Modelling the social and economic impacts of a volcanic eruption in Auckland


By Robert Cardwell

June 2021

A research team led by University of Auckland PhD candidate and Market Economics researcher Robert Cardwell recently published new research on the simulation of economic and urban development recovery pathways following the eruption of a new volcano in the Auckland region.

The research, published in the Journal of Volcanology and Geothermal Research, demonstrates the capability to simulate not only the direct physical impacts associated with infrastructure and land use destruction and disruption, but also population and business evacuation and relocation. The team also modelled the socio-economic and land use planning implications associated with recovery and also the longer-term growth-related pressures on city development.

Figure 1 shows the impact on value added (the counterpart to Gross National Product at the regional level) of the volcanic eruption scenario over a 20 year period following the eruption compared to a baseline scenario where the eruption did not occur.


Figure 1: Eruption and Baseline integrated scenario total value added for industrial and commercial sectors

Figure 2 depicts the difference in development of residential and industrial land use that has occurred by September 2040 in the volcanic eruption scenario compared to the baseline scenario under a ‘Fast Recovery’ scenario.


Figure 2: Change in residential and industrial land use between 1st October 2019 and 30th September 2040 in the ‘Fast Recovery’ scenario

Figure 3 depicts the difference between two scenarios at September 2040 where in one scenario the land physically impacted has been remediated within a timeframe of 5 years, and another in which the remediation has taken 10 years.


Figure 3: Land use in ‘Fast Recovery’ scenario as at 30th September 2040 vs ‘Slow Recovery’ scenario as at 30th September 2040

The differences between the ‘Fast Recovery’ and ‘Slow Recovery’ scenarios demonstrate how taking longer to remediate the area directly impacted by the volcanic eruption can result in more urban development on the urban-rural boundaries than might have otherwise occurred.

As far as the authors are aware, this research is the first instance of simulation of land use and economic changes over time after a natural hazard event has occurred. The developments demonstrated in this research will enable planning authorities to assess trade-offs with competing objectives in mitigation and redevelopment strategies to respond to volcanic eruptions and other natural hazard events.

The research was co-authored by Robert’s supervisors Assoc Prof Liam Wotherspoon (University of Auckland) and Dr Garry McDonald (Market Economics), with support from Prof Jan Lindsay (also University of Auckland). Funding was provided by the Resilience Challenge Economics and Urban research programmes with support from DEVORA (Determining Volcanic Risk in Auckland).


Challenge research used in
fuel security report

By Dr Nicola McDonald, Market Economics

In September 2017, the pipeline that distributes petroleum-based fuels from Marsden Point Oil Refinery to Auckland ruptured, halting the delivery of fuels into Auckland for 10 days and instigating a major rationing response for jet fuel usage at Auckland Airport, with significant disruption to air transport. This incident was not unforeseen: as part of repeated reviews of the nation’s fuel security the government had previously commissioned compilation of a set of potential supply disruption ‘scenarios’ to inform resilience planning. A pipeline disruption of the type and duration of the 2017 event had indeed been considered among these scenarios.


The pipeline connecting Marsden Point Oil Refinery and Auckland Airport was breached in 2017. Photo: Refining NZ

To assist the government in further resilience planning, and to provide information to support the ensuing Government Inquiry into the Auckland Fuel Supply Disruption, ME Research (assisted by Resilient Organisations) was commissioned to undertake an analysis of the economic consequences of the fuel supply disruption scenarios, including consideration of the benefits of investing in mitigation measures such as additional storage facilities. Although economic analyses of the disruption scenarios had been undertaken previously, this study was unique in that it utilised a systems-based approach involving the identification of cascading consequences of fuel disruptions and utilised MERIT (Measuring the Economics of Resilient Infrastructure Tool), a decision support system that has been partly developed and extended within the Resilience Challenge.  The final report Economics of Fuel Supply Disruptions and Mitigations considered five of the disruption scenarios, focusing particularly on those that would impact the Auckland region.[1] The report also provided commentary on how this type of scenario-based analyses might sit within a policy/planning framework and the points of difference between the economic modelling and common cost-benefit type approaches.

The Inquiry Members drew on the findings of this report, in particular noting limitations in the application of cost-benefit analysis to evaluate ‘resilience’ investments. The Inquiry Members also noted the report’s findings that the main contributions of scenario-based economic impact analyses are to (1) reduce uncertainty by helping to explore the causes, effects, and potential outcomes of disruption events; and (2) provide a common focus for debate on the appropriate level of societal investment in mitigation steps. Various measures to improve Auckland’s fuel security were suggested by the Inquiry Members. [2]



Student profile: Cuong Nguyen





A bit about me 


I was born in Vũng Tàu city, Vietnam. My hometown is a beautiful coastal city in the south-east of the country. With its golden sandy beaches and inactive volcanic mountains, my little city draws many visitors from Ho Chi Minh City who can arrive by hydrofoil. In 2008, I moved to London to complete a B.Sc. and M.Sc. in Economics at Queen Mary University of London and worked as nail technician and assistant accountant. In 2014, I went back to Vietnam to work at Vietnamese-German University in Bình Dương.

During my childhood, I enjoyed studying geography and watching disaster-related movies, especially the Godzilla series. From that knowledge, I learnt more about how public institutions respond during critical times and how communities resist and recover after a catastrophic event.

On top of that, I frequently saw sad news about the effects of storm surges, floods and landslides in Vietnam and Japan. I felt I needed to be more informed and contribute my knowledge to enhancing society’s resilience. As a result, in September 2016, I decided to pursue a PhD in Economics under the supervision of Professor Ilan Noy (Chair in the Economics of Disasters) at Victoria University of Wellington.

In my spare time, I enjoy swimming and playing tennis. Currently, I am in the top 1000 tennis players in New Zealand, based on interclub ranking.


Cuong at Mount Victoria summit and the Wellington waterfront

My project


Through my PhD, I work on several projects regarding earthquake insurance and the property market. In New Zealand we have the privilege of having access to rich datasets from many different sources, which makes it possible to answer many interesting research questions. My focus is on the 2010 -2011 Canterbury Earthquake Sequence (CES). Using the insurance claim payment data from EQC, I examine the role of insurance in residential recovery in Greater Christchurch. I also compare the functions and capability of different earthquake insurance schemes for countries that have similar earthquake risk profiles to New Zealand. In addition, I investigate homeowner’s decisions regarding the Crown’s offers for houses in the residential red zones in Christchurch. I’m currently working on investigating the change in risk perception for property markets post-CES and the effect of coastal hazards on Kāpiti Coast property sale prices.

Local councils and Crown research institutes are interested in my projects and have been very supportive during my research process. I feel that my research outcomes can contribute to their decision-making around natural hazards and resilience. The PhD program provides me an opportunity to enhance my skill set; I have learnt to use different software and programming languages such as ArcGIS, QGIS, Stata, R and Python. For example, I extracted nightlight intensity from NASA satellite imagery to use in one of my research projects.


Night-time light imagery for Greater Christchurch

As part of the Resilience to Nature’s Challenges Economics toolbox, my supervisor and I are collaborating with Resilient Organizations and Wellington Region Emergency Management Office (WREMO) on a research project developing resilience indicators for the Wellington region. As part of this project, I am currently gathering resilience metrics from various ministries and local councils.


Next steps


When starting the PhD program, I did not foresee how interesting the process would be. After two years pursuing the degree, I still think about many ideas for potential research projects that would serve both academic and applied purposes.

My next step is to finish writing my current papers. After that, I will work on one or two additional research projects before my PhD completion. I plan to submit my thesis by summer 2019 and continue my journey in disaster research.


The Wellington Lifelines Regional Resilience Project



By Nicky Smith

How can co-ordinated infrastructure planning and investment improve the resilience of New Zealand’s capital city, Wellington, to earthquake events? To what extent can targeted infrastructure investments reduce the economic consequences experienced by Wellington, and the rest of New Zealand, should a large-scale event occur? These were among the key questions considered by a group of researchers, including members of the Resilience to Nature’s Challenges Economics Toolbox, in the Wellington Lifelines Regional Resilience Project.
The Wellington Fault running through Thorndon, Kelburn and Wellington city

Wellington’s vulnerability to a major earthquake event is well known, with a probability of a local magnitude 7.5 or greater earthquake of around 20% over the next 100 years. Evidence from New Zealand and abroad indicates that when key infrastructure is out or operating at degraded levels of service following a major hazard event, people leave, productivity drops, and communities and the economy suffer. What is more, utilities are also strongly interconnected, with the operation and restoration of one infrastructure often critically dependent on one or more other infrastructures. Despite these interdependencies, utilities have historically planned their resilience investments independently, often leading to diluted assessment of resilience arguments and sub-optimal investment outcomes.

Recognising the deficiencies in routine approaches, the Wellington Lifelines Regional Resilience Project has taken a step forward in integrated resilience planning. Drawing on knowledge from all 16 Wellington utility providers and local government, it identified and presented a preferred, accelerated, programme of infrastructure investments for Wellington. The project was supported by GNS Science’s RiskScape and Post Disaster Cities (PDC) teams who modelled and mapped infrastructure service outages and restoration after a M7.5 Wellington Fault event, and its associated perils (fault rupture, ground shaking, liquefaction, landslides, lateral spreading, subsidence).  The likely scale and timeframe of infrastructure service losses under current investments were compared with that which might occur should further investments be made in infrastructure resilience.

The outcomes from this work then fed into modelling performed within MERIT (Measuring the Economics of Resilient Infrastructure Tool), a novel economic simulation tool developed by Market Economics, Resilient Organisations and GNS Science under the Economics of Resilient Infrastructure research programme and being extended under Resilience to Nature’s Challenges, that enables high-resolution assessment across space and through time of the economic consequences of infrastructure failure, business response and recovery options.  With the preferred programme of investments, the MERIT modelling demonstrated a multi-billion-dollar reduction in economic losses and improved community outcomes in the event of a major seismic event in Wellington.  In addition, the investment programme was also recognised to have co-benefits in reduction of losses and faster community and economic recovery arising from smaller earthquakes, and from threats arising from other perils.

Linkages between the various stages of damage loss assessment and economic impact analysis for the Wellington Lifelines Regional Resilience Project

The Wellington Lifelines Regional Resilience Project is an excellent example of multi-stakeholder co-ordination for a common goal and demonstrates a positive step forward in resilience planning in New Zealand.

Further information and demonstration of the MERIT tool is available at:

Building the resilience of local economies to the flow-on impacts of natural hazard events



By Emily Harvey, Stuart Mead and Morag Ayers 

When a natural hazard event occurs, there are many direct impacts on people, infrastructure, and businesses in the area. These direct impacts often lead to substantial indirect impacts on other businesses in the area and further afield when goods or services are suddenly not able to be produced, purchased or transported. Using the Waikato Region as a case study, researchers in the Resilience to Nature’s Challenges Economics programme are investigating how the interconnections and interdependencies in the economy affect how these flow-on impacts propagate. Finding out what causes the effects to be amplified or reduced will help to inform resilience building strategies.
5 January 2018 storm event near Thames. Photo: Nick Van’t Wout

Co-creating natural hazard scenarios


In the first stage of this project we held a workshop process with stakeholders around the Waikato Region, including local and regional council members, CDEM, and the Waikato Lifeline Utilities Group. Within this process, researchers were able to listen to stakeholders and work out the key social, economic, cultural, and environmental impacts of natural hazard events that they were concerned about. From this workshop two natural hazard scenarios were selected based on the type of impacts they would have, the availability of data to model them, and the level of concern of stakeholders. The selected events were: a 6.9M Kerepehi Central fault earthquake that produces strong shaking in the Hauraki Plains region, and moderate shaking in other Waikato regions, and an extra-tropical cyclone storm event that produces a storm surge, high winds, and extreme rainfall.


Waikato region earthquake scenario from a Kerepehi Central fault

Understanding interconnections in local economies


In the modern economy, infrastructure, businesses, and communities in local regions are interconnected in increasingly complex ways. The flow-on effects of a natural hazard event depend critically on the structure of the local economy and the infrastructure network, as well as on the interconnections between local economies. As local economies become increasingly complex, these impacts become increasingly challenging to analyse, with the potential for unforeseen consequences. In this project we are looking at local economies at the Territorial Authority (TA) level. Using datasets called Input Output Tables, we have quantifications of the flows (in monetary terms) between all the industry sectors within and between the Waikato TAs, as well as to the rest of New Zealand and the rest of the world (imports and exports). An example of the interconnections you see in this data when you look at the economy as a network is shown below.


Example of the network you would produce from the Input-Output table data for the Waikato region (Region 1), north of the Waikato (Region 2), and south of the Waikato (Region 3) with industry sectors aggregated into primary (Ind1), manufacturing (Ind2), and services (Ind3)

Disrupting local economic networks


In this project we are developing a model that considers the disruptions to industries and regions that are up and downstream from direct impacts, and incorporates potential adaptations (e.g. sourcing goods from outside the region). We will use the two co-created natural hazard scenarios to investigate how impacts propagate following a natural hazard event in the Waikato region. By looking at a detailed local economy structure level, this work will help to identify consequences that are not directly obvious. We will also test artificial isolated disruptions to individual industries and industry connections to identify critical sectors or links.




Through this research we will identify the characteristics of local economies that enhance (or reduce) their resilience to natural hazards, as well as identifying ‘hot spots’ (critical sectors or critical links) within the economy that amplify disruptive impacts. Within the Waikato Region, the TAs range in structure and include some that are dominated by agriculture, some tourism driven, and the urban economy in Hamilton city. By looking at the impact that disruptions have in a range of both rural and urban local economies, our results will be applicable around New Zealand more broadly.

An economic investigation of disaster insurance



By Ilan Noy

The Christchurch Earthquake of 2011 was the most insured earthquake ever, anywhere, with 98% of Christchurch properties being insured when the quakes occurred (Nguyen and Noy, 2017). In comparison, a recent New York Times article stated that only 13% of houses in California are currently insured. Yet the public’s view of the way our public-private insurance system functioned in the aftermath of the quake is dim, at best.

What were the impacts of all this insurance cover during the recovery in Christchurch? How will our insurance system deal in the future with weather risk as climate change begins to bite more? And what can we do to fix any problems? These are some of the questions being asked by Resilience to Nature’s Challenges’ Economics Toolbox researchers. 

Damage being assessed after the 2011 Canterbury quake

In July 2018, the Earthquake Commission (EQC) reported that it was still processing 3,476 claims for damage from the 2010-2011 Canterbury Earthquake Sequence (CES). During that month, it closed 811 claims, but also received 767 new claims associated with the CES (some for previously completed repairs). So, in July, the net reduction in the number of open claims was 44. Yet, the CES was the most heavily insured event in history, and as such our insurance system managed to shield New Zealand from much of the financial consequence of the earthquakes. The cost of the earthquake recovery was largely transferred overseas (to insurance companies across the Tasman, and further afield to reinsurers in London and other international financial centres).

When insurance functions properly it manages to transfer the financial risk from individuals, families, and companies, to the financial markets. In addition, the best disaster insurance systems also manage to incentivise risk reduction ex ante and to speed up recovery ex post. It should also encourage and enable investment in productive opportunities and protect the most vulnerable in our society from falling into poverty in the aftermath of an event. While our insurance system did achieve its first goal, it delivered less successfully on the other four. 

In the first years of the Resilience Challenge, our team (Ilan Noy, together with post-grad students Cuong Nguyen, Sally Owen, Jacob Pastor-Paz, Polly Poontirakul, and Belinda Storey) has documented how and when our insurance system does and does not deliver on its promise. This research agenda included examining both the Earthquake Commission insurance, the private residential insurance that sits on top of the EQC cover, and commercial insurance (both for commercial property and business interruption).

For example, we documented that EQC insurance payments assisted in the recovery of residential areas of the city, which was measured by night-time luminosity as captured by satellites. More surprisingly, this research also documented that cash payments were more conducive to local economic recovery, as measured by nightlights, than house repairs done through the managed repair programme (though we do not know whether the damage was repaired with this cash). In another research project, we documented that firms that had insurance and were paid promptly recovered much better than both firms with no insurance and firms that had insurance but whose claim payments were delayed. Unexpectedly, we found that there was no observable difference in recovery trajectories between those firms that had no insurance at all, and those who did, but whose payments were belated.


Average annual night-time light in 2013 at the area unit level from Nguyen and Noy (2018)

In a remarkably less lauded government initiative, the presence of so much insurance helped us (tax and rate payers) by funding most of the costs associated with the Residential Red Zone programme. The programme was the largest ‘managed retreat’ effort undertaken in a developed country to reduce risk exposure, through which residents were ‘bought out’ from risky Red Zone areas of the city. We are now examining the choices Red Zone property owners made, and determining what lessons we can learn from that programme for the future implementation of managed retreat plans elsewhere.


Who pays for managed retreats?

Based on the insights gained from this research, we recently made two submissions to the government for its Insurance Law Review. These submissions were written together with Roger Sutton, ex-CEO of CERA, Leanne Curtis, director of Breakthrough Services in Christchurch, and Samuel Becher, a Professor of Consumer Law. In the submissions, we advocated for several changes and additions to the Insurance Law; changes that we think will make New Zealand more ‘resilient to nature’s challenges.’

The importance of personal relationships in risk mitigation:
Meeting with business owners in North Canterbury



By Robert I Radics and Linh N K Duong 

Rural value chains are rich in personal relationships and these play a key role in risk mitigation.

A team of researchers have been interviewing local businesses in North Canterbury to explore the resilience of rural supply chains to natural hazards as part of Resilience to Nature’s Challenges. Coincidentally, the magnitude 7.8 November 2016 Kaikōura earthquake occurred during the study period, providing an opportunity to investigate how this event affected businesses in the area.

A value chain is the process by which a business adds value to produce its product. Generally, it involves the business receiving raw materials, adding value to the raw materials through various processes to create a finished product, and then selling that finished product to customers. On the other hand, a supply chain is the system of organizations, people, activities, information, and resources involved in moving a product or service from the supplier to the customer.

Many North Canterbury businesses’ supply chains were disrupted by the 2016 Kaikōura earthquake when State Highway One (SH1) was closed between Seddon and Cheviot. The road was closed for just over a year, and the main railway line was closed for ten months (some services are still not restored).

Researchers have been exploring the effects of the disaster on the larger economy, as well as the recovery mechanism. This work is an in-depth analysis, looking at how a collapsed or partly functioning value chain affects the individual businesses and their value chains, and how we could mitigate the risks of a bullwhip (increasing swings in inventory in response to shifts in customer demand), snowball (propagation and amplification of disruptions) or ripple effect, for example.

The team is using a bottom-up approach to explore how the networks of rural supply chains, their connections with rural communities, infrastructure and dependency on partners further afield, were affected by the 2016 Kaikōura earthquake.

The bottom-up approach is useful as it gives us the ability to identify different behaviour and bounce back mechanisms of affected businesses after the quake. We have been able to compare behaviour and recovery of businesses on SH1 in North Canterbury, which were essentially isolated after the road closure, with businesses on the bypass roads, which a higher level, bird-eye view would not have been able to do.


The centre of Cheviot before the reopening of SH1 in 2017 – close to zero traffic



We met with business owners and managers in North Canterbury and asked them about their supply chains and how they were affected by the quake. The interviewees were generous with information that went beyond the answers of our questions, and the stories and perspectives shared resulted in interesting and rewarding conversations.

As part of participating, each interviewee will receive a map of their value chain and evaluation of its strengths and weaknesses. This will include identification of opportunities in the business’ supply chain management for improvement, with accompanying benchmarks.


Example of a value chain map

Supplier information

  Your company score Industry average District average
Reliability 9.0 N/A* 8.4
Communication 6.7 N/A* 7.7
Personal relationships 5.3 N/A* 7.1
Willingness to compensate 4.3 N/A* 6.3
Easy to do business without this supplier 7.3 N/A* 5.3
Transparency of the supply chain 4.7 N/A* 6.0

Client information

  Your company score Industry average District average
Reliability 10.0 N/A* 8.8
Communication 10.0 N/A* 8.1
Personal relationships 6.7 N/A* 8.2
Willingness to compensate 10.0 N/A* 7.0
Easy to do business without this client 3.0 N/A* 5.6
Transparency of the supply chain 3.3 N/A* 5.4

* More information on industry averages will become available as data collection continues.

Examples of part of the feedback report.




The Kaikōura event was not the direct focus of these discussions, but talking about it was unavoidable. Farmers and service providers exposed to the event had improved their preparedness since, and were more resilient than those who had not personally experienced a disaster. Some of the preparedness activities that they had undertaken included increasing inventories, improving partnerships and in some cases adding infrastructure to substitute the centralised systems like generators and waste water cleaning systems.

The togetherness and interdependence in the North Canterbury community was evident. After the earthquake, farmers checked on their neighbours personally if they were not available via phone, and provided immediate help to each other. Employers called and visited employees who were potentially affected by the quake too.

During and after the natural disaster event, most of the partners across the value chain were ready to offer help to others far beyond their contractual obligations. It seems as though a commitment beyond the business’ interests comes into play when a community needs immediate help.

Data collection and linking the nodes in the complex value chain network will continue until early next year. We will send another information package to our interviewees with a picture and description of their extended value chain before the final report and publications are produced. This transparency and information flow will give participants another layer of knowledge and certainty, as well as providing a risk mitigation opportunity for them