New Zealanders’ Judgments of Earthquake Risk Before and After the Canterbury Earthquake: Do they Relate to Preparedness?

John McClure, Victoria University of Wellington

Celine Wills, Victoria University of Wellington

David Johnston, Joint Centre for Disaster Research, Massey University and GNS Science

Claudia Recker, Victoria University of Wellington

Previous research has examined judgments about earthquake likelihood after citizens have experienced an earthquake, but has not compared judgments in the affected region with other regions. Following the Darfield (Canterbury) earthquake, this research compared earthquake risk judgments in the affected region and those outside the region. Participants in Christchurch, Wellington and Palmerston North judged the likelihood of an earthquake before and after the 2010 Canterbury (Darfield) earthquake, near Christchurch. Wellington was chosen as there had been higher expectations of an earthquake in that area. Palmerston North was chosen to be comparable to Christchurch before the Darfield earthquake, in that many New Zealanders have expected an earthquake in Wellington, not Palmerston North. Participants judged earthquake likelihoods for their own city, for the rest of New Zealand and for Canterbury. Christchurch participants also indicated their preparedness before and after the earthquake. Expectations of an earthquake in Canterbury were low before the Darfield earthquake in all three regions and rose significantly after that earthquake. In contrast, Wellingtonians’ judgments of the likelihood of an earthquake in Wellington were high before the Darfield earthquake and did not rise after that earthquake. Christchurch participants’ risk perceptions showed only a weak relation to their preparedness. These results clarify how disasters such as major earthquakes affect judgments of earthquake risk for citizens inside and outside the affected area. The results show that these effects differ in cities where an earthquake is expected. Broader issues about preparing for earthquakes are also discussed.

This paper focuses on the Lichtenstein, 1982; Spittal, McClure, One factor affecting risk Christchurch earthquake in relation to Walkey, & Siegert, 2008). This is perception is personal experience; risk perception and preparedness. Risk partly because many citizens are usually, if people have a personal assessment is not the main factor in fatalistic and think that their actions experience of the hazards, they take preparedness; in fact, risk assessment will not make any difference the risk more seriously (Burger & is often a weak predictor of being (McClure, Allen, & Walkey, 2001; Palmer, 1992; Helweg-Larsen, 1999; ready or prepared, as in the case of McClure, Walkey, & Allen, 1999; Sattler, Kaiser, & Hittner, 2000). The preparing for an earthquake. Some Spittal et al., 2008; Spittal, Siegert, Police are very familiar with this, and research finds no relationship between McClure, & Walkey, 2002). However, say that the comment they hear most the two (risk assessment and recognition of the risk is a prerequisite often after accidents is: “I never preparing) (Cowan, McClure, & for voluntary action, and unless people thought it would happen to me.” This Wilson, 2002; Hurnen & McClure, recognise the risk, they are unlikely to effect interacts with a second bias 1997; McClure, Sutton, & Sibley, take action. where people think disasters are going 2007; Slovic, Fischhoff, & to happen to other people, not to

New Zealand Journal of Psychology Vol. 40, No. 4. 2011 • 7 • Judgments of Earthquake Risk Before and After the Canterbury Earthquake themselves (Burger & Palmer, 1992; Helweg-Larsen, 1999; Mileti & Darlington, 1995; Spittal, McClure, Siegert, & Walkey, 2005; Weinstein, 1980).

The Christchurch Risk Study To clarify these risk biases, this paper reports a study that we carried out after the first Canterbury earthquake, the Darfield earthquake in September 2010 (McClure, Wills, Johnston, & Recker, 2011). We were interested in how people in Christchurch, Wellington and Palmerston North changed in their perception of risk of a future earthquake after a significant local earthquake. acquaintances in Christchurch.  Figure 1. The perceived likelihood Christchurch participants were also of an earthquake occurring in or near The questionnaire asked for asked about their preparedness for an Christchurch before and after the Christchurch citizens’ recall of their earthquake. pre-earthquake risk perception: Canterbury Earthquake. (1= not at all likely, 5 = very likely) “Before the Darfield earthquake, how The results have been published in probable did you think it was there McClure et al. (2011). In judgments of These judgments of earthquake would be a big earthquake in or near the likelihood of an earthquake in likelihood before the earthquake are Christchurch?” A second question Christchurch before and after the recall data and thus subject to recall asked: “Since the Darfield earthquake, September earthquake, likelihoods effects, but they are consistent with how probable do you rate a future rose significantly in all three cities data collected before the earthquake earthquake in Christchurch?” The (Figure 1). There was a main effect for (Becker, 2010). same questions were asked in the ‘before and after the earthquake’ Wellington and Palmerston North. time factor, which shows that it was a With regard to the question: “If similar effect across all three cities. you’ve previously thought an Questions also asked Wellington Time showed that these likelihood earthquake in or near Christchurch and Palmerston North participants for judgments were correct; after the first was unlikely, why was that?” the most their recall of the likelihood of an earthquake (i.e., after September frequent reply was “because earthquake in their own city – and in 2010), there was another big Christchurch is not on a fault line”. any other part of New Zealand (NZ). earthquake in Canterbury in February Most Cantabrians are aware of the Questions then asked for their 2011. Alpine Fault, and they thought an judgments of the likelihood of a future earthquake was more likely near the earthquake in each of these three areas Interestingly, Wellingtonians Southern Alps or the Alpine Fault that (Christchurch, their own city, and expected an earthquake in Wellington runs down those Alps. They assumed another part of NZ). Judgments were prior to the earthquake more than there were no fault lines near given on Likert scales. A related Cantabrians did for Canterbury, and Christchurch. Secondly, they said that question asked “If you’ve previously that expectancy in Wellington showed there have not been any earthquakes in thought an earthquake in or near no change after the earthquake in this region before (some qualified this Christchurch was unlikely, why was Christchurch. So Wellingtonians did view with the term ‘recently’). They that?” with space for open ended not think an earthquake in Wellington found out after the earthquake that responses that were coded by two was any more likely after the they were wrong: there had been coders. September earthquake; and they were earthquakes in the region. Many New correct. Zealanders, including Cantabrians, did Questions on other issues asked not know that the spire of “Were you aware of information and In contrast, for Palmerston North, Christchurch Cathedral, which warnings to prepare prior to the the rise in their expectancy of an collapsed in the February earthquake, earthquake?” and “did you see this earthquake looks modest (Figure 2) had been knocked down by information as relevant to you?” and but is statistically significant. Thus earthquakes twice before, in 1888 and for Christchurch citizens “Did you Palmerston North citizens saw a 1901. The Cathedral was damaged suffer serious damage to your home or likelihood of an earthquake in less significantly by earthquakes in not”; and for Wellington and Palmerston North (and also in another 1881 and 1922. As this all happened Palmerston North citizens, whether part of New Zealand) as more likely some time ago, people had either they had friends, family or close after the Darfield earthquake. never known about it or forgotten.

• 8 • New Zealand Journal of Psychology Vol. 40, No. 4. 2011 McClure, Wills, Johnston & Recker

that it is not just Wellington that is at risk, and fortunately New Zealand’s building standards partially reflect this expert knowledge.

What Is Preparedness and How Do We Increase It? There are three strands to action: legislation, incentives, and personal voluntary action.

Legislation The first is legislation. New Zealand does have building regulations for new buildings and, positively, New Zealand is a relatively non-corrupt country that enforces these regulations. As a consequence, information as relevant to you”, they there are not 2,000 or 10,000 people  Figure 2. The perceived likelihood said they thought it wasn’t relevant to dead following the Canterbury of an earthquake in participants’ own them because the earthquake was earthquakes. Some countries have the city before and after the Canterbury going to be in Wellington. same building regulations as New Earthquake in Wellington and Zealand, but in earthquakes most of Palmerston North. (1= not at all likely, People who have suffered harm or the buildings collapse due to 5 = very likely) damage in an accident such as a car corruption that has resulted in building Thirdly, they thought that accident usually see the future risk of standards not being enforced. The Wellington (or the North Island) was that hazard as higher. But in this case, New Zealand regulations are also at greater risk. And perhaps that was when asked “Did you suffer serious being steadily upgraded. No accurate in terms of probabilities. But damage to your home or not”, commercial building built since 1985 events don’t always follow Christchurch citizens who suffered killed anybody, and nor did almost probabilities. damage did not see the future risk any private homes. In New Zealand, from earthquakes as higher than those commercial buildings are made so that This judgment reflects an who had not experienced major they won’t kill people in an interesting pattern. It is as if damage. This differs from the usual earthquake, and those built since 1985 Christchurch people thought that finding and may reflect a ceiling did not. However, perhaps they should Wellington is more likely to have the effect. One novel finding is that be built as in Japan, so that they will earthquake, and therefore they don’t Wellington and Palmerston North still be useable after the earthquake. It (or didn’t) see a need to prepare in citizens with friends, family or close would only cost 5-10% more in Christchurch. This view suggests a acquaintances in Christchurch saw the construction costs. Some old buildings dangerous leap in people’s thinking, risk of another earthquake in in Christchurch had been analogous to middle-aged people Christchurch as higher. This finding is strengthened, including both thinking: “teenagers more likely to interesting, because it suggests that if Christchurch Cathedrals. They are have car accidents, so therefore I don’t people have an emotional bond to now badly damaged by the need to drive safely.” someone in Christchurch, they see the earthquakes, but if they had not been risk in Christchurch as higher. From a strengthened, they would have been Often the perceived likelihood of personal perspective, having a like other historic buildings that are earthquakes does not relate to daughter in Christchurch who works now a pile of rubble. preparation. However, in this study, every Saturday in an Addington café there was a weak but significant that is made of brick, which Engineers knew that there is a lot relationship between Christchurch fortunately has been strengthened, one of soft soil in the East of the city, and people’s recall of the likelihood of an authour can understand that. that if Christchurch were to have an earthquake before the Darfield earthquake, there would be earthquake and their preparation. On In summary, these data offer some considerable damage in these areas. the question “Were you aware of lessons for preparation strategies and Articles and TV documentaries in the information and warnings to prepare lessons for the media. First, it is not 1990s reported this risk. New prior to the earthquake”, there was no just Wellington that is at risk of an Zealanders cannot leave this issue to difference between the three cities. earthquake. New Zealanders have a Councils; this is too dangerous, as Interestingly, Christchurch people said fixation on Wellington’s risk; this is Councils permit buildings on soils that they knew all about the warnings, but analogous to adults who think that are likely to liquefy in an earthquake. on the question “Did you see this young people are the only ones with New Zealand may not have high alcohol problems. Seismologists know levels of corruption, but Councils can

New Zealand Journal of Psychology Vol. 40, No. 4. 2011 • 9 • Judgments of Earthquake Risk Before and After the Canterbury Earthquake obviously be pressured by developers time bombs in this country, they may include strengthening buildings, and to permit building on unsuitable land. find to their regret it is likely to replacing or reinforcing brick or New Zealand therefore needs a collapse on them in an earthquake. unreinforced masonry. An example is national regulation stipulating that if the Hunter building at Victoria people are building on soft soil or sand Incentives University. The University they need better foundations. In addition to legislation, another strengthened this quaint old building, Engineers could provide a formula for strategy is to use incentives, as giving and put concrete and steel bracing this; indeed, this principle is already people information is not enough to inside it – it is like inserting a being applied in the rebuild of get them to prepare. Many insurance backbone in a jellyfish. Yet despite the Christchurch, where there are different companies did not apply this principle importance of these damage building specifications for sites with to houses. If people want to drive a mitigation actions, preparedness different soil composition. dangerous motorbike or are under 25, messages focus almost solely on they pay more insurance, but at least survival actions. On the day after the Legislation could also require before the Christchurch earthquakes, September earthquake, the Dominion Councils to make the earthquake people could have a house that was Post newspaper ran a big feature status of buildings public. Two totally unsafe, but pay no more saying: “Have you got your councils (Hastings and Timaru) tried insurance. When the disaster happens, emergency kit, etc.” One author wrote to conceal and withhold the known the insurance companies must recover a letter saying that this is good civic earthquake state of local buildings losses or go bankrupt, so everyone duty you are performing, but it would after the Canterbury earthquakes. must pay (much) more insurance. be useful to also remind people that They justified this action by saying Insurance could be targeted, or be they need to have a builder check their that the information would make more conditional. For example, if a house. The newspaper duly printed the people panic. However, their lawyers house owner has a brick chimney, letter, but when the next earthquake told them that they legally had to instead of the usual $400 premium, happened on 22 February, the next reveal this information. In Wellington, they might have to pay $800 a year, or day, the Dominion Post repeated the some buildings in the city have been have a higher excess. mantra “Have you got your emergency ‘red stickered’, with a notice on the kit, etc.” The lesson about the window announcing that the building importance of actions to mitigate Personal readiness is a dangerous earthquake hazard. This damage was not learned. There needs is useful, but the notices are small, A4 A key issue is focusing not just on to be a shift to focus on prevention, size, and given the hazard, they should response and recovery, but on not just survival. be a metre wide. The message should readiness. The concept of civil be strong. Legislation requires these defence is based on the analogy with When an earthquake happens, the messages to be prominent on cigarette military defence. Clearly, to have an importance of building strength and packets. New Zealand also requires effective military defence, if people at soil type is obvious. People need to more retrofitting of old buildings. It is war are facing guns and tanks get a builder to check their house or expensive. But if, after Hawkes Bay, shooting at them, the best defence is chimney. Our questionnaire found that Christchurch had just strengthened not just an emergency kit to patch only two out of 200 had done this. We just four buildings a year, many of them up after they’ve been injured or have mentioned the analogy with those lovely old brick buildings in maimed. For readiness and good military defence. After the Canterbury Canterbury would have survived, as defence, people need armour that earthquakes, New Zealand had a would their inhabitants. The retrofitted protects them from being injured or window of opportunity where people buildings are often not as strong as killed. In civil defence, people need in other cities were buying more new buildings, but many of those in readiness as well as response and bottles of water and other actions. But Christchurch that had been recovery. Civil Defence did not use this mild strengthened survived the earthquakes. anxiety to get people to prepare. That An important issue underpinning seems to be a missed opportunity. Is Another useful, low-cost this concept of defence is that there not this what these agencies are being legislative requirement would be to are different types of preparedness. paid for? As a consequence, we have put the earthquake rating of houses on Having an emergency kit and water is buildings like the Dean’s beautiful the title. New Zealand has warrant of one class of action, and in Wellington, house, with three layers of brick, fitness requirements for cars but not this may be more important than destroyed in the earthquake. The for houses that are worth about Christchurch, because there are fewer earthquake damage has cost 15 billion $400,000 each on average. So New access routes into Wellington, and dollars. Preparedness in the form of Zealand can do more with legislation. Fonterra (a large milk supplier) may strengthening buildings would have A warrant of fitness requires not convert its milk trucks into water been expensive, but if builders had mechanics to check numerous safety trucks to rescue Wellingtonians. Thus, strengthened all the brick and features, but anyone can buy a brick survival actions such as compiling an unreinforced masonry buildings in house that will collapse like a pack of emergency kit are important. Christchurch, this would have cost cards in an earthquake. If an However, actions to mitigate or much less than 15 billion dollars. It’s Australian comes to New Zealand and prevent damage made in advance of not surprising that Canterbury people doesn’t know that brick houses are the quake are also important. These now think New Zealand should have

• 10 • New Zealand Journal of Psychology Vol. 40, No. 4. 2011 McClure, Wills, Johnston & Recker stricter building regulations (Mitchell, McClure, J., Sutton, R M., & Sibley, C. G. Kahneman, P. Slovic, & A. Tversky 2011). (2007). Listening to reporters or (Eds.) Judgment under uncertainty: engineers: How different messages Heuristic and biases (pp. 463-492). about building design affect earthquake Cambridge: Cambridge University References fatalism. Journal of Applied Social Press. Becker, J. S. (2010). Understanding Psychology, 37, 1956-1973. disaster preparedness and resilience in Spittal, M. J., McClure, J., Siegert, R. J., & Canterbury: Results of interviews, McClure, J., Walkey, F., & Allen, M. Walkey F. H. (2005). Optimistic bias in focus groups and a questionnaire (1999). When earthquake damage is relation to preparedness for survey. GNS Science report 2010/50. seen as preventable: Attributions, locus earthquakes. Australasian Journal of of control and attitudes to risk. Applied Disaster and Trauma Studies, 2005-1, Burger, J. M., & Palmer, M. L. (1992). Psychology: An International Review, 1-10. Changes in and generalization of 48, 239-256. unrealistic optimism following Spittal, M., McClure, J., Walkey, F., & experiences with stressful events: McClure, J., Wills, C., Johnston, D., & Siegert, R. (2008). Psychological Reactions to the 1989 California Recker, C. (2011). How the 2010 predictors of earthquake preparation. earthquake. Personality and Social Canterbury (Darfield) earthquake Environment and Behavior, 40, 798- Psychology Bulletin, 18, 29-43. affected earthquake risk perception: 817. Comparing citizens inside and outside Cowan, J., McClure J., & Wilson, M. the earthquake region. Australasian Spittal, M. J., Siegert, R. J., McClure, J., & (2002). What a difference a year Journal of Disaster and Trauma Walkey, F. H. (2002). The Spheres of makes: how immediate and anniversary Studies, 2011-2, 1-10. Control scale: The identification of a media reports influence judgments clear replicable factor structure. about earthquakes. Asian Journal of Mileti, D. S., & Darlington, J. D. (1995). Personality and Individual Differences, Social Psychology, 5, 169-185. Societal response to revised earthquake 32, 121-131. probabilities in the San Francisco Bay Helweg-Larsen, M. (1999). (The lack of) area. International Journal of Mass Weinstein, N. D. (1980). Unrealistic optimistic bias in response to the Emergencies and Disasters, 13, 119- optimism about future life events. Northridge earthquake: The role of 145. Journal of Personality and Social personal experience. Basic and Applied Psychology, 39, 806-820. Social Psychology, 21, 119-129. Mitchell, J. (2011). Community insights on events from 4th to late September 2010. Authors’ Note Hurnen, F., & McClure, J. (1997). The Report to the Regional Emergency effect of increased earthquake Management Office. Canterbury The empirical results included in knowledge on perceived preventability CDEM Group. this paper have previously been of earthquake damage. Australasian reported in McClure, Wills, Johnston, Journal of Disaster and Trauma Sattler, D. N., Kaiser, C. F., & Hittner, J. & Recker, (2011) and the figures are Studies, 3. 1-10. B. (2000). Disaster preparedness: reprinted here with permission from Relationships among prior experience, Australasian Journal of Disaster and McClure, J., Allen, M. W., & Walkey, F. personal characteristics, and distress. Trauma Studies. H. (2001). Countering fatalism: Causal Journal of Applied Social Psychology, information in news reports affects 30, 1396-1420. judgements about earthquake damage. Correspondence should be Basic and Applied Social Psychology, Slovic, P., Fischhoff, B., & Lichtenstein, addressed to John McClure at: 23, 109-121. S. (1982). Facts versus fears: [email protected] Understanding perceived risk. In D.

New Zealand Journal of Psychology Vol. 40, No. 4. 2011 • 11 •