Towards an understanding of social vulnerability and resilience to disasters

29 May 2014

Shelby Canterford and Valdis Juskevics

Science can tell us much about the nature of floods, how fires behave and the response of a building to strong winds. However, not as much is known about how a person will behave or react in the face of unfolding disaster. Like the landscape, different people react to all these things in different ways according to who they are and their locality to produce different levels of vulnerability and resilience. To understand these differences Geoscience Australia has developed a social research capability.

Continuing the success of Geoscience Australia's well-established post-disaster building damage surveys has enabled the social research survey responses to be linked back to detailed damage information taken in the immediate aftermath of major floods. This has been invaluable to the engineering vulnerability program and has led to the development of new models.

To date Geoscience Australia has conducted four household and three business postal surveys covering the flooding in Brisbane and Ipswich in 2011 and 2013, Tropical Cyclone Yasi in 2011 as well as flooding in Bundaberg in 2013. An additional Brisbane and Ipswich building survey was undertaken, with a focus on how dwellings had been rebuilt, either as like-for-like, with flood mitigation or as a complete rebuild.

Postal surveys have the advantage in that they are more cost effective to collect large amounts of data. Large numbers of responses are achievable making the results statistically significant, and questions can cover a broad range of topics.

However, there are some disadvantages. Asking detailed questions relating to a specific topic is not always possible. It is also not possible to question a respondent to identify why they answered a question in the way they did or to identify if the respondent had difficulty in understanding a particular question. One example of this comes from survey questions about warnings. Many respondents advised that they had been unaware of any warnings, however they also advised that they removed or relocated belongings, stocked supplies and sandbagged their home. Some even stated they were too busy packing to hear a warning. Such responses reveal the ambiguity that may exist within surveys, in terms of both research questions and participant responses.

Survey fatigue: a drain on respondents

For the initial survey of Brisbane and Ipswich 1267 households responded representing 26 per cent of those surveyed. While this percentage may appear low, it is actually good for this type of survey. Survey fatigue is an issue that all researchers face and can dramatically reduce response rates; and other forms of official information gathering can add to this effect. For example, application forms for assistance also contribute to survey fatigue.

One respondent provided the following comment on the Brisbane and Ipswich follow up survey
"I am tired of surveys. Please pass this info onto the relevant government bodies."

This position is understandable, as this household may already have completed forms from their insurance company, Centrelink, the Queensland Government, the Queensland Premier's Disaster Relief Fund Appeal, service organisations like the Red Cross, other research organisations, and two survey forms from Geoscience Australia. The problem is the lack of coordination between organisations about surveys for households. There is no "relevant government body" to pass the message on to and it is likely some that other organisation may still want to survey them in the future.

Figure 1: Complete rebuild of a house
following the 2011 floods. Before and after
rebuild images are shown.

A changed understanding of mental health

Our social survey capability is continually evolving based on responses from previous surveys. One example of this is the area of mental health. Rather than asking about mental health directly, Geoscience Australia used a subjective well-being test [1]. Subjective well-being is often used as part of quality of life assessments, and in some cases is referred to as a "happiness" test. The results from the subjective well-being questions showed that 9 per cent of respondents' well-being declined and 19 per cent experienced an improvement. While the subjective well-being measure can provide a more nuanced view of respondent's mental health, it appears to be better suited to interview-based data collection.

To supplement the subjective well-being section of the survey, Geoscience Australia asked whether household members felt anxious or nervous when it rained heavily. In the Brisbane and Ipswich follow-up survey 72 per cent of households said "yes". The stress of the 2011 floods has had an impact on many families. Some people said they now realise what is really important to them and are living in better ways. Others have become depressed or anxious and remain on medication to cope with their everyday lives. One respondent said that completing the survey was the first thing they had achieved in months. Some respondents said they were closer together as couples, yet many were torn apart, some permanently. There were two reported cases of a partner taking payouts and disappearing, leaving the remaining partner with a ruined house, the children and no cash. This sort of behaviour isn't very well studied in Australia but it is well documented overseas, along with dramatically increased reports of domestic violence [2],[3].

Figure 2: Modification of existing house as
a rebuild following the 2011 floods. Before
and after modification images are shown.

The rebuilding phase

The process of rebuilding has been difficult for some people. Some respondents spoke of insurance being perceived as a double edged sword. Negotiating with insurance companies and inflexible rebuilding arrangements were seen to cause their own problems. Some people felt that dealing with insurers and builders caused more stress than the actual flood event. There were also a number of reports of builders not completing work, price-gouging or undertaking substandard work. Many insurance companies repaired the house on a like-for-like basis, which did not allow home owners to make improvements that would decrease the home's vulnerability to future flood events (See Figure 1 for examples). However there are benefits to the extra cost during the rebuild, with 35 households in the follow-up survey reported that the mitigating actions undertaken post January 2011 resulted in their home sustaining little or no damage in the January 2013 flood.

Sixty-four per cent of respondents with full insurance had made no changes to their home, compared with 52 per cent of those with no insurance cover. There were reports of insurance premiums in Brisbane and Ipswich post January 2011 being four or more times higher than they had been, with many respondents quoting new insurance costs of $5,000-$6,000 a year, and some as high as $9,000. The initial Geoscience Australia survey undertaken in in 2012, identified that 19 per cent of respondents said insurance was either unaffordable or that they weren't eligible; and another 10 per cent did not seek insurance.

Figure 3: A like-for-like repair as a rebuild
response following the 2011 floods. Before
and after repair images are shown.

An evolving capability

This survey capability has provided new insights and enabled the development of a new understanding of vulnerability and resilience; however there is still a lot of work to do. Mental health is very important to recovery, but Geoscience Australia currently does not have the capability to study it appropriately. Recovery generally needs to be studied in a more holistic way, while focusing on the dimensions of housing, economic, social, and health (including mental health) and how they interact. Analysis of the survey data has shown that there is a potential disconnect between the warnings issued and people hearing them, but the reasons for this remain unclear. The choices that people make about post-disaster housing and how to help people who move into unfinished houses also needs further analysis.

References

[1] ^Cummins, RA 1997, Comprehensive quality of life scale - adult , 5th Edition (ComQol-A5), Australian Centre on Quality of Life, Deakin University, accessed 26/03/2012.

[2] ^World Health Organisation, 2002. Gender and Health in Disasters [PDF 730KB], Department of Gender and Women's Health, World Health Organisation, Geneva Switzerland, accessed 22 May 2008.

[3] ^Wilson J, Phillips BD and Neal DM 1998, "Domestic violence after disaster", in Enarson E and Morrow BH (eds), The gendered terrain of disaster: through women's eyes, Praeger Publishers, Westport, CT , USA, pp. 115-122.