes agricultural produc- First responders. Hurricane damage evalua-
ment of Agriculture tion, commercial fishing, and aquaculture in Puerto tions. Provision of information on recovery aids
(DAPR) Rico; administers rural development, credit, and and programs. Coordination of animal care and
conservation programs designed to implement mortality management (livestock and poultry).
national growth policies; conducts scientific and Support in debris removal
technological research in all areas of agriculture

Natural Resources Provides conservation planning and technical assis- Assist farmers with the implementation
Conservation Service tance to land managers. Designs and promotes land of agricultural conservation practices that
(NRCS)—Caribbean Area management conservation practices and programs address hurricane-related concerns. Provides
(PR and USVI) technical and financial assistance after
disasters to reduce threats to life and property
through easements and recovery activities

Farm Insurance Administers and distributes government-subsidized Assist farmers in acquiring farm insurance.
Corporation of Puerto agricultural insurance to farmers against losses or Evaluate farm damages after hurricane events
Rico damages caused to plantations and crops by natural and process crop insurance compensation
risks such as hurricanes, floods, and named storms

APHIS—Veterinary Protects and promotes agricultural health, regulat- Provide support during emergencies through
Services ing genetically engineered organisms, administer- the FEMA Emergency Support Function (ESF)
ing the Animal Welfare Act and carrying out wildlife #11, including responses to animal and agri-
damage management activities cultural health issues; provision of technical
expertise, coordination and support of animal
and agricultural emergency management

Land Authority of The Land Authority is a programmatic and opera- First emergency responders after hurricanes.
Puerto Rico tional component of the DAPR. Its goal is to acquire, Work in debris removal, farm access, flood con-
conserve and preserve land of high agricultural trol, and recovery assistance
value, and to facilitate the use of this land for agri-
culture production through land leases or sales

Assist in natural resource management—eg,

Conservation districts are units of government erosion and sediment control, stormwater

established under state law to develop locally driven management, flood control, and water use effi-

Soil and Water solutions to natural resources concerns. Districts ciency—respond to natural disasters with clean-
Conservation District
work with landowners and operators to manage and up efforts and restoration—eg, cropland and

protect land and water resources on private and drainage system cleanup, repair of conservation

public lands best management practices, livestock mortality

issues, and waste management systems

USVI Department of The Virgin Islands Department of Agriculture devel- First responders. Hurricane damage evalua-
Agriculture ops, supports, and promotes an economically lucra- tions. Provision of information on prepared-
tive agricultural industry in USVI while protecting ness strategies to reduce losses, recovery aids
farmers, consumers, and the environment and programs. Coordination of animal care and
mortality management (livestock and poultry)

This table describes the main public institutions that provide direct assistance to farmers, and is not meant to be comprehen-
sive, as other agencies that deal with disaster preparedness and management might also provide direct or indirect services to
farmers—eg, PR Department of Natural and Environmental Resources, PR Emergency Management Bureau, and Virgin Islands
Territorial Emergency Management Agency.

Special Issue on Puerto Rico

Journal of Emergency Management 139
Vol. 19, No. 8

The instability in the electrical and communica- and low-median prevalence and labeled them as
tion services lasted beyond the hurricane events, having an importance-prevalence gap. Summary sta-
affecting internet and phone lines in several agencies tistics of the quantitative data were calculated using
participating in this study, resulting in challenges SPSS statistical software.32
with distributing an online survey. Thus, we adapted
the online survey to a paper version and distributed Finally, open-ended questions about the topics,
these among the agencies without consistent internet (1) challenges faced by farmers and ranchers dur-
access. To distribute the survey, we requested that ing recovery, (2) challenges faced by advisors during
agency directors and Deans or Associate Deans for recovery, (3) and additional resources considered
Cooperative Extension send the survey to staff who necessary by advisors to better assist farmers and
assist farmers and who are involved in hurricane pre- ranchers with hurricane preparedness and recovery,
paredness and recovery. The request for participation were analyzed for themes following thematic analysis
was sent three times between November 2018 and methods using Dedoose.33,34 Responses were coded
January 2019, following the Tailored Design Method.31 independently by two coders from the research team;
The survey and revision to the survey (Spanish trans- mismatched codes were reviewed collectively by both
lation) were approved by the University of Florida’s coders, who came to a consensus.
IRB protocol #IRB201801856 and Revision 1 for IRB
Study #IRB201801856. RESULTS

Survey questions focused on three main topics: (1) Description of survey respondents
the main effects associated with hurricanes and asso- and expert knowledge
ciated impacts on farmers and ranchers, (2) the main
challenges faced by farmers, ranchers, and advisors in A total of 202 agricultural advisors from eight
dealing with hurricanes in the Caribbean, and (3) the institutions responded to the survey. The largest num-
relative importance and perceived prevalence of hur- ber of respondents were affiliated with Cooperative
ricane preparedness and recovery strategies among Extension in PR and USVI (39 percent). This was
land managers. Question formatting included Likert followed by advisors affiliated with PRDA (31 per-
scales, multiple-choice, and open-ended queries. To cent), NRCS-Caribbean Area (14 percent), and Farm
assess the importance of hurricane preparedness Insurance Corporation (FIC) (6 percent) (Table 2).
and recovery strategies, we provided a list of short- Forty-six percent of the respondents held a gradu-
term hurricane preparedness strategies (measures ate-level degree, 37 percent a bachelor’s degree, 10
taken to prepare for a forecasted hurricane arriving percent other qualifications, eg, Associates degree
in less than a week), long-term hurricane prepared- and graduate courses, and 7 percent preferred not to
ness strategies (measures taken to protect farms answer. Although many of the advisors work both in
from hurricanes that may come within months or Puerto Rico and USVI, 94 percent are based in Puerto
years), and hurricane recovery strategies (measures Rico and only 6 percent in USVI.
taken to assess and repair damage after a hurricane).
The list of strategies was gathered from Cooperative The advisors surveyed reported using multiple
Extension and NRCS publications and from con- resources and organizations for information on hur-
versations with experts. Respondents were asked to ricane preparedness and recovery for professional
estimate the proportion of farmers and ranchers they purposes. Among a list of eight types of resources and
work with who use a strategy (hereafter referred to organizations, advisors indicated that they mostly
as “prevalence”), as well as how important each strat- rely on information through training, workshops,
egy is for successful hurricane preparedness/recovery and webinars (44 percent), technical reports (38
(hereafter referred to as “importance”). We then iden- percent), resources from Cooperative Extension (35
tified strategies with both a high-median importance percent), social media (31 percent), and the expertise
of other colleagues (31 percent). Advisors also rely
on peer-reviewed literature or journals (25 percent)
and information from relief agencies or groups (22

Special Issue on Puerto Rico

140 Journal of Emergency Management
Vol. 19, No. 8

percent), eg, the Puerto Rico Emergency Management USVI according to their level of impact on a 5-point
Bureau and the Virgin Islands Territorial Emergency Likert-type scale (1 = no impact; 2 = low impact; 3 =
Management Agency. Overall, most advisors (73 moderate impact; 4 = high impact; 5 = devastating
percent) feel confident or very confident in their impact) (Table 3). Seventy-one percent of respondents
ability to assist land managers on topics related to indicated that loss of power had a devastating impact,
hurricane events in general. However, confidence and another 23 percent indicated that loss of power
levels varied substantially with regard to specific as having a high impact. Fifty-eight percent indicated
hurricane-related effects. While more than 60 per- that loss of phone/internet communication as having a
cent of advisors indicated that they feel confident or devastating impact, and another 31 percent as having
very confident advising on the topics of soil loss or a high impact. Fifty-two percent indicated impassable
erosion, wind damage, stormwater inundation, and roads as having devastating impacts, and 37 percent
landslides, advisors feel less confident advising on as having a high impact. Likewise, 52 percent indi-
topics related to coastal hazards. Only 27 percent cated fallen trees as having devastating impacts, and
reported that they feel confident or very confident 32 percent as having a high impact. Sorting by mean
advising on coastal flooding and storm surge, and perceived impact produced similar results. The hurri-
21 percent reported the same level of confidence on cane-related effects rated as having the most devas-
saltwater intrusion topics. tating impact on agriculture were loss or power (μ =
4.61), loss of phone/internet communication (μ = 4.42),
Effects and challenges—agricultural impassable roads (μ = 4.37), and fallen trees (μ = 4.37).
lands and farmers and ranchers Coastal flooding and storm surge flooding ranked
as the effect with the lowest impact on agricultural
Advisors ranked the perceived effects of the lands in the islands (μ = 2.91) (Table 3).
2017 hurricanes on agriculture in Puerto Rico and

Table 3. Main effects associated with hurricanes impacting farmers [scale impact: no impact (1), low i­mpact
(2), moderate impact (3), high impact (4), and devastating impact (5)]. Sorted by devastating impact

Impact N No impact Low impact Moder- High impact Devastat- Mean SD
(percent) (percent) ate impact (percent) ing impact
(­ percent) (percent)

Loss of power 184 2 2 2 23 71 4.61 0.76

Loss of phone/internet 185 1 3 7 31 58 4.42 0.82
communication

Impassable roads 183 1 2 8 37 52 4.37 0.79

Fallen trees 188 1 2 9 36 52 4.37 0.80

Lack of potable water 186 2 7 13 32 47 4.15 1.00

Flooding from rain 186 1 7 20 31 41 4.03 1.00

Gas shortages 184 3 4 10 42 41 4.14 0.96

Landslides 186 4 8 21 34 33 3.86 1.08

Coastal flooding/storm 176 30 14 13 20 23 2.91 1.57
surge flooding

Evacuation requirements 175 10 20 33 22 15 3.13 1.19

Special Issue on Puerto Rico

Journal of Emergency Management 141
Vol. 19, No. 8

Main challenges faced by farmers and Table 4. Main challenges faced by land managers
ranchers during hurricane recovery
Percentage of
We analyzed the open-ended question, “In your
opinion, what is the biggest challenge the farmers # of times respondents
and ranchers you work with face during hurricane
recovery?” Advisors frequently mentioned multiple Coded challenges men- who men-
challenges (Table 4). Eighty-three percent of the
respondents mentioned challenges related to farm tioned tioned this as
management issues, which included unavailability of
farm supplies such as seeds, fertilizers, and feed for a challenge
animals (mentioned by 16 percent of total respond-
ents), unavailability of equipment and machinery (14 Farm management 147 83
percent), and infrastructure damages and repair (12
percent). Farm supplies unavailable, ie, 29 16
seeds, fertilizers, and feed
In looking more specifically at farm management
issues and challenges reported by respondents, sev- Equipment and machinery una- 24 14
eral of the open-ended answers indicated the need vailable and losses
for large equipment particularly for recovery efforts
post-disaster. There was “[a] lack of heavy equipment Infrastructure damage and 22 12
(such as bulldozers, excavators, or diggers) needed repair
for clearing the roads, accessing farms, preparing
land to sow new crops, disposing of dead animals, Animal care, losses, and dead 16 9
and complying with health parameters.” Other chal- animals disposal
lenges included a delay in time to obtain equipment,
“The heavy machinery we needed to clear the trails Labor unavailable 15 8
and roads to the farm were not available for many
months after the hurricane,” and “damaged equip- Materials for repairs unavailable 10 6
ment such as tree-cutting saws, water cisterns, and
the water well motors that draw water for their Debris removal and clean up 10 6
animals.”
Crop losses and damaged crops 6 3
Thirty-one percent of participants mentioned
facing challenges with electricity and fuel shortages, Salvaging and handling down or 5 3
while 23 percent mentioned access and transporta- damaged timber
tion. In reporting challenges with electricity and fuel
shortages, some responses indicated a lack of prepa- Hurricane plan unavailable 4 2
ration led to the reliance on electric generators. One
respondent stated, “Farmers do not prepare ahead of Other 63
time, so they do not have the vital fuel reserves that
are necessary for cattle ranchers. Since the islands’ Electricity and fuel shortages 55 31
electrical system is fragile, they depend on electric
generators.” Access and transportation 40 23

One respondent indicated concern over provid- Financial 39 22
ing aid to all areas due to transportation problems,
stating, “The biggest challenge is how to manage Environmental/ecological 39 22
the recovery aids. Most of the farmers are from very
Potable water 27 15

Flooding and excess rain 85

Landslides and erosion 42

Government assistance and aid 37 20

Communication 31 18

Markets 14 8

Personal effects on family and 10 6
friends

Information needs 74

Bold font indicates parent codes, and normal font indicates
subcodes (total number of respondents, n = 177).

Special Issue on Puerto Rico

142 Journal of Emergency Management
Vol. 19, No. 8

rural areas where the access to roads are drastically Twenty-nine percent mentioned challenges related
affected after a hurricane.” Furthermore, 22 percent to working for government agencies. Specifically,
of participants mentioned concern over financial respondents mentioned lack of planning and coordi-
issues. Other challenges include those related to nation or lack of leadership (mentioned by 14 percent
environmental and ecological problems—eg, potable of total respondents), lack of educational materials/
water and flooding, mentioned by 22 percent of total information and insufficient training (10 percent),
respondents—government assistance and aid (20 and insufficient financial resources (7 percent) among
percent), communication (18 percent), markets (8 other government challenges.
percent), personal impacts on family and friends (6
percent), and information needs (4 percent) (Table 4). Government agency challenges were often dis-
Of the communication issues, one respondent said, cussed as to leading to additional challenges. One
“The main problems were the [lack of] telephone com- respondent stated, “[For government agencies, the
munication and road access.” main challenge is] the lack of agency preparation and
organization. There is no real plan to follow, [forcing
Frustration over government assistance and aid us] to work blindly. We are not given information or
and political impacts were also evident in some of training on the steps to follow.” Several others dis-
the responses. One respondent stated, “[A major issue cussed government agency challenges as a “lack of
is the] high level of bureaucracy and the very slow tools and applicable practices” and a “lack of financial
payment of subsidies and incentives from regulatory resources required for immediate help.”
agencies. [Another component of this issue is the]
high politicization of the state [government].” Communication obstacles, eg, internet and phone
lines being down, were shared by several advisors
Main challenges faced by advisors assisting farmers (29 percent). Difficulties with financial assistance for
and ranchers during hurricane recovery recovery were also shared by several respondents (17
percent), including slow, unavailable, or inadequate
We subsequently analyzed the second open-ended aid (10 percent), bureaucratic red tape and exces-
question, “What is the biggest challenge you face in sive paperwork (5 percent), and lack of information
assisting land managers/landowners during hurri- about aid (4 percent). Other challenges frequently
cane recovery?” Advisors frequently mentioned mul- mentioned related to employees’ personal needs (16
tiple challenges (Table 5). Forty-three percent of the percent), farmer challenges (13 percent), and utilities
respondents indicated difficulty reaching farmers, (10 percent) (Table 5).
which included inability to access farms, impassable
roads, and landslides (mentioned by 36 percent of Several of the personal challenges mentioned
total respondents), transportation issues and lack of were the same that were encountered by farmers
4 × 4 vehicles (8 percent), and inability to reach farm- and ranchers. Personal challenges included “insuf-
ers in a timely manner (4 percent). ficient supply of water and gas for your own family
before going out to help [farmers] and quickly run-
Open-ended responses indicated that the inabil- ning of our stored supply.” One respondent stated,
ity to access farms and communicate with farmers “[One of the biggest challenges for me was that] I
were major challenges following hurricanes. Many was in the same situations as [the farmers I was
respondents indicated transportation due to lack of meant to help]: no food, no water, no gasoline, and
a 4 × 4 vehicle, landslides, runoff, and storm debris [damages to] my house.” In assisting with farmers
made it very hard to reach farmers. One respondent and ranchers, advisors also indicated challenges
stated, “The biggest challenge is reaching the farms. with assisting with farmer personal challenges and
My car is not a 4 × 4 and in cases where there are emotional support. One respondent stated, “One sig-
fallen trees or poles, it is impossible for me to access nificant challenge was the lack of preparedness to
the farms to collect data on damages and to offer deal with the farmers’ emotions and reactions after
recovery recommendations.” the disaster.”

Special Issue on Puerto Rico

Journal of Emergency Management 143
Vol. 19, No. 8

Table 5. Advisors’ main challenges Table 5. Advisors’ main challenges
assisting landowners assisting landowners

(total respondents, n = 163) (total respondents, n = 163) (continued)

Coded challenges # of times Percentage Coded challenges # of times Percentage
mentioned of respond- mentioned of respond-

ents who ents who
mentioned mentioned

this as a this as a
challenge challenge

Reaching farmers 78 43 Must attend to personal and 3 2
family needs
Unable to access farms/impass-
able roads/landslides 58 36 Farmer challenges 26 13

Challenges with transportation 13 8 Insufficient financial resources 11 7
or lack of 4 × 4 vehicles

Unable to reach farms in a 7 4 Land tenure or farm record 7 4
timely manner issues

Government agency 68 29 Insufficient resources (mate- 5 3
challenges rials, equipment, seeds, and
machinery)

Lack of planning and 23 14 Lack of planning 32
c­ oordination or lack of 16 10
leadership Utilities 24 10

Lack of educational materials/ Fuel shortages 96
information and insufficient
training Lack of potable water or
utilities in general
Insufficient financial resources 12 7 85

Insufficient resources (mate- Power outages and shortages 7 4
rials, equipment, seeds, and
machinery) 8 5 Bold font indicates parent codes, and normal font indicates
subcodes.

Insufficient personnel 64

Inadequate recovery practices/ 3 2 Preparedness and response
programs
The relative importance and perceived prevalence
Communication challenges/ 47 29 of hurricane preparedness and recovery strategies
internet and phone lines down among land managers (farmers and ranchers) in
the Caribbean. A list of 13 strategies for hurricane
Financial assistance for 31 17 preparedness and recovery was given to advisors
­recovery to gather their perspectives on how important and
prevalent these strategies are in agricultural lands.
Aid slow/unavailable/ 16 10 Responses are on a 4-point Likert-type scale (1 = not
inadequate important; 2 = somewhat important; 3 = important;
4 = very important); mean response (μ) and standard
Bureaucratic red tape and 8 5 deviation (σ) are reported in parenthesis. Regarding
excessive paperwork importance, the top strategies for short-term pre-
paredness are stock up on potable water (μ = 3.80;
Lack of information about aid 7 4

Personal employee challenges 26 16

Provision of emotional or 23 14
motivational support or feeling
powerless to help

Special Issue on Puerto Rico

144 Journal of Emergency Management
Vol. 19, No. 8

σ = 0.51), stock up on fuel for generators (μ = 3.79; From the list of 16 hurricane-related strategies for
σ = 0.46), stock up on feed (μ = 3.74; σ = 0.58), stock up long-term preparedness, the following ranked as the
on emergency cash (μ = 3.73; σ = 0.51), and reinforce top-five in terms of their perceived importance: cre-
the roofs of structures (farmsteads, dairies, and enclo- ate a hurricane preparedness plan (μ = 3.75; σ = 0.53),
sures) (μ = 3.72; σ = 0.54). Three of the most impor- invest in erosion control (μ = 3.67; σ = 0.59), establish
tant practices were also among the most prevalent a water storage system resistant to hurricane winds
practices observed by the advisors in the agricultural (μ = 3.67; σ = 0.61), create and secure a seed bank (μ =
lands they serve: stock up on potable water (μ = 3.47; 3.66; σ = 0.64), and invest in generators (μ = 3.59; σ =
σ = 1.15), stock up on fuel for generators (μ = 3.42; 0.68). The long-term preparedness strategy ranked as
σ = 1.08), and inventory livestock (μ = 3.08; σ = 1.23) most prevalent is invest in generators, which is gener-
(Figure 1). ally perceived to be applied by about 50 percent of land
managers. All the remaining strategies listed are con-
Following Wiener et al.,12 we identified strategies sidered important or very important, though applied by
with a high importance and perceived low prevalence less than 50 percent of the land managers. Therefore,
and categorized them as strategies with an impor- all the long-term preparedness strategies demon-
tance-prevalence gap. Specifically, these included strated an importance-prevalence gap (Figure 2).
strategies with a median importance of very impor-
tant and a median prevalence of about 50 percent or From the list of nine hurricane-related strategies
lower, as well as strategies with a median importance for recovery, the following ranked as the top-three in
of important and a median prevalence of less than terms of their perceived importance: apply for disaster
50  percent or lower. Except for stock up on potable assistance (μ = 3.82; σ = 0.49), make crop insurance
water and stock up on fuel for generators (median val- claims (μ = 3.79; σ = 0.46), and inventory livestock (μ =
ues indicating very important and more than 50 per- 3.77; σ = 0.54). These strategies also ranked as the
cent prevalence), all the short-term strategies listed top-three in terms of prevalence. Examine livestock
had an importance-prevalence gap (Figure 1). for infectious diseases, spraying livestock with insect

Figure 1. Mean importance and mean perceived prevalence of short-term hurricane preparedness strategies
(bars). Vertical lines represent median importance and median prevalence values. Asterisks indicate strategies
with an importance-prevalence gap.

Special Issue on Puerto Rico

Journal of Emergency Management 145
Vol. 19, No. 8

Figure 2. Mean importance and mean perceived prevalence of long-term hurricane preparedness strategies
(bars). Vertical lines represent median importance and median prevalence values. Asterisks indicate strategies
with an importance-prevalence gap.

repellent, hiring temporary labor, using alternative on how we can effectively educate farmers.” Others
chemical application techniques, and testing flooded were more specific regarding overall training needs.
crops for contaminants are considered important or very “We need intense workshops on how to preserve
important recovery strategies. However, advisors per- water, air and land resources. It is not only important
ceive that these are applied by less than 50 percent or to give the farmers incentives, but also to make them
none of the land managers they serve and as such dem- aware of the importance of preserving and caring for
onstrated an importance-prevalence gap (Figure 3). the resources.” Another response indicated a need for
mental health training. “It might be helpful to have
What is needed—advisors psychology workshops to equip us to better attend to
Finally, we analyzed additional resources or infor- farmers during the first visits [after the disaster].”
Respondents also suggested methods and logistics of
mation needed by advisors to better assist land training. One respondent stated, “[We need a] series
managers/owners with hurricane preparation (short of videos and lectures on how to prepare for these
and long term) and recovery (Table 6). Qualitative events. The duration should be no more than 30
responses revealed a substantial need for professional minutes and it should be aimed at farmers. I suggest
development resources to better assist in hurricane content could be composed of information that is 25
preparedness and recovery. Approximately 65 percent percent [focused on actions to take] before the event,
of the respondents indicated the need for more train- 25 percent during the event, and 50 percent on the
ing and workshops on hurricanes and related topics recovery process.”
for the three stages: short-term preparedness (men-
tioned by 33 percent of total respondents), long-term Forty-four percent of the respondents indicated
preparedness (35 percent), and recovery (33 percent). the need for additional educational resources and
information, principally information on prepared-
Several respondents were specific regarding the ness and recovery practices and fact sheets, reports,
training content that was needed including “first aid guides, or manuals to deliver to farmers. The respond-
and rescue,” “recommended steps to follow during ents’ request for additional educational resources
recovery that do not rely on subsidies” and “guidelines

Special Issue on Puerto Rico

146 Journal of Emergency Management
Vol. 19, No. 8

Figure 3. Mean importance and mean perceived prevalence of hurricane recovery strategies (bars). Vertical
lines represent median importance and median prevalence values. Asterisks indicate strategies with an impor-
tance-prevalence gap.

and information was relatively higher for long-term means such as 4 × 4 vehicles to reach farms after
preparedness (35 percent). hurricane events and machinery for debris removal,
along with readily available gas and diesel to supply
In addition to training and workshop, open-ended government vehicles. Some respondents (14 percent)
responses included a need for information, financial urged for more financial resources and improved
resources, and government coordination. One respond- agencies coordination during emergencies, among
ent stated, “up to date on what the agency’s plan is, other needs (Table 6).
what services and support we can [and can’t] provide,
and how we can best serve the agency in order to ben- DISCUSSION
efit the farmers.” Others stated a need for “assistance
to fill out the high amount of disaster assistance appli- Effects
cation paperwork that is required to aid landowners.” This study examined the effects of hurricanes on
One respondent stated, “[We need better] coordina-
tion in the provision of assistance and services from US Caribbean agriculture as perceived by agricultural
government agencies. [We need them to prioritize the advisors using survey data from over 200 participants
agricultural sector.] For example, [we need assistance across eight institutions. The most prominent effects
with] animal burial and road cleaning. [It would also on the agricultural sector highlighted by advisors in
be helpful to] inventory the crops before and after the our study—power outages, communication issues, and
hurricane to determine where the real needs are in impassable or closed roads—are consistent with chal-
terms of opening imports.” lenges identified in hurricane effects evaluations in
both territories.35,36
Other information requests include information
on aid, crop insurance, and programs (24 percent) As noted in the introduction, hurricanes Irma
and information on various management issues (19 and María resulted in the collapse of 80 percent of
percent), particularly during the recovery stage (20 the power lines and a blackout that left all 3.4 million
percent). Additionally, multiple advisors (18 percent) residents in Puerto Rico without electricity for several
called for suitable machinery and transportation weeks. In some areas, the blackout lasted more than
1 year, with only 65 percent of the island’s electricity

Special Issue on Puerto Rico

Journal of Emergency Management 147
Vol. 19, No. 8

Table 6. Additional resources or information needed by advisors to better assist land managers/owners
with hurricane preparation (short and long term) and recovery (total respondents, n = 97; total number of

respondents who commented on short term (n = 79), recovery (n = 49), and long term (n = 60))

Percentage Percentage Percentage Percentage Percentage

of respond- # of times of respond- # of times of respond- of respond- of respond-

Coded needs # of times ents who mentioned ents who mentioned ents who ents ents who

mentioned mentioned under ST mentioned under REC mentioned mentioned mentioned

this as a needs under ST needs under REC under LT under LT

need needs needs needs needs

Trainings and work-

shops on hurricanes 63 65 26 33 17 35 21 35

and related topics

Educational

resources and infor- 56 44 21 19 7 14 28 35

mation needs

  Information on pre-

paredness/recovery 20 21 6 8 2 4 12 20

practices

  Factsheets/reports/

guides/manuals to 17 18 6 8 3 6 8 13

deliver to farmers

  Educational 10 10 6 8 1 2 3 5
resources general

  Other education 6 6 2 3 1 2 3 5
recommendations

  Technical resources 3 3 1 1 0 0 2 3
for outreach

Information on aid,

crop insurance, and 24 25 10 13 10 20 4 7

other programs

Information on

various management 19 20 7 9 6 12 6 10

issues

Machinery/ 18 18 7 9 9 16 3 5
transportation

  4 × 4 vehicles 8 8 2 3 5 10 2 3
applied

  Heavy equipment 8 8 4 5 3 6 1 2

  Gas/diesel 22111200
availability

Financial resources 15 14 3 4 2 4 10 15

  Financial resources 6 6 1 1 2 4 3 5
general

Special Issue on Puerto Rico

148 Journal of Emergency Management
Vol. 19, No. 8