Volcano icon

Disaster Stress Following a Hurricane:
The role of religious differences
in the Fijian Islands

The Australasian Journal of Disaster
and Trauma Studies
ISSN:  1174-4707
Volume : 1999-2

Disaster Stress Following a Hurricane:
The role of religious differences in the Fijian Islands

Matt Gillard & Douglas Paton, School of Psychology, Massey University, New Zealand. Email: D.Paton@massey.ac.nz

Keywords: Disaster, vulnerability, religion, culture, stress

Matt Gillard
Douglas Paton

School of Psychology,
Massey University
New Zealand


This paper explores the influence of religion on disaster stress in Fijian Islanders. Interview data revealed that religious groups could be segregated in regard to the assistance afforded them and the demands made upon them by religious organisations. A questionnaire examining these dimensions, together with a stress and traumatic stress measure, was used to compare the impact of Hurricane Nigel (1997) on Christian Fijians, Indians who follow Islam, and Indians who practice Hinduism. The results revealed that religious denomination exercised a differential impact on vulnerability, although differences were partially dependent on the measure of vulnerability used. Explanations for these differences, and their implications for intervention, are discussed.

Disaster Stress Following a Hurricane:
The role of religious differences in the Fijian Islands


While their unpredictable and uncontrollable nature precludes the possibility of mitigating most of the direct physical effects of natural disasters, several of their secondary consequences (e.g., mental health, economic) are less immutable. Recognition of this fact has focused attention on understanding the factors that influence vulnerability to hazard consequences. For example, religious beliefs, socioeconomic status, and education can influence the nature, intensity and distribution of traumatic and disaster stress reactions (Bolin & Klenlow, 1988; Eng, Hatch & Callan, 1985; Gerrity, 1994; Goodman, Saxe & Harvey, 1991; Riad & Norris, 1996; Walls & Zarit, 1991).

Regular exposure to hurricanes, and vulnerability to volcanic and earthquake hazard threats, makes the Fijian islands an appropriate group to study, particularly because the occurrence of a disaster is often accompanied by substantial foreign aid and assistance. Irrespective of its altruistic intentions, unless this aid is designed in a manner consistent with community and cultural needs, it may, rather than assisting recovery, produce response generated demands (Quarantelli, 1985) that compound community problems. Two problems with external aid provision are readily apparent. Firstly, aid practices and procedures, particularly those aimed at mitigating mental health consequences, are often developed in a cultural context that differs substantially from that prevailing within the affected country. This raises questions ranging from the validity of theoretical constructs and intervention procedures to the degree of cultural safety afforded by their use. Secondly, the ad hoc nature of intervention, largely as a consequence of the unpredictable (with respect to time, place, severity etc) nature of disasters, typically precludes attempts to identify the vulnerability characteristics that must be accommodated within culturally-safe interventions and which need to be considered if attempting to develop resilient communities (an important issue in countries like Fiji where natural disasters occur frequently). This pilot study will focus on the latter, but will comment on the former.

Several studies (e.g., Dewey, 1988; Walls & Zarit, 1991; Eng et al, 1985; Pargament, 1986) have shown religion to be an important determinant of traumatic reactivity. Common to Christian, Islamic, and Hindu groups alike in Fiji is an extremely high adherence rate. These religions are actively practiced by at least ninety five percent of the population. Practicing these religions extends beyond merely claiming membership to a certain group. Activities include regular attendance at the appropriate meeting house, regular and often public displays of faith (e.g., burning of incense in a dedication to a God on the front lawn of a residential home), frequent visitation from religious officials (for prayers or the collection of donations), maintenance duties at the local Church, Temple or Mosque, formal instruction in the ways of the faith for both children and adults, attending denominational schools for children, informal attendance at group get-togethers at members houses, and attending the many weddings and funerals. Religion is thus a highly salient element of Fijian cultural life, an done which may influence the post-disaster experience of members of these groups.

The nature and intensity of post-disaster stress reactions may be affected by the demands placed upon people by their respective religion. A religious denomination which, through overseas support (e.g., the Christian Church), can provide considerable material support (e.g., rebuilding houses, providing building materials, or providing shelter in large, well-constructed and equipped churches) may have an entirely different impact upon individuals or families than a religious organisation which, due to financial hardship, is forced to lobby its members to provide financial, material, and manual assistance following a disaster. Membership of a particular religious denomination can thus either assist coping efforts or exacerbate the situation by placing further demands upon individuals who are already heavily burdened with the demands of a disaster.

This paper describes an exploratory study undertaken to assess the influence of religion on stress reactions to natural disasters in the Fijian Islands. While it did not set out to examine the impact of a natural disaster per se, the coincidental occurrence of Hurricane Nigel (in 1997) shortly after the initial study was completed provided an unplanned opportunity to examine its impact on indigenous Fijians who follow the Christian religion, Indians who follow Islam, and Indians who practice Hinduism.


The first stage of the research involved conducting open-ended interviews with twenty members of each of the three religious denominations. Interviews were conducted on the hurricane belt down the west coast of the island of Viti Levu in Fiji. The setting was the town of Lautoka, a garment-manufacturing town that had been hit by most of the major hurricanes in previous years. The levels of education and literacy differed, although it was generally of a low/medium standard. Participants were selected at random on a door to door basis and interviews were conducted in the interviewee's private home. Interviewees were asked to describe how their religion affected their personal experience of hurricanes.

Summary of Interview Data

Some 94% of Fijian Christians expected their church to render assistance to them. There was a confident anticipation that the church would "get them back on their feet" no matter how severe the damage. Reports of the last hurricane included stories of the church building new houses for members whose homes were destroyed. The church often supplied members with food and provisions over and above standard government aid. There was a strong belief that relief money was no obstacle for the church and that a high level of competent support could be expected. With the availability of foreign aid (from overseas Christian Churches) the Christian organisations of Fiji are able to provide good support on all levels, including reconstruction of housing, relocation, limited financial aid, and household food and supplies. Members of the Christian Church do not have any demands placed upon them apart from attending church and their usual church duties. The Christian Church can be classified as providing an "assistance" environment.

While still anticipating some assistance, the expectations of members of Islamic Mosques and Hindu Temples were lower than those of their Christian counterparts. After a disaster, some 75% of Hindus and 63% of Muslims expected assistance to be forthcoming from their Temple or Mosque. Stress reactions could be exacerbated or compounded if this support fails to materialise or falls short of expectations.

Another difference between Christians, Hindus and Muslims was evident in regard to the expectations on the part of their respective religious centres to require and request assistance from its members. Hindus and Muslims are expected to provide manual labour (particularly by trades persons), food supplies (for the Mosque staff and Temple committees as well as the poorer of their members), building provisions (for the Mosques, Temples and their poorer members), and financial aid. Due to limited resources Muslim Mosques and Hindu Temples come under the "demand" category of post disaster interaction. That is, the members of these organisations are expected to make contributions toward the repair and continuation of their Mosques and Temples, as well as providing support and assistance to their poorer fellow members. Operating within a "demand" environment may increase the psychological vulnerability of members of these religious denominations.

On the basis of these observations, membership of the Hindu and Muslim religions could constitute a response generated demand. We could thus anticipate higher levels of post disaster stress amongst members of "demand" religions compared with "assistance" religions. The occurrence of Hurricane Nigel provided an opportunity to test this hypothesis. This was accomplished by the second stage of the research.

The second stage involved the content analysis of interview data to compile a demand/ assistance questionnaire (Gillard, 1998). This questionnaire was designed to assess differences between the three religious groups with respect to the a) the post-disaster demands (e.g., physical, financial) made upon them and, b) the assistance offered (e.g., spiritual, physical, financial) by their respective Church, Mosque or Temple. The questionnaire included a general stress symptom measure, the "HSCL-21", (Green, Walkey, McCormick, & Taylor, 1988), and a traumatic stress symptom measure, the "Impact of Events Scale" (IES - Horowitz, Wilner, & Alvarez, 1979) to assess differences in psychological vulnerability between religious groups. Because of concerns about the cultural validity of traumatic stress measures (Marsella, Friedman, Gerrity, & Scurfield, 1996), two measures of psychological vulnerability were used.

Although a third world nation, Fiji has 87% literacy. To minimise the risk of language problems affecting data collection the questionnaire was administered door-to-door to thirty members of each religious denomination. This ensured that the researcher was available to deal with language problems prior to respondents completing the questionnaire. Data was collected 3 - 4 weeks after hurricane Nigel. The sample population was representative of the population at large. Respondents ranged in age from sixteen to seventy two, included an equal proportion of men and women, and comprised individuals of high, middle, and low socioeconomic status.

Questionnaire results

Religious denomination was perceived to play a role in assisting respondents to cope with the hurricane. Some 89% of Fijians, 76% of Hindus, and 63% of Muslims stated that their respective religious beliefs were helpful during the crisis.

Table 1: Means and standard deviations for the IES, HSCL, Demand and Assistance scales for members of each religious denomination.

IES HSCL Demand Assistance

Means and standard deviations for each scale are recorded in table 1. Analysis of variance applied to the questionnaire data yielded significant differences between religious groups for demands (F (2, 94) = 10.238, p<0.001), assistance (F (2, 94) = 7.939, p < 0.001), HSCL-21 scores (F (2, 94) = 2.994, p < 0.05), but not on their IES scores (F (2, 94) = 0.749, p < 0.476).

Christian Fijians received higher levels of assistance from their Church than Muslims did from their Mosque. While Hindus received relatively high levels of assistance from their Temple, concomitantly high levels of demand may have countered any benefits from the former (table 1). A different perspective is afforded by table 2 where the proportions of each group receiving spiritual, physical and financial assistance is described together with the level of perceived satisfaction or adequacy of the assistance received. Not only were Fijian Christians more likely than their Hindu and Muslim counterparts to receive spiritual, physical and financial assistance, they were more likely to perceive this assistance as adequate for meeting their post-disaster needs.

Table 2:  The proportion of each group receiving spiritual, physical, and financial assistance and the level of perceived satisfaction/adequacy of received assistance.

Christian Hindu Muslim

Furthermore, Hindu's and Muslims both anticipated, to a significant level, that their respective religious organisations would place higher demands upon them after a hurricane than did Christian Fijians. While no Fijian Christian reported their church making demands upon them, 17% of Hindus and 20% of Muslims reported experiencing demands being made upon them by their temple or Mosque (see also table 1).

Religious organisations had a limited input into practical mitigation and preparation for hurricanes, with only 17% of Hindus, 7% of Fijians and 7% of Muslims reporting receiving information on preparation for hurricanes. Religious organisations did play a more active role in providing information on the religious aspects of the causation and nature of hurricanes (Hindus - 17%; Fijians - 23%; Muslims - 20%), but this did not include claims about its occurrence being linked to religion.


The members of all three groups reported that religious beliefs helped them cope with their disaster experience. Significant differences between them were evident with respect to the amount of assistance provided and the demands made upon them by their respective religions after the hurricane. In addition to it being a coping strategy, religious affiliation, by imposing additional response demands on certain groups, acted to heighten their psychological vulnerability. Overall, Fijians received higher levels of assistance and had lower levels of demands placed upon them after the hurricane than their Indian counterparts. These results also revealed that Hindu and Muslim expectations of assistance were not fully realised. This discrepancy can itself constitute a disaster stressor (Raphael, 1986).

Against a demand/assistance criterion, a relationship between vulnerability and religious affiliation can be discerned. Hindu and Muslim Indian groups are less likely to receive assistance from their Temples and Mosques, more likely to perceive assistance as being inadequate for their needs, and more likely to find their religious organisation making demands upon them after a disaster. Mosques and Temples, being of limited financial means, were unable to provide their members with financial support after a hurricane; instead requesting financial and material assistance from their members, increasing their vulnerability to experiencing disaster stress reactions.

Conversely, because Fijian Christians can anticipate high levels of assistance from their church, with no, or only minimal, demands being placed upon them after a hurricane, they can be placed in a lower risk category. The financially stable Christian Church was able to provide considerable assistance to its members, supplying food, clothing, and building materials, and even building replacement houses for those who had lost their homes. Fijian Christians were aware of this and were confident that such support would be available in future.

These results suggest that further research into the role of religious denomination both as a coping resource and as a vulnerability factor is warranted. Because some 95% of the population actively practice one of these religions, religious organisations are well placed to provide disaster education, plan and implement mitigation

programmes, and disperse aid. The perceived importance of religion as a coping resource, and the fact that religious organisations already play a small role in disaster preparation activities, suggests that it would be worth exploring whether it is a lack of resources and expertise that limits their playing a more active role in this respect. Channeling aid and preparatory activities through religious organisations would optimise dissemination, ensure that relevant information is translated appropriately for each group, and facilitate the implementation of mitigation strategies in a culturally appropriate manner.

When a traumatic or disaster stress measure of vulnerability was used, the picture was less clear. While a difference between groups was evident on the general stress measure (HSCL - 21), no significant difference in traumatic stress (IES) was evident. Several explanations can be proposed to account for this difference. For example, this observation may have been affected by the timing of data collection. The IES and HSCL-21 data were collected 3-4 weeks after the hurricane struck. In this context, the traumatic stress reactions triggered by the direct effects of the hurricane could have started to decline during this period, reducing any inter-group differences. In other words, religious affiliation could have influenced traumatic symptomatology during the immediate aftermath of hurricane impact, but the time lapse between then and the point of data collection may have obscured this influence. Since data were not collected immediately after impact no objective measure of the severity of the hurricane was available. Nor was data on the extent and uniformity of its impact or on rates of symptom change available. Consequently, we can only speculate on this issue. The assumption of traumatic symptom decline during this period can, however, be justified on several grounds.

For example, population familiarity with and prior (and regular) experience of hurricanes could facilitate prompt recovery. Further, the commencement of reconstruction activities in the immediate aftermath of the hurricane could provide a community environment that helped ameliorate traumatic stress symptoms. However, differences in the recovery demands associated with religious denomination during this period would manifest themselves as differences in general stress. When these data were collected (3-4 weeks post-impact) religious demands may have become a more salient stressor, and one whose impact was more accurately reflected by the content of the general stress measure.

Further uncertainty is injected into the interpretive process by a lack of pre-event scores on either measure with which to gauge actual impact and/or the rate of change in symptom score over time and in relation to changes in the demands experienced. There is a clear need for the use of longitudinal methodology (Paton & Smith, 1995) to examine disaster stress phenomena. A longitudinal framework provides a more appropriate vehicle for understanding how the stressors/demands that emanate from hazard effects change over time and how they interact with vulnerability factors. Pre- and post-event measurement is also essential if an accurate understanding of disaster stress phenomena is to be compiled.

Another possible explanation for the difference in perceived vulnerability concerns the validity of the traumatic stress and stress constructs that underpin the measurement instruments used here (Keane, Kaloupek & Weathers, 1996; Paton, Smith, Ramsay, & Akande, 1999). This issue is particularly salient in cross-cultural psychology where the characteristics of the population under study (e.g., in regard to its meaning systems, values, practices, and beliefs) may differ radically from those upon whom the test was developed. Several authors (e.g., Jenkins, 1996; Keane et al., 1996; Marsella et al., 1996; Paton et al., 1999) have argued that measurement of traumatic stress reaction remains a serious problem in cross-cultural research because the existing instruments often do not include indigenous idioms of distress or causal conceptions of traumatic stress reactions. While Marsella et al (1996) concluded that existing research suggests the existence of a universal biological response to traumatic events they argue that the use of standard definitions in identifying traumatic stress reactions can result in inaccurate conclusions regarding symptom prevalence rates.

Marsella et al (1996) outlined several strategies for improving the validity of cross-cultural studies. These suggestions included comparing the nature and frequency of symptoms between ethnocultural groups and standard western samples, and exploring symptom profiles between different ethnocultural groups (while controlling for age, gender, social economic status, education, and social class). Further suggestions include the examination of culture-specific disorders such as “latah,” “koro,” and “susto.”

The manner in which ethnicity and culture influence the meaning attributed to natural hazards, mitigation strategies and the expression of traumatic reactions and other emotional sequelae also deserve more detailed analysis (Jenkins, 1996; Marsella et al., 1996; Rosaldo, 1984). For example, the groups studied here differed from Western populations with respect to the influence of religious fatalism on the meaning attributed to disasters and on their social and psychological reactions. It remains to be determined whether, and how, this impacts on the traumatic stress construct and meaning of disaster. These authors argued for more research into defining the ethnocultural experience of traumatic events, evaluating similarities and differences between cultures, defining the nature of the constructs that underpin the interpretation of this experience and the expression of reactions, and the use of these findings to develop culturally-appropriate measurement instruments and intervention procedures. The need to develop research and intervention methodologies for use in cross-cultural settings with diverse contexts and populations has also been highlighted (Keane et al., 1996; Paton et al., 1999).

Finally, the scope of this study was limited by a lack of in-depth knowledge of the language, culture and religions of the three cultural groups studied. A working knowledge of Fijian and Hindi would have been beneficial during the interview stage. Even though the respondents spoke and wrote (87% literacy) English to a moderate degree, speaking in their first language may have caused them less anxiety and allowed them to have provided more detailed descriptions of their experiences. Apart from general language difficulties, it is impossible to gauge whether any subtle interpretations or meanings were lost in translation.

In general, research in this area will be facilitated by more detailed knowledge of the characteristics of the respective cultures and religions (Marsella, Friedman, & Spain, 1992; Marsella et al, 1996). Religious denomination may have implications beyond those discussed here. For example, there are strong relationships between religion and other factors likely to affect vulnerability, including education, socio-economic status, housing location and quality. In some contexts, such as Kosovo, political and inter-ethnic conflict would be highly salient variables and would have to be included within cross-cultural research agenda.

It is evident from this and other studies that vulnerability to disaster stress represents a complex web of religious, cultural, social and psychological factors. While these are more starkly contrasted in smaller, Pacific island communities, the issues raised here have implications for the more developed countries within the region. These dynamics may operate equally in Auckland, Melbourne, Sydney, or Wellington, but not be so obvious.

The complexity and uncertainty inherent within vulnerability analysis, coupled with the scarcity of research opportunities in this area, means that we rarely have the chance to get it right before intervening. Building a model of vulnerability, hazard and risk requires cooperation between scientists and practitioners to facilitate the evolution of understanding, methods and interventions, the comprehensive identification and analysis of vulnerability characteristics, the development of culturally-safe interventions (or interventions that, at least, accommodate indigenous factors) and the utilisation appropriate local media and systems for information dissemination and intervention. This will include the analysis of salient psychological constructs and the development of instruments capable of assessing them. Until this is done, research and intervention activities conducted within the international arena should, irrespective of underlying altruistic motives, proceed with caution.


Bolin, R.C. & Klenow, D.J. (1988) Older people in disaster: A comparison of Black and White victims. International Journal of Ageing and Human Development, 26, 29-43.

Dewey, D. (1988). When a congregation cares: Organising ministry to the bereaved. Special Issue: Cultural and religious perspectives of death. Death Studies, 12, 123-135.

Eng, E., Hatch, J., & Callan, A. (1985). Institutionalising social support through the church and into the community. Health Education Quarterly, 12, 81-92.

Gerrity, E.T. (1994). Relocation stress following natural disasters. In R.J. Ursano, B.G. McCaughey, & C.S. Fullerton, Eds.Individual and community response to trauma and disaster: The structure of human chaos, Cambridge: Cambridge University Press, pp. 220-247.

Gillard, M.J. (1998) Traumatic stress reactions: Cultural differences. Unpublished MA Thesis, Massey University, Palmerston North, New Zealand.

Goodman, L.A., Saxe, L., & Harvey, M. (1991). Homelessness as psychological trauma: Broadening perspectives. Special Issue: Homelessness. American Psychologist, 46, 1219-1225.

Green, D.E., Walkey, F.H., McCormick, I.A., & Taylor, A.J.W. (1988), “Development and evaluation of a 21 item version of the Hopkins Symptom Checklist with New Zealand and United States respondents”, Australian Journal of Psychology, 40, 61 - 70.

Horowitz, M., Wilner, M, & Alvarez, W. (1979) Impact of Event Scale: a measure of subjective stress. Psychosomatic Medicine, 41, 209 - 218.

Jenkins, J.H. (1996). Culture, emotion, and PTSD. In A.J.Marsella, M.J.Friedman, E.T.Gerrity, & R.M.Scurfield (Eds.), Ethnocultural aspects of posttraumatic stress disorder : Issues, research, and clinical applications. Washington, DC: American Psychological Association.

Keane, T.M., Kaloupek, D.G., Weathers, F.W. (1996). Ethnocultural considerations in the assessment of PTSD. In A.J.Marsella, M.J.Friedman, E.T.Gerrity, & R.M.Scurfield (Eds.), Ethnocultural aspects of posttraumatic stress disorder : Issues, research, and clinical applications. Washington, DC: American Psychological Association.

Marsella, A.J., Friedman, M.J., Gerrity, E.T., & Scurfield, R.M. (Eds.). (1996). Ethnocultural aspects of posttraumatic stress disorder : Issues, research, and clinical applications. Washington, DC: American Psychological Association.

Marsella, A. J., Friedman, M., &Spain, H. (1992). A selective review on the literature on ethnocultural aspects of PTSD. PTSD Research Quarterly, 2, 1-7.

Pargament, K.I. (1986). The psychology of religion : A clinical/community psychology perspective. Journal of Psychology and Christianity, 5, 68-72.

Paton, D & Smith, L.M. (1995) Methodological issues in the evaluation of work-related psychological trauma. Australian Psychologist, 30, 200 - 209.

Paton, D., Smith, L.M., Ramsay, R., and Akande, D. (1999) A structural re-assessment of the Impact of Event Scale: The influence of occupational and cultural contexts. In Gist, R. and Lubin, B. (eds) Response to Disaster: Psychosocial, Community and Ecological Approaches, Philadeplhia, Taylor & Francis.

Quarantelli, E (1985) Social Support Systems: Some behavioural patterns in the context of mass evacuation activities. In B. Sowder (ed) Disasters and Mental Health: Selected Contemporary Perspectives. Washington, D.C. DHSS Publication No. (ADM) 85 - 1421.

Raphael, B. (1986), When Disaster Strikes, Hutchinson, London,

Riad, J.K., & Norris, F.H. (1996). The influence of relocation on the environmental, social, and psychological stress experienced by disaster victims (Hurricane Andrew). Environment and Behaviour, 28, 163-182.

Rosaldo, M. (1984). Toward and anthropology of self and feeling. In R.A Shweder & R.A LeVine (Eds.),

Culture theory: Essays on mind, self, and emotion. Cambridge, MA : Cambridge University Press.

Walls, C.T. & Zarit, S.H. (1991). Informal support from Black churches and the well being of elderly Blacks. Gerontologist, 31, 490-495.


M Gillard & D Paton © 1999. The authors assign to the Australasian Journal of Disaster and Trauma Studies at Massey University a non-exclusive licence to use this document for personal use and in courses of instruction provided that the article is used in full and this copyright statement is reproduced. The authors also grant a non-exclusive licence to Massey University to publish this document in full on the World Wide Web and for the document to be published on mirrors on the World Wide Web. Any other usage is prohibited without the express permission of the authors.

| Home | Current | Back Issues | Reports | Conferences | Books | Links | Information |

Comments to
Massey University, New Zealand
URL: http://www.massey.ac.nz/~trauma/

Last changed August 31, 1999
Copyright © 1999 Massey University