Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

Preparing for High-impact,

Bernice Lee and Felix Preston, with Gemma Green

Low-probability Events
Lessons from Eyjafjallajökull
A Chatham House Report
Bernice Lee and Felix Preston, with Gemma Green

ISBN 9781862032538

Chatham House, 10 St James’s Square, London SW1Y 4LE

T: +44 (0)20 7957 5700 E: [email protected]
F: +44 (0)20 7957 5710 www.chathamhouse.org
www.chathamhouse.org
Charity Registration Number: 208223
9 781862 032538
Preparing for High-impact,
Low-probability Events
Lessons from Eyjafjallajökull

Bernice Lee and Felix Preston, with Gemma Green

A Chatham House Report

January 2012

www.chathamhouse.org
 Chatham House has been the home of the Royal Institute of International Affairs
for ninety years. Our mission is to be a world-leading source of independent
analysis, informed debate and influential ideas on how to build a prosperous and
secure world for all.

© The Royal Institute of International Affairs, 2012

Chatham House (The Royal Institute of International Affairs) in

London promotes the rigorous study of international questions and is
independent of government and other vested interests. It is precluded
by its Charter from having an institutional view. The opinions
expressed in this publication are the responsibility of the authors.

All rights reserved. No part of this publication may be reproduced or

transmitted in any form or by any means, electronic or mechanical
including photocopying, recording or any information storage or
retrieval system, without the prior written permission of the copyright
holder. Please direct all enquiries to the publishers.

The Royal Institute of International Affairs

Chatham House
10 St James’s Square
London SW1Y 4LE
T: +44 (0) 20 7957 5700
F: + 44 (0) 20 7957 5710
www.chathamhouse.org

Charity Registration No. 208223

ISBN 978 1 86203 253 8

A catalogue record for this title is available from the British Library.

Designed and typeset by Soapbox, www.soapbox.co.uk

Printed and bound in Great Britain by Latimer Trend and Co Ltd

The material selected for the printing of this report is Elemental

Chlorine Free and has been sourced from sustainable forests. It has
been manufactured by an ISO 14001 certified mill under EMAS.

ii

www.chathamhouse.org
Contents

About the Authors v

Acknowledgmentsvi
Executive Summary and Recommendations vii

1 Introduction 1
1.1 Spotlight on systemic resilience 1
1.2 Questions and dilemmas 3
1.3 Preparing for high-impact, low-probability events  4
1.4 Outline of the report 5

2 Mapping Economic Impacts and the Role of Business Resilience  7

2.1 How shocks cascade through the global ecomony 7
2.2 The vulnerability of manufacturing industries 9
2.3 The consequences of a prolonged disruption 11
2.4 Uneven impacts across regions and society 13
2.5 The importance of business resilience 15

3 The Contest over Science, Certainty and Legitimacy 17

3.1 Scientific uncertainty in high-impact, low-probability events 17
3.2 Understanding risk and decision-making during crises 18
3.3 Competing stakeholder interests 21

4 The Battle for the Airwaves 23

4.1 Crisis management in the age of social media 24
4.2 Lessons for stakeholders 25

5 Conclusions and Policy Implications 29
Strengthening processes and transparency 29
Stepping up communications in a crisis 29
Enhancing business resilience and responses to shocks 30

iii

www.chathamhouse.org
 Appendices
Appendix 1: Acronyms 32
Appendix 2: Methodology 33
Appendix 3: Chatham House Survey – Questions and Responses 35
Appendix 4: Measuring the Economic Costs of a High-impact Event: the Stylized GDP Model 37
Appendix 5: Agencies and Institutions Involved in an Ash-cloud Crisis (Selected Countries) and
Decision-making Linkages during the Ash-cloud Event 39
Appendix 6: National Approaches to Decision-making during the Ash-cloud Event 44

iv

www.chathamhouse.org
About the Authors

Bernice Lee OBE is Research Director, Energy, Environment

and Resource Governance at Chatham House.

Felix Preston is Research Fellow, Energy, Environment

and Resource Governance at Chatham House.

Gemma Green is Programme Manager, Energy, Environment

and Resource Governance at Chatham House.

Contributing Authors

Rob Bailey (Chatham House)

Suw Charman-Anderson (Independent consultant)
Charles Emmerson (Chatham House)
Antony Froggatt (Chatham House)
Sarah Helm (Cambridge IP)
Ilian Iliev (Cambridge IP)
William Jackson (Chatham House)
Cleo Paskal (Chatham House)
Vanessa Rossi (Chatham House)

www.chathamhouse.org
 Acknowledgments

The authors would like to thank the contributing authors

for their substantial input into the report, particularly
Vanessa Rossi and William Jackson for their work on
the economic consequences of high-impact events, and
Suw Charman-Anderson and Ilian Iliev and the team at
Cambridge IP for their extensive analysis of media and
communication during a crisis. James Norman supported
the team with data analysis and designed many of the
diagrams and tables in the report. Cleo Paskal, Rob Bailey,
Charles Emmerson, Antony Froggatt, Jens Hein, John
Mitchell, Estelle Rouhaud, Elizabeth Stevens, Grant
Rudgley, David Heymann and Sudeep Chand at Chatham
House all provided valuable input to the drafts, and we are
also grateful to Nick Mabey, Shane Tomlinson, Jonathan
Haslam and Sam Hardy for their comments. Special
thanks to the team at Skoll for providing guidance and
critical feedback. The authors would also like to thank the
companies and organizations that participated in inter-
views and completed the survey. Thanks also to Margaret
May and Nicolas Bouchet for copy-editing the report and
overseeing the production process at Chatham House.This
study was conducted with funding received from the Skoll
Global Threats Fund.
BL, FP, GG

vi

www.chathamhouse.org
 and to the just-in-time business model. The consequences of
HILP events spread rapidly across sectors and borders, often
with second- or third-order impacts that are hard or impos-
sible to predict. The 2003 SARS outbreak, for example, cost
businesses $60 billion, about 2 per cent of East Asian GDP.
Executive The devastating earthquake in March 2011 may have lost
Japan 10 per cent of its capital stock – equivalent to around
Summary and 20 per cent of the country's GDP – with wider knock-on
impacts for global companies such as Toyota and Sony,
Recommendations which were forced to halt production.
In an increasingly connected global economy and
society more people are (and will continue to be) affected
by shocks, irrespective of whether ‘high-impact events’ are
actually becoming more frequent or not. To explore our
preparedness for HILP events in this context, Chatham
The frequency of ‘high-impact, low-probability’ (HILP) House has examined the ash cloud that spread across
events in the last decade signals the emergence of a new Europe in April 2010 to draw lessons for other HILP
‘normal’. Apparent one-off high-profile crises such as 9/11, events. In particular the analysis considered the nature
Hurricane Katrina, the Macondo oil spill and the Japanese of decision-making and coordination before, during and
earthquake and tsunami were all mega-disasters requiring after the ash cloud; the impact of scientific uncertainty; the
rapid responses at a global level, marking the beginning of economic consequences and the role of communications.
a crisis trend. But lower-profile, persistent events such as
flooding, droughts and cyclones have been shown to have
equally serious impacts, raising new questions about the way A complex risk environment
in which we perceive risk and prepare for disruptive events.
These events can manifest themselves not only as Despite considerable efforts to improve scientific under-
‘black swans’ – which by nature are impossible to predict standing and reform risk management approaches,
– but also as known hazards such as floods, hurricanes governments and businesses remain insufficiently prepared
or earthquakes, which, owing to the low likelihood of to confront HILP crises and effectively manage their
occurrence or the high cost of mitigating action, remain economic, social, political and humanitarian consequences.
un- or under-prepared for. There are also crises such as Current contingency planning often assumes the return
pandemics which typically unfold over weeks, months or a of the status quo ante after a crisis. But this approach may
few years, for which the scope or timing remains unknown be inadequate in a world of complex economic and social
even with preparations. Events such as the 2011 drought risks, especially when combined with slow-motion crises
and subsequent food crisis in East Africa have also raised like climate change and water scarcity. Slow-motion crises
troubling questions about the way in which the interna- such as these build over many years, but are likely to result
tional community responds to ‘slow-motion’ disasters in a higher frequency and greater severity of shocks. Often
which build up over several or many years. there are several steps between an event ‘trigger’ and the
The globalization of production and optimization of social consequences that result.
supply chains have increased systemic efficiencies in the National risk management structures – based on
global economy but have exacerbated the speed and scope of classifying events by tiered levels of threat and imple-
contagion in the event of shocks. They pose particular threats menting specific contingency measures – may therefore
to key industries – especially high-value manufacturing – need to be reconsidered. Instead, senior leaders and vii

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

decision-makers will need to develop and roll out a week. This was confirmed by a survey of businesses about
overarching strategies which consider the full range of the 2010 ash cloud – many said that had the disruptions
preparedness and response capacities, and establish clear continued for a few days longer, it would have taken at least
frameworks for crisis decision-making. a month for their companies to recover. It is also the case
Risk matrixes which categorize risks by common conse- that planning by government and industry organizations
quences that require a generic response (such as earth- for an ash-cloud event had failed to consider a timeframe
quakes or floods) and those that require a more specific of more than about three days. One week seems to be the
response (such as pandemics) can provide a more useful maximum tolerance of the ‘just-in-time’ global economy.
framework for decision-making. This approach has its Yet for business, deviating from the just-in-time model
own limitations; it may not always capture interrelated means potentially offsetting short-term profitability. The
risks (that flooding could lead to foot and mouth disease challenge therefore for both business and governments is
for example). But building generic institutional capacity to establishing how to balance the cost of resilience and the
plan and respond to any type of event will create a broader impact of worst-case scenarios – and who should pay.
platform to ensure greater preparedness overall.
Beyond certain thresholds governments are the
responders of last resort – they are often expected to Navigating conflicting interests amid
step in and take charge of emergency responses during uncertainty
major crises. However, sectoral responses are also
critical, especially where crises involve major engineering The existence of competing and mutually exclusive claims to
challenges or have highly technical dimensions. This is certainty is often unavoidable during any crisis situation. As
the case, for example, in the rapid production of vaccines the 2010 ash cloud over Europe demonstrated, pre-existing
or in technology failures like the Macondo oil spill and rules and guidelines will come under severe pressure during
Fukushima meltdowns. a crisis particularly if worst-case scenarios have not been
explored and in the absence of flexible but credible deci-
sion-making structures. Policy-makers have some freedom
Speed and scope of economic contagion to take emergency measures in response to a short-term
crisis, but uncertainties and conflicts of interests will inevi-
Instruments of risk management have traditionally concen- tably surface during a longer-term event, complicating the
trated on ‘normal’ procedures which regard extremes as response process as political and economic pressures grow.
unlikely. Recent shocks highlight the need to plan also for Transparency, especially during and after a crisis, can
worst-case scenarios given the nature of our increasingly help ensure the decisions are made on the basis of the
globalized and interconnected world. best available evidence (recognizing uncertainties), build
The impacts of future crises are unlikely to remain public confidence and manage vested interests. Policy-
local – regardless of their origins – and will likely affect makers need to give close attention to mapping the
more than one country or region. The vulnerabilities of complex political, institutional and industrial interests
globalized supply chains and particularly the just-in-time that surround the key stakeholders in critical areas of the
business model are likely to be exposed by any disruption economy, during and after an event.
lasting more than a few days.
Evidence from a range of recent events, notably the 2010
ash cloud, the March 2011 earthquake and tsunami in Japan Battling for the airwaves
and the floods in Thailand in 2011, indicates that key sectors
and businesses can be severely affected if a disruption to Scientific and technology uncertainty is notoriously
viii production centres or transport hubs persists for more than difficult to communicate, especially when it comes to

www.chathamhouse.org
 Executive Summary and Recommendations

articulating risks to the public – whether over climate be held to account. Innovative mechanisms to mobilize
change, bird flu or terrorism threat levels. resources automatically once warning systems are triggered
The public would benefit from increased and quality should be explored.
coverage of scientific opinion by the media. Governments Scientists need to work collaboratively with civil
could help give voice to independent scientific opinion servants, the private sector and civil society to agree on
by involving scientists in public briefings and other the most appropriate set of recommended actions and
information dissemination activities. There is also a present these to decision-makers in a transparent fashion.
critical window of opportunity for authorities to engage In the case of a continuously evolving crisis this needs
effectively during a crisis situation. Reacting slowly to be a fluid and iterative process; the recommended
can cede control of the message to other stakeholders actions should be presented together with analysis to help
who have quite different interests. On the other hand, decision-makers identify which courses of action are most
acting rapidly but without a clear strategy will affect amenable to their specific risk preferences.
credibility.
Communications strategies across all forms of tradi-
tional and social media should also be built into scenario- Creating a robust process for resilience
planning and exercises. Organizations that engage with
the public and key stakeholders in normal times, building To get the right balance between planning for specific
their presence, reputation and network, enjoy a significant ‘known’ events and creating generic responses for events
advantage when disaster strikes. This is especially true that are rare or unexpected, governments must strengthen
of social networks. But it is clear that traditional media planning processes to anticipate and manage shock
continue to be hugely influential, including in the social events: from clarity in the chains of command (espe-
sphere. Stakeholders also need a contingency plan in case cially where multi-jurisdictions are involved) to activating
systems are compromised; recent crises have shown that and connecting independent knowledge networks with
modern communications networks can be fragile and lack policy-makers, to building common approaches in the
redundancy. management of complex risks.
There are common activities and actions that are relevant
in the majority of disruptions. For example, evacuation
Improving information and coordination processes will remain largely the same whether for hurri-
mechanisms canes, earthquakes or a terrorist attack such as 9/11.
Planning for specific threats will bear fruit only if the reality
Governments must also ensure that science and uncer- matches the scenario-planning. However, governments and
tainties are translated into a set of recommended actions. stakeholders can identify robust – but not necessarily
Identifying ‘no-regret’ options in such strategies makes ‘threat-specific’ – processes to mitigate disruption.
sense whether or not a specific threat actually material-
izes in the future. For example, existing social safety net
programmes can build contingency arrangements so that Recommendations
the delivery of cash transfers or execution of public works
after natural disasters can be rapidly scaled up. This same Stress-testing risk mechanisms
capacity can be used to cope with ‘man-made’ crises such 1. Industry bodies and safety regulators should work in
as food, fuel and financial shocks. coordination with governments and businesses to stress-
Early warnings, which are by their nature uncertain, test risk-related practices in critical infrastructure sectors
must be quickly followed by recommended steps, making and to examine whether policies reflect the real costs
it easier for decision-makers to take timely action and and risks associated with future infrastructure decisions ix

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

in worst-case scenarios. This should be supported including a robust website (for example, a ‘dark site’
by interactions – before, during and after an event – prepared in advance but only made available to the
between scientific advisers and national civil contin- public when a crisis hits). National science institu-
gency agencies to ensure that decision-making during tions should work together to develop, strengthen and
a crisis is based as far as possible on scientific and promote effective guidelines for the communication
technical evidence. of scientific and risk-related information for media
2. Red-teaming HILP scenarios with key decision-makers and science institutions during a crisis, reflecting the
(politicians as well as agencies) is essential to enhance new opportunities and challenges presented by social
preparedness in coping with the unexpected. A multi- media.
sector voluntary agreement on participation in planning, 6. There should be independent, high-quality hubs
exercises and crisis response should be established, (national or regional) for up-to-date risk notification
led by governments and industry. Transport and and provision of scientific information in a crisis –
communications are two priority sectors, as they are supported by governments, businesses and industry
critical in any crisis response. These scenario-building associations – that are critical scientific institutions
exercises can also help identify particularly affected that can be expected to play a role in future crises.
social groups and countries to enable rapid financial For example, a one-stop centre should be created
and practical support where national organizations to aggregate information and advice from official
are unable to cope or where the consequences are sources with information provided by individuals via
cross-border in nature. social media networks. This would become known in
3. Sharing best practice and, where relevant, capacity, advance as the go-to place in a crisis for stakeholders,
especially among industrial sectors and governments. with enhanced capacity to meet huge increases in
There are a limited number of cross-cutting responses traffic during a crisis.
to the consequences of a crisis (rapid technological
‘fixes’, evacuation, treating sick people, communi- Enhancing business resilience and responses to
cations systems etc.), compared with hundreds of shocks
potential risks. Company-led and sectoral responses 7. Governments should work with the insurance industry
are especially critical when it comes to highly technical to set up a global pooling system for reinsurance to
issues or engineering failures. address future disruptive events and review existing
4. Emergency preparation and response mechanisms arrangements regarding the provision of state support
should be transparent and subject to public account- to businesses during HILP events. Although state aid
ability. Governments should introduce a requirement can fulfil a vital role in alleviating paralysis during
for competent authorities to conduct post-crisis and immediately following an event, concerns remain
impact assessments. These would consider how crisis around issues of anti-competitiveness legislation and
decisions were taken, the basis of risk decision- market distortion.
making processes and the consequences (positive and 8. A multi-disciplinary reference library for quantifying
negative) for the environment, society and economy. the impact of shocks should be established in relevant
This would both help ensure continuous improvement international institutions such as the World Bank or the
in future crises, and enhance the transparency of risk- International Monetary Fund. Analysts can system-
based decisions to the public. atically build up a library of observations that can be
drawn on when preparing for similar shocks in the
Stepping up communications in crises future. Mistakes made in impact studies can also be
5. All actors, especially regulators and government bodies, used to improve predictions, creating a more reliable
x should step up planning for communications in a crisis reference system to provide faster and more accurate

www.chathamhouse.org
 Executive Summary and Recommendations

analysis when faced with recurring events, and to 10. Businesses should undertake cost-benefit analysis of
improve policy planning. options such as shifting to regional hubs and storage
9. The private sector needs to invest additional resources in centres for non-perishable goods to avoid urgent inter-
training and investment in ‘business resilience’, supported continental transportation. While transport risks will
by governments, especially for small businesses. A new be more difficult to overcome for perishable goods
international standard for preparedness and continuity trade, in some instances different packaging and
management systems (ISO 22301) for organizations is storage methods may permit delivery by land and
due to be published in spring 2012. In parallel, govern- sea instead of air. Indicators of business resilience
ments could also promote the ISO mark across industry should be developed that can actually be audited or
and the public. This would help ensure a competitive reported on and passed on to stakeholders or the
advantage for those demonstrating a commitment to stock market, to bolster incentives for investing in
robust business continuity management. resilience.

xi

www.chathamhouse.org
 event that has not been conceived in advance as a realistic
threat or whose frequency is unknown. It remains unclear
in many cases how to balance the costs of preparing
against the costs of responding to a crisis, and where the
economic burden should fall.
1. Introduction
1.1 Spotlight on systemic resilience

This confluence of incidents in recent years has brought

renewed concerns over our systemic resilience to external
The beginning of this century has been marked by extreme shocks. Governments and businesses remain insufficiently
events – from floods, famines, droughts and earthquakes prepared to manage high-impact, low-probability (HILP)
to commodity price spikes and the global financial crisis. crises and shoulder their economic, social and humani-
In 2010 in Europe alone, three separate natural events – tarian consequences.
a cold spell at each end of the year and the eruption of A number of underlying factors have heightened
the Eyjafjallajökull volcano and subsequent ash cloud in awareness of the risks of such events. Decades of indus-
April – wreaked havoc as the continent struggled to regain trialization and urbanization – a phenomenon no longer
economic momentum following the global downturn. limited to a few countries – have resulted in a concen-
Early 2011 saw a drought in Australia and an earthquake tration of both population and wealth in relatively compact
in New Zealand, the most severe earthquake in Japan geographical areas, especially cities and mega-cities.
for the last hundred years – followed by a tsunami and Additionally, such events can disrupt the intercon-
meltdowns at three nuclear reactors – and the worst nectedness upon which modern society depends. The
flooding that Pakistan and Thailand have experienced for globalization of production processes and optimization
decades. All of these crisis events have caused disruption of supply chains have stimulated greater efficiencies in the
in different countries, regions and sectors, but typically it global economy. At the same time, they have increased
is the poorest communities among those affected that are the level of exposure to risk should a disruption to the
the worst hit. system occur, particularly in the transport and power
We have always had risks to face. Two things seem sectors, and they have raised the likelihood of second-
to have changed today: the frequency of catastrophes or third-order impacts that are hard or impossible to
seems to be increasing; and our population remains predict.
relatively unaccustomed to the magnitude and proba- With growing interdependence, the impacts from any
bility of the risks we are currently facing. In the past, national or local crisis – whether from an earthquake, a
events such as floods and earthquakes had significant but hurricane, a pandemic or a terrorist attack – can spread
largely localized impacts. Today, these events act as harsh beyond national borders. For businesses, this makes it
reminders of the vulnerability of our interdependent hard to establish effective resilience measures: the security
social and economic production systems and the fragility of having more than one supplier for a part may be under-
of just-in-time business models. The impacts of such mined by disruption to a single major transport hub,
occurrences tend to escalate and spread, surging and while the manufacture of a complex product that requires
stabilizing as new sectors or countries are caught in the thousands of parts can be halted by the absence of a single
chain reaction. component. High-impact, low-probability events can be
Unforeseen shocks, such as the 9/11 terrorist attacks broadly divided into three types according to the general
raise questions of how we build capacity to respond to an level of preparedness: 1

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

1. ‘Black Swans’ – events which are beyond the realm of Traditional instruments of risk management concen-
normal expectations in history, science, finance and trate on normal procedures and tend to disregard
technology and therefore impossible or extremely extremes, yet consideration of these extremes is essential
difficult to predict. Preparedness for black swans
1
given the nature of our interconnected world. Scenario-
may be enhanced by strengthening generic processes, and horizon-planning therefore run the risk of preserving
systems and institutions to enable a more effective the prevailing assumptions and mindsets in terms of
response to the consequences, but specific planning risk management. Yet, as the German sociologist Ulrich
for such an event is close to impossible. Beck has pointed out, ‘the exceptions that only apparently
2. ‘Known and prepared for’ – rare events which confirm the rule must be the primary focus of attention’.2
pose a significant threat (real or perceived), often What is required is a method for rational analysis of
as a result of historical experience or technological the worst-case scenario, and then the ability to manage
advance; this category includes flu pandemics, the situation with the resources that will be available,
extreme weather conditions, flooding, nuclear rather than those that the scenario dictates should be
accidents and terrorist attacks. To a greater or lesser available.
extent, governments and businesses invest in a wide Some crises, such as floods, hurricanes, earthquakes
range of preventative actions, warning systems and and terrorist attacks, unfold over minutes or hours.
security measures to limit the impacts. These Others, including food crises and pandemics, are ‘slow-
reflect important choices about the magnitude, burn’ events which typically play out over weeks,
scale and duration of an event that preparations months or a few years. Slow-motion events emerge even
are designed to cope with. Evaluating the costs more slowly, with the impacts building up over several
and benefits of different levels of preparedness or many years – these include climate change, water
is difficult, however, given that an event may not scarcity, biodiversity loss, non-communicable diseases
occur for years or decades. and AIDS.
3. ‘Known but unprepared for’ – rare events which The sheer number of potential types of crises and
governments and businesses may have identified as a their impacts presents a challenge in itself. However, in
potential threat, but where little or no action is taken most cases a limited number of cross-cutting or ‘shared’
to prevent or mitigate the impacts. An event may fall responses to different crises can be identified – for
under this category if preparations have only been example, evacuation processes, treating sick and injured
made for a less severe occurrence – in other words, people, providing information, maintaining public order
if worst-case scenarios have been discounted. The and effective coordination across agencies. Focusing on
tsunami and consequent flooding of the Fukushima these (including cross-border sharing of capacity) could
nuclear plant, and the 2010 Icelandic ash cloud are be the key to effective practical responses.
examples. Lack of preparedness may be a result of,
for example, extremely low perceived likelihood;
low political sensitivity; a disconnect between scien- 1.2 Questions and dilemmas
tific communities and decision-makers; socially
‘acceptable’ consequences (at least relative to the HILP events raise many challenges for governments, busi-
cost of preparedness); or a belief that, because nesses and decision-makers. To improve preparedness
the expected impacts are so extreme, preparedness they will need to address many new questions on disaster
measures are futile. and crisis management. These include:

2 1 Nassim N. Taleb, The Black Swan (London: Allen Lane, 2007).

2 Ulrich Beck, World Risk (Cambridge: Polity Press, 1999).

www.chathamhouse.org
 Introduction

• What and who defines a crisis? Where do responsi- • How can we deal with persistent uncertainties despite
bilities lie in preparing for and responding to HILP important advances in scientific understanding? For
events – considering both individual shocks and example, although climate change is widely expected
the slow-motion trends that tend to increase their to bring serious negative impacts, the specific nature,
magnitude and frequency – and how should they frequency and location of climate-related extreme
be shared among key actors such as governments, climate events (such as flooding, high temperatures
regional and international organizations, the private and rainfall volatility) cannot be predicted accurately.
sector, scientists and civil society? The response In addition, the estimated probabilities are often
to Hurricane Katrina highlighted, for example, misinterpreted, sometimes deliberately. This is a
fundamental tensions between federal and local challenge for detailed preparedness-planning and also
government. Climate change is the classic example makes it harder to dispel scepticism.
of a slow-motion crisis with multiple timescales and • In a media-saturated world, how should communi-
stakeholders with divergent interests. cation activities be managed during a crisis when
• How should competing needs be prioritized – consistency of information is absolutely paramount?
minimizing the loss of human life, minimizing The public can be swayed by the most vocal, the most
economic losses, maintaining public confidence in active or the most politically powerful participant
regulatory institutions, or a combination of these? The rather than the best informed or the most legitimate.
2010 Icelandic ash cloud over Europe showed how a Ash-cloud events in 2010 and 2011 resulted in a
precautionary approach can lead to severe economic public debate to a great extent dominated by airlines.
and political pressure to change the whole basis of A Congressional investigation into Hurricane Katrina
risk-management procedures in real time. concluded that the US had ‘an analog government in
• How can short-termism be avoided? Early and sustained a digital age’.4
action is required to avoid a problem escalating, but the
most serious impacts occur years or decades into the An evidence-based approach or practice in crisis or
future (climate change, ageing population etc.). There risk management may not be possible, especially during
is a temptation to delay action rather than investing an emergency when decision-making is considerably
the required political capital. Similarly, most businesses more time-sensitive. The application of the precautionary
focus on performance over much shorter timescales. principle is also far from straightforward where both action
How do we create incentives for long-term thinking and inaction can incur astronomical associated costs
and for more rapid responses to early warnings? (see Chapter 3). The existence of multiple layers of
• How can sensible decision-making be achieved when uncertainty – or competing and mutually exclusive claims to
juggling stakeholders with different interests, different certainty – accentuates the dilemmas. For example, during
assessments of the hazard, fundamentally different the 2010 ash cloud there was uncertainty over the location
tolerances to risk, and hence different approaches to of the ash, greater uncertainty over ash concentration at
risk management? The response to the 2009 H1N1 different locations, and further questions about the impact
pandemic is a good example: different perspectives of ash on aeroplane engines. Each of these led to arguments
on the risk led some to praise the rapid response and between stakeholders and confusion for the public.
others to argue that commercial lobbying had driven One of the cardinal principles of risk management is
decision-making. 3
that risk is borne most effectively at the level at which

3 Martin Enserink on the Science magazine website: ‘Facing Inquiry, WHO Strikes Back at "Fake Pandemic" Swine Flu Criticism’, 14 January 2010,
http://news.sciencemag.org/scienceinsider/2010/01/facing-inquiry.html.
4 Tom Davis et al., ‘A Failure of Initiative: Final Report of the Select Bipartisan Committee to Investigate the Preparation for and Response to Hurricane 3
Katrina’, Congressional Reports: H. Rpt. 109-377, US Government Printing Office, 2006.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

it arises. Otherwise individual actors are absolved of 1.3 Preparing for high-impact,
responsibility, increasing moral hazard and socializing the low-probability events
downsides of some risks.
However, this approach may be difficult or impossible How countries prepare for HILP events is shaped by
to apply during a complex crisis where there is an ongoing their historical experience of natural disasters, threats to
process of defining risk parameters and serious infor- public health and national security concerns – and how
mation asymmetry. This rule also poses a challenge for the recently these have materialized. Because the systems
most vulnerable countries and communities, which may and processes tend to evolve in response to specific
often be closest to the level at which a risk arises, but may events they may not be well suited to an unexpected
lack the capacity and resources needed to manage it effec- scenario. Moreover, they reflect national institutional
tively. Additionally, where the challenge is quintessentially arrangements, posing challenges for coordination
systemic it cannot be owned by any constituent part of the across borders. Planning for a disruption to aviation
system, rendering the traditional ‘hands-off ’ approach to in the United States is security-focused – driven by
risk management insufficient. critical infrastructure protection, heavily strengthened
The international or transnational aspect of such a in response to 9/11 – while crisis arrangements for
crisis – involving decision-making, coordination and inter-modal transport were introduced in the wake
implementation by different jurisdictions – also of Hurricane Katrina. UK government planning for a
compounds the operational challenges. Agencies from major transport disruption is shaped by concerns related
multiple jurisdictions had to come to agreement on to counter-terrorism, severe winter weather and the
adjusting the rules on air travel in an ash-cloud situation fuel price protests of 2000. Other developed countries
and manage large-scale interruption to international air (including Japan, Australia and New Zealand) focus on
traffic. Aligning incentives between stakeholders will coping with major natural disasters.
be more complicated across different cultures of risk Most countries have a structure for coping with civil
management. The question of responsibilities may become emergencies. The designated agencies – often known as
harder to pin down, particularly in a fluid or unantici- civil contingencies departments – frequently sit within
pated situation, or one in which the geography of the crisis government bodies and are responsible for preparedness
changes markedly over its duration. Stakeholders have and for coordination during a crisis. For example, the UK
also expressed concern that leadership at the European Civil Contingencies Secretariat sits within the Cabinet
level was weak, leading to poor coordination and a loss of Office. In France the relevant authority is the Direction
confidence in decision-making. de la Sécurité Civile, while Germany has its Federal Office
During a crisis, which sector takes the lead during for Civil Protection and Disaster Assistance (Bundesamt
each phase can vary by country, and the interactions für Bevölkerungsschutz und Katastrophenhilfe). In the
among governments, the private sector and civil society EU, most responsibilities for civil contingencies lie with
are often affected, for better or worse, by the media the individual member states; however, the Community
(see Chapter 4). However, ultimately all stakeholders Mechanism for Civil Protection can provide some
look to the state to provide effective guidance and to financial support and coordination activities during a
impose order, even though governments may be beset by crisis.
competing government agencies, different interest groups Such agencies typically plan for a crisis by identifying
and incomplete or inadequate information. potential threats; taking action (together with a range of
Many of the adverse effects can be mitigated by actions stakeholders) to reduce the likelihood or magnitude of
taken before, during and after the event. However, this an event; clarifying roles and responsibilities, including
assumes that the risk has already been identified and prior- running scenario exercises in advance to test these
4 itized by key stakeholders. out; and, during an event, coordinating the response

www.chathamhouse.org
 Introduction

of multiple agencies and ensuring consistent and clear through scenario-planning is important, but there is no
communications. Despite having broadly similar aims, guarantee that the reality will match estimations. Building
these agencies naturally differ in their priorities (the generic processes to respond to any type of event with
types of threats they have identified for their country shared consequences creates greater institutional capacity,
and citizens) and institutional norms, which may but will provide less institutional capacity on the specific
include: nature and implications of a given event.
Politicians and government officials, under pressure to
• Direct intervention by governments versus a more take action from the media and the public, may change
hands-off approach where the focus is more on course during a crisis, potentially undermining plans
enabling the private sector or individuals to respond made by bureaucrats. Resources may be directed to a
by providing information and advice; politically sensitive issue that would not be a priority
• The balance of responsibility between state agencies, from a more dispassionate stance. To take one example,
the private sector and voluntary organizations; during the 2010 ash-cloud event the UK’s decision to
• The level of citizen participation; send three large naval vessels to rescue stranded citizens
• Emphasis on a centralized national response versus appeared to have been made in ministerial circles. A major
decentralized regional action; challenge for civil contingencies is how to balance the
• Emphasis on civil organizations versus a more practical implementation of a response to a crisis, which
security- and military-focused approach. is institutionally driven, against the impact of political
decision-making.
While civil protection agencies have a cross-cutting
approach, preparations for specific risks are also made
by more targeted agencies, responsible for arenas such as 1.4 Outline of the report
counter-terrorism, nuclear safety, the environment, public
health and aviation safety. Each issue is normally the Chapter 2 sets out the economic costs of HILP events
responsibility of one or more government departments, and how the impacts of a shock spread across sectors and
but the day-to-day activities may be devolved to an arm’s- countries in today’s globalized world. The chapter high-
length organization (such as an environment or safety lights the importance of the duration of an event. ‘Just-
agency, health service or civil aviation authority). in-time’ business models and the complexity of product
The linkages between central coordination and more supply chains mean that costs can escalate rapidly once
specific agencies are part of the architecture for crisis transport networks (or major production centres) are
response. This brings significant operational challenges disrupted for more than a few days.
in practice. For example, the different organizations have Chapter 3 explores two critical dimensions of
their own decision-making dynamics. Some agencies the decision-making environment during a crisis –
have international dimensions that are not mirrored omnipresent questions of scientific and technological
in the structure of national civil contingencies, which uncertainty, and the competing economic and political
often have limited links with their foreign counterparts. interests of key stakeholders. Successful management of
Further tensions can emerge between detailed planning HILP events means skilfully navigating these two inter-
and generic preparation; deep technical understanding related challenges.
and more flexible tools; and the role of the state versus that Chapter 4 argues that effective messaging and commu-
of the private sector. nications have never been more important in the
Getting the balance right between planning for specific management of high-impact events. The chapter draws
events and divising generic responses is at the heart of on systemic analysis of social media to understand how
preparing for HILP events. Preparing for specific events the public discourse is shaped; highlights the window of 5

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

opportunity to influence media messaging; and draws The Appendices include a reference list of acronyms,
lessons for how the media should handle scientific an outline of the research methodology and supporting
uncertainties. empirical research relating predominantly to the unfolding
Chapter 5 summarizes the conclusions and recommen- of the ash-cloud event that was used as the central case
dations for governments and other stakeholders. study in the report.

www.chathamhouse.org
 Unpredictable and uncontrollable events can pose risks that
stretch beyond the normal range of economic variables.
In an increasingly globalized world, a disruption in one
sector can swiftly cascade to other parts of the economy
and society. Even relatively ‘small’ events can generate
2. Mapping Economic significant overall effects across regions and around the
world. These may be low- or high-probability events of
Impacts and the Role short or long duration – and all of these characteristics can
alter the nature of the results. Notably, even when initial
of Business Resilience effects are low, if change is persistent rather than short-
lived, the impact is likely to build up. The effects tend to be
uneven, rising and falling as new sectors, or countries, are
caught up in the chain reaction.

2.1 How shocks cascade through the

Key messages in this chapter global economy
z Even though globalization has created greater
efficiencies through integrated supply chains and Economic activity is driven by different kinds of expendi-
concentrated assets in major production hubs, ture and activity, reflecting both domestic and external
it has also increased the scope and speed of demand and production, and exhibiting varying degrees of
contagion should a disruption to the system occur. sensitivity to cycles and shocks. Depending on the origin
z The consequences of high-impact, low-probability of any particular shock, some parts of the economy will be
events often spread rapidly and unevenly across directly affected and others indirectly.
sectors and borders. They pose particular threats to Discretionary consumer spending and business expend-
key industries – especially high-value manufacturing iture tend to be the initial drivers and key swing factors in
– and to the just-in-time business model. determining the economic impact of shocks. Spending on
z Businesses and organizations can improve their essential goods and services by households and businesses,
resilience by adapting their business models and long-term infrastructure and public-sector goods and
improving emergency decision-making capacities. services – and the exports and imports associated with
z Governments are not well equipped to manage the these demands – tend to be relatively inert and display
effects of a prolonged disruption to critical trade and only low volatility in reaction to many shocks. In contrast,
transport networks. The vulnerabilities of the just-in- consumer spending on leisure, travel and tourism is
time business model are likely to be exposed by any quickly and strongly affected by any such shift in behaviour.
disruption lasting more than a few days. This is because, as non-essential expenditure, it can be
z Impacts from HILP events are felt unevenly across rapidly adjusted in the event of any change in circum-
the world, with poorer communities most at risk to stances. Advanced economies with a large proportion of
shocks. Crisis-led policy- and decision-making and discretionary spending will be susceptible to short-term
intervention processes can also have unintended volatility linked to ‘news’, whereas the repercussions of
consequences for those most vulnerable. shocks may appear much later in the poorer developing
countries, especially those less open to trade.5

7
5 This was the reason why North Korea was seen as the country least affected by the 2009 global recession.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

In response to a negative shock, high-volatility in the first round are largely composed of job-intensive
expenditure tends to fall rapidly as news breaks, with industries). Through trade, effects spread to other
discretionary consumer spending and services exports countries even if they have no direct connection to the
being cut. These first-round losses then create secondary initial shock. But international contagion also spreads
effects in the local economy, curbing imports and immediately via the impact on consumer and business
hitting employment (especially if the sectors affected confidence (see Appendix 5).

Box 1: The impact of SARS on Asian economies

Severe Acute Respiratory Syndrome (SARS) generally spreads by close person-to-person contact. During the 2003
outbreak, international aviation served as the key mechanism for wider dispersion, yet the economic contagion
transcended physical distance.  It is notable that the travel and tourism sector, a major part of so-called discretionary
consumer spending, now plays a major role in creating almost instantaneous global economic contagion. This rapid
shock-transmission channel can cause economic damage long before any significant direct effects (e.g. flight cancel-
lations or cases in epidemics) are seen – in this case, the sizeable impact on Asian economies from SARS even in
countries with no actual cases.
Disruption to aviation in the region lasted about two months compared with less than two weeks for Europe’s
ash-cloud event, with losses in travel and tourism about double those of the 2010 event – overall, therefore, SARS
probably cost around eight times as much in economic terms. Estimates after the event suggested that SARS caused
an average loss in regional GDP for East Asia of about 0.6–0.7% for 2003 (about double the loss in foreign tourist
revenues). Taking one-eighth of this loss as a proxy for the ash-cloud impact on Europe would therefore imply an
equivalent GDP loss for the EU of around 0.05% for the year, or 0.2% for the second quarter.
In Figure 1, historical data for growth in revenue passenger-kilometres show past losses in 2001 and 2003 versus
2009 and 2010, with the 2010 ash-cloud incident standing out.

Figure 1: Passenger-kilometre growth, by destination (%)

Intra-Europe Total
25

20

15

10
Growth (%, y-on-y)

-5

-10

-15

-20

-25
00

01

02

03

04

05

06

06

07

08

09

10

10
l0

l0

l0

l0

l0

l0

l0

l0

l0
n

l
Ju

Ju

Ju

Ju

Ju

Ju

Ju

Ju

Ju

Ju

Ju
Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

Ja

8 Source: Association of European Airlines

www.chathamhouse.org
 Mapping Economic Impacts and the Role of Business Resilience

In this sense it was not the 2008/09 global recession just-in-time business model – lean systems with minimal
that was surprising but the fact that the world economy inventories and tight schedules – is that the production of
had not seen large, synchronized disturbances of this kind final goods such as cars or mobile phones can be delayed
more often over the previous decade – the so-called great if essential components are not available almost immedi-
moderation was the unusual feature. For example, just a ately via the global transportation system. Manufacturers
slightly longer outbreak of SARS in 2003 (see Box 1), with of these intermediate products aim for a rapid turnaround
a few more cases in Europe and the United States, might on delivery to meet customer needs. A disruption can
have turned this event into a global recession on a similar therefore result in rapidly escalating costs if production is
scale to that of 2009. And a few months of the ash-cloud forced to slow or pause.
problem over the summer of 2010 could have provoked a Up to a third of the total value of international trade
slump back into recession for Europe. is transported by air, and the share is growing.6 For any
economy that is highly interconnected with the rest of
the world, aviation provides vital trade links in today’s
2.2 The vulnerability of manufacturing globalized economy. For example, in its trade with non-EU
industries countries, the UK exports 55% of manufactured goods by
air. A similar share of critical intermediate components for
As governments across the world navigate their economic manufacturing – such as machinery, mechanical appliances
recovery after the financial crisis, manufacturing has and electric and electronic equipment – is air-freighted
returned to centre stage. The sector is perceived as key from non-EU countries.7
to addressing long-term economic prosperity, growth Air transport is used predominantly for high value-to-
and competitiveness questions. Even among indus- weight products, perishable goods, emergency deliveries
trialized nations, manufacturing has been offering a for unanticipated shortages, and products requiring the
glimmer of hope amid the recession and many developed security of increased attention. But high value-to-weight
regions, including the United States, Japan and the EU, ratios mean that there is a relatively light transportation
have bolstered their industrial policies in response. cost burden and if critical components (such as micro-
Disruptions to high-value manufacturing are therefore chips) which form part of a complex supply or distribution
a significant threat to the economic health of some chain are not delivered in a timely manner, significant
countries. assets may lie idle.8
Any disruption to the international trading system – Trade in perishable goods such as flowers, fish and
and particularly to air transport – has the potential for a fresh fruit depends on air transport, with many of these
wide-ranging impact. Moreover, given the concentration products being distributed across Europe from key hubs
of economic production and value in a limited number in the Netherlands and Luxembourg. Clearly, any delay to
of geographic locations (south of Tokyo in Japan, the highly perishable goods leads to a significant decrease in
Ruhr Valley in Germany, the Pearl River Delta in China product value.
and so on) the impact of a local shock can quickly spread A systematic analysis of goods imported and exported
throughout the global manufacturing sector. by air can shed light on the dependence of specific high-
Today’s manufacturing sector is highly dependent on value supply chains on aviation. Figure 2 displays the
air-freighted intermediate products. One consequence of a range of goods exported from the EU in 2009 by total

6 John D. Kasarda, Stephen J. Appold and Makoto Mori, ‘The Impact of the Air Cargo Industry on the Global Economy', presented at the International Air
Cargo Association Air Cargo Forum, Calgary, Canada, 13 September 2006. Available at www.tiaca.org/images/TIACA/PDF/The%20Impact%20of%20
the%20Air%20Cargo%20Industry%20on%20the%20Global%20Economy.pdf.
7 Oxford Economic Forecasting, ‘The Economic Contribution of the Aviation Industry in the UK’, 2006, available: www.oef.co.uk/Free/pdfs/
Aviation2006Final.pdf. 9
8 Kasarda et al., ‘The Impact of the Air Cargo Industry on the Global Economy’.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

value and the percentage that departed by air.9 Goods to air-freighted to their destinations. The total value of
the top right of the chart have a high value and high share imports of these categories to the EU was €7 billion, and
of air freight. Key intermediate goods such as integrated about 80% of them were air-freighted. Data-processing
circuits, sensors, data storage and fine chemicals can machines and data-storage devices also rank within the
be identified – these are components in a wide range of top 20 imports by value, with over half of these goods
manufactured goods. Other notable categories include being air-freighted. Each of these electronic components
industrial diamonds, aerospace components and medical is critical for an ever-increasing range of applications,
products, including vaccines. Analysis of imports to the such as personal computers, mobile phones, automobile
EU highlights a similar range of products. electronics and other telecommunications equipment. The
In 2009 about 90% of the semiconductors, microprocessors total value of the final products for which intermediate
and digital signal processors exported from Europe were products are required far exceeds their direct value.

Box 2: The Tōhoku earthquake and the supply-chain knock-on effect

Beyond the initial threat posed by the earthquake and consequent tsunami, emergent secondary and tertiary impacts
have cascaded unpredictably, testing emergency response and support infrastructure to or beyond their capacity, with
slow-burn disruption threatening recovery in the longer term.
For example, the early rolling blackouts which beset waste treatment plants, supermarket refrigerators and reactor core
cooling pumps persisted into the subsequent months. With reactors remaining offline and utilities unable to meet summer
demand, the Japanese government imposed restrictions across users in order to bridge the residual electricity supply gap.
The 15% reduction from the peak demand mainly affected the western Kansai region and northeast mainland area, further
blighting manufacturing in the hard-hit Tōhoku region, which accounts for around 8% of the country's GDP.
A stable electricity supply is essential for the high-precision manufacturing techniques employed in the electronics
industry – the premier industrial sector in Japan, accounting for around a quarter of exports by value. Despite the lifting
of the 15% restriction in September 2011, uncertainty surrounding the longer-term availability of Japan's nuclear plants
remains, particularly following pressure from local government to close plants if they fail stringent safety assessments.
In the event of such a failure across the fleet, it is conceivable that all 54 nuclear reactors in Japan could be shut by
March 2012. The Japan Centre for Economic Research estimates that, without nuclear power, GDP in 2012 would be
1.6% lower than it would otherwise be.a
There is clear evidence of disruption to global supply chains, according to an analysis by HSBC. The bank notes that
“Japan is not as crucial to the global economy as it was 20 years ago, but it remains a big producer of cars and electronic
components, especially D-Ram and flash memory.”b Moreover, the thirst for outsourcing has made many supply chains
geographically diverse and complex. As firms rationalized their businesses and contained costs, their production processes
became more vulnerable to problems at individual suppliers or ports. According to the World Bank, global industrial produc-
tion declined 1.1% in April 2011 in the wake of the tsunami and earthquake, probably reflecting supply-chain disruptions.c

a Japan Centre for Economic Research, ‘Impact to last decade or more if existing nuclear plants shut down – GDP could drop 2% on power shortages’,
25 April 2011: www.jcer.or.jp/eng/research/pdf/pe(iwata20110425)e.pdf.
b HSBC, ‘Double Trouble’, 2011, www.businessthinking.hsbc.co.uk/double-trouble.
c World Bank, ‘Japan tsunami and earthquake – Prospects Weekly: Global industrial production declined 1.1% in April in the wake of the tsunami and
earthquake in Japan’, 26 June 2011, available at: http://blogs.worldbank.org/prospects/category/tags/japan-tsunami-and-earthquake.

10
9 For clarity, only goods with a total import value of €2bn are shown.

www.chathamhouse.org
 Mapping Economic Impacts and the Role of Business Resilience

Figure 2: Exports of selected products to the EU

Non-industrial diamonds unworked

% Diamonds, worked, but not mounted or (excl. Industrial diamonds)
set (excl. Industrial diamonds) Parts of turbo-jets or
100 turbo-propellers
Vaccines for human medicines
Electronic integrated circuits as
Articles of jewellery and processors and controllers
parts thereof
90
Heterocyclic compounds with nitrogen
Electronic integrated circuits hetero-atom(s) only
(excl. items such as processors, Turbojets of a thrust
Antisera and other blood
controllers, memories and amplifiers)
% of goods air freighted by value

fractions and modified > 25 KN

immunological products
80
Diagnostic or laboratory Nuclear acids and their salts
Parts of aeroplanes
reagents on a backing or helicopters Mobile telephones

Parts of telephone sets Medical

70
Machines for the reception, Electrical components
conversion and transmission of
voice, images or other data
Electronic equipment
Aerospace
60 Precious metals/diamonds
Chemical
Instruments and appliances used in
Parts and accessories of automatic medical, surgical or veterinary sciences
data-processing machines

50
Parts of gas turbines
Medicaments containing hormones or
steroids used as hormones

40
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000

Total imports into the EU in 2009 (€ millions)

Source: Authors' analysis of Eurostat (external trade database), 2010

2.3 The consequences of a prolonged investment and employment levels permanently reduce
disruption the growth potential of economies.
However, in response to a continuous, critical disruption
The exposure and vulnerability of our global economy, of transport networks, a coalition of governments and
environment and society are sometimes not fully recog- enterprises might emerge to lead investment efforts aimed
nized until a disruption becomes persistent. Recovery at providing alternative systems to circumvent the risks
becomes more complex as networks and systems stall or posed by weather, ash clouds or similar problems. Such
fail. efforts are essential for loss mitigation in the face of threats
Overall business confidence and investment prospects but investment would nevertheless be costly, especially
tend to deteriorate in the face of a shock where production for countries already facing economic losses. Only those
and distribution grind to a halt. But the scale of any economies with the resources to pay for new transport
change in plans depends very much on the expected infrastructure could hope to avoid the threat of deterio-
duration of a problem: a short-lived, one-off shock (such rating long-run growth prospects.
as an unusual snowfall) might be passed off with little Through all these channels, and depending on the
reaction, whereas events that could persist or recur (such duration of risk factors, there would be further reper-
as repeatedly harsher winters) encourage more radical cussions on financial variables: a deteriorating long-run
change. economic outlook tends to weaken financial markets
In the face of persistent disruption, some businesses and capital inflows, for example, while new investment
would start to impose big cuts in investment and jobs projects that could offer solutions to transport problems
or to consider closing down. This would be a dangerous and restore growth may reignite hopes for recovery and
outcome because, over a prolonged period, lower encourage new inward investment. 11

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

Evidence from a range of recent events, notably the 2010 confirmed by a survey of businesses about the 2010 ash
ash cloud, the March 2011 earthquake and tsunami in Japan cloud – many said that had the disruptions continued for a
and the floods in Thailand in 2011, indicates that businesses few days longer, it would have taken at least a month for their
can be severely affected if a disruption to production centres companies to recover. One week seems to be the maximum
or transport hubs persists for more than a week. This was tolerance of the ‘just-in-time’ global economy.

Box 3: Potential climate-related disruptions to trade and critical infrastructure

Climate-related infrastructure threats are likely to accelerate, with looming uncertainty over environmental liability and
costs. Environmental change – whether extreme weather events, water shortages, changing sea levels or melting
glaciers – will generate new threats to critical infrastructure that underpins traditional energy production and delivery
systems.a

Figure 3: Global energy shipping routes and world ports

Ports
High density: 100

Low density: 0

Sources: NCEAS (shipping routes), FAO (ports)

Figure 3 illustrates the location of the world's ports and the density of a handful of shipping lanes upon which global
energy trade depends. On this basis, regions with the most vulnerable energy infrastructure include the east coast
of North America, Europe, Northern Asia (mostly former Soviet Union), Southeast Asia, Japan, and the Middle East,
many of which are key producers of fossil fuel for the global market. These vulnerabilities highlight the imperative
of investments that are resilient to climate change and disruptions to energy supplies and developments that will
prepare the world for the ‘once in a century’ energy transformation. Extreme weather events are set to increase –
even in the best-case ‘2 degree’ climate stabilization scenario – and the international community has yet to compre-
hend the potential disruptions these and other climate-related environmental changes will bring to global trade.b

12 a Cleo Paskal, The Vulnerability of Energy Infrastructure to Environmental Change, Chatham House Briefing Paper, April 2009.
b Cleo Paskal, Global Warring: How Environmental, Economic, and Political Crises Will Redraw the World Map (Basingstoke: Palgrave Macmillan, 2010).

www.chathamhouse.org
 Mapping Economic Impacts and the Role of Business Resilience

2.4 Uneven impacts across regions and to Africa’s comparatively low levels of economic devel-
society opment, rendering it more vulnerable to shocks as the
number of natural hazards increases.
Recent disruptions have demonstrated that any shock to Poorer communities have low incomes and few assets,
global trade routes is likely to create distinctive impacts making it harder for them to smooth consumption
for different countries and, in turn, different social patterns during an economic shock. They are more likely
groups. In part this is because impacts are unevenly to live in hazard-prone areas with weaker infrastructure
spread, but structures, development and openness also and yet are unlikely to have access to a wide variety of
vary substantially across different areas. For instance, resources and provisions such as financial services, health
while Africa is not the most disaster-prone region in care and social protection. They also spend a greater
the world, it is the only continent over the last ten years proportion of their income on food – up to 80% in the
which has experienced an increase in the number of poorest countries – so food price spikes hit them hardest.
reported disasters. This is particularly troubling owing
10
Overall, disaster mortality rates in the poorest countries

Box 4: The Kenyan flower industry during the 2010 Eyjafjallajökull ash-cloud event

As the United Nations Development Programme (UNDP) has noted, ‘it is the poor who often suffer the greatest
cost of an event in terms of lives and livelihoods, and rebuilding their shattered communities and infrastructure’.a
During the ash-cloud event in May 2010, Kenya was one of the worst-affected countries globally in terms of
disruption to exports to the EU. Roses constitute the single largest air-freighted commodity, with approximately
one-third of the cut flowers sold in the European Union grown in Kenya. As a result of an ash cloud thousands
of kilometres away, Kenyan horticulture, the top foreign exchange earner and therefore a critical component of
the national economy, lost $3 million a day and shed 5,000 jobs.b Overall impacts tend to obscure the variation
of impacts on different social groups. Within Kenya’s cut-flower industry, 75% of employees are female, and over
65% of the total workforce are employed on a temporary, seasonal, or casual basis.c Across a number of African
horticulture exporters, social norms dictate that women may be more predisposed to certain types of work (such
as horticulture) and that the income they receive is a ‘supplementary’ contribution to the household, rather than the
mainstay.d This perception leads to many workers being employed on a temporary basis, which offers little security
in the face of a disruption like the ash-cloud event.
Under normal circumstances in Kenya, fresh produce is harvested, packed and chilled on the same day, trucked
to Nairobi’s international airport and loaded onto passenger flights to Europe. Typically Kenyan produce then arrives
in supermarkets within 72 hours of being pulled from the fields.e However, during the six days of airport closures,
thousands of tons of fresh flowers were widely reported as being left rotting in storage units and warehouses around
the growing areas of the country.

a Mark Malloch Brown, Foreword to UNEP, Reducing Disaster Risk: A Challenge for Development, 2004, http://www.undp.org/cpr/whats_new/rdr_english.pdf.
b ‘With Flights Grounded, Kenya’s Produce Wilts’, New York Times, 19 April 2010, http://www.nytimes.com/2010/04/20/world/africa/20kenya.html?hp.
c C. Dolan, M. Opondo and S. Smith, Gender Rights and Participation in the Kenya Cut Flower Industry, Natural Resources Institute Report No. 2768
SSR, Project No R8077 2002–4, Chatham Maritime, 2003.
d S. Smith et al., ‘Ethical Trade in African Horticulture: Gender Rights and Participation’, IDS Working Paper 223 (Brighton: Institute of Development Studies, 2004).
e Oxford Economics, ‘The Economic Impacts of Air Travel Restrictions Due to Volcanic Ash’, 2010, http://www.airbus.com/fileadmin/media_gallery/
files/brochures_publications/Volcanic-Update.pdf.

10 ‘Disaster Risk Reduction for Sustainable Development in Africa: Guidelines for Mainstreaming Disaster Risk Assessment in Development’, 2004: 13
http://www.unisdr.org/africa/af-hfa/docs/africa-guidelines-mainstreaming-dr-assessment-development.pdf.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

Box 5: Hurricane Katrina

According to Dr Susan Cutter, Director of the Hazards Research Lab, “the revelations of inadequate response to
Hurricane Katrina’s aftermath are not just about failures in emergency response at the local, state, and federal levels
or failures in the overall emergency management system. They are also about failures of the social support systems
for America’s impoverished – the largely invisible inner city poor.”a
Socially created vulnerabilities are easily overlooked in disaster management plans and national risk registers because
they are so hard to measure and quantify.b Social vulnerability is partially a product of social inequalities – those social factors
and forces that create the susceptibility of various groups to harm, and in turn affect their ability to respond, and their resil-
ience after the disaster. Aside from the direct physical impact of a disaster process, a combination of pre-impact conditions
and prejudices in emergency management interventions can increase the burden on specific segments of each community.
In the wake of Hurricane Katrina, an evacuation call assumed erroneously that all victims had access to private
transport, when in reality many people without vehicles were unable to escape and were left stranded on rooftops –
or worse – during the ensuing floods.c

a Susan Cutter, ‘The Geography of Social Vulnerability: Race, Class, and Catastrophe’, 2006, http://understandingkatrina.ssrc.org/Cutter/.
b Ibid.
c Maureen Fordham, ‘Social Vulnerability and Capacity’, Natural Hazards Observer, 2007, www.colorado.edu/hazards/o/archives/2007/nov07/NovObserver07.pdf.

are almost fifty times higher than in the richest ones.11 The was low, especially in the most indebted economies
earthquake and resulting tsunami that hit Japan in 2011 such as Ireland, Greece, Portugal, Spain and Romania;
were catastrophic, killing nearly 16,000 people. However, • Economic and socio-economic factors which
the earthquake that hit Haiti the year before is estimated to determine levels of vulnerability to shocks such as
have cost over 300,000 lives, despite being several orders of per capita income and other human development
magnitude smaller. indicators, or access to social protection schemes and
Relative impacts are usually driven by the following safety nets;
factors: • The examples in Boxes 4 and 5 demonstrate how
crises can have disproportionate impacts upon poor
• The geography of the shock and how this may affect people – either as a result of their greater vulnerability
key global hubs; (the labourers in the Kenyan cut-flower industry had
• The composition of GDP and consumer spending, the most precarious jobs in the value chain) – or as a
including the proportion of discretionary spending result of inappropriate policies (policy-makers in the
in the economy and (where there is a disruption to Katrina response failed to take proper account of the
passenger transport) the importance of tourism and needs of low-income groups).
travel within the local economy;
• The ownership and structure of key industries, such Poor communities are more vulnerable to shocks – but
as the companies most affected by the disruption and they are also more likely to be marginalized economi-
their employment patterns; cally, politically or socially. As a result, their needs may be
• Scope for government policy responses: in the wake of inadequately considered in planning for, responding to or
the 2009 recession, the scope for government assistance recovering from a crisis. This can leave them trapped in a

11 Analysis by the International Red Cross and Red Crescent Federation found that between 1991 and 2000, there were 23 deaths per disaster in
14 countries with high human development indices, compared with 1,052 in countries with low human development indices. See International Federation of
Red Cross and Red Crescent Societies, World Disaster Report 2001 (Geneva: IFRC, 2001), pp. 161–4.

www.chathamhouse.org
 Mapping Economic Impacts and the Role of Business Resilience

downward spiral of increasing poverty and vulnerability, During any major disruption to transport systems,
as each crisis leaves them more susceptible to the next. management structures can be dislocated and employees
As key stakeholders in crisis management and response, left stranded in different locations. Restrictions on flights
civil society organizations play an important role in limiting following 9/11 led some – but by no means all –
disproportionate impacts upon poor people, for example by companies to change their business practices so that they
campaigning for civil, political and economic rights, or would not be paralysed by a future shock.15 Information
delivering emergency responses to affected communities. and communication systems can be held at multiple
Ultimately, however, responsibility for ensuring that the locations and accessed remotely, enabling activity to
needs of vulnerable communities are properly considered continue if a local disruption occurs. Energy systems
– and that the associated vicious cycle of increasing vulner- can also be backed up. All these responses may require
ability is broken – lies with government agencies. significant changes to business culture and investment in
The humanitarian community has made progress on resilient systems.
agency-specific preparedness missions, but recent crises As is the case for governments, stress tests and scenario
highlight the importance of investing more resources in devel- analyses can be used by firms to understand the potential
oping national and global capacities. Coherence and coordi- negative impacts from rare events that are typically omitted
nation across development and disaster agencies need to be in risk models.16
addressed and there is a need to shift away from contingency Enhancing resilience at the strategic level may mean a
planning to multi-year preparedness strategies. Much progress more fundamental challenge to business models. Firms
has been made in improving early warning systems, but this might choose to hold larger inventories, use multiple
does not always translate into early action on the ground. suppliers for key manufacturing inputs, regionalize opera-
tions to spread risk across different locations, and invest
in more resilient infrastructure along supply chains.
2.5 The importance of business resilience Yet many of these measures mean deviating from the
just-in-time model that businesses strive for in order to
HILP events pose significant risks to businesses. Nearly maximize short-term profitability. Businesses are likely
18,000 were dislocated, disrupted or destroyed, for example, to weigh the impacts of such measures, which potentially
by the events of 9/11. Over a ten-year period, Deloitte
12
reduce the efficiency and profitability of their operations
found that half of the 1,000 largest global companies to some extent, against reduced vulnerability to future
suffered declines in share prices of more than 20% in a shocks. In an extreme case these shocks could result in
one-month period. They identified ‘high-impact, low- overwhelming losses. However, they may not materialize
probability’ risks as a major cause of these losses. A quarter for many years.
of the companies affected had to wait a year before their In a survey carried out by the Business Continuity
share price recovered – and another quarter had failed to Institute, 50% of business respondents working in
recover by the end of the study’s timeframe. Airlines and 13
twelve different countries and across twelve different
the travel industry suffered huge costs during the 2010 ash sectors had tried to optimize productivity and profit
cloud and, given wider economic conditions, many were margins by outsourcing, consolidating suppliers
pushed close to bankruptcy. 14
and adopting just-in-time or lean manufacturing

12 Gail Makinen, ‘The Economic Effects of 9/11: A Retrospective Assessment’, Congressional Research Service, 2002, www.fas.org/irp/crs/RL31617.pdf.
13 Deloitte, Disarming the Value Killers: A Risk Management Study, 2007, available at: www.corpgov.deloitte.com/binary/com.epicentric.contentmanagement.
servlet.ContentDeliveryServlet/CanEng/Documents/Risk%20Oversight/DisarmingTheValueKillers.pdf.
14 IATA, ‘Volcano Crisis Cost Airlines $1.7 Billion in Revenue – IATA Urges Measures to Mitigate Impact’, 2010, www.iata.org/pressroom/pr/Pages/2010-04-21-01.aspx.
15 Amy E. Hurley-Hanson, ‘Organizational responses and adaptations after 9-11’, Management Research News, 2006, Vol. 29, Issue 8, 2006, pp. 480–94,
www.emeraldinsight.com/journals.htm?articleid=1571697&show=html. 15
16 Deloitte, Disarming the Value Killers.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

techniques.17 Businesses choosing such options were about the volcanic ash event, 50% of companies for which the
shown to be more likely to experience supply chain disruption, closure of European airspace was identified as a significant
with 83% of those using low-cost-country solutions experi- problem did not have this risk on their register; and for those
encing disruption owing to failures in the transport network that did, one week was generally the longest period that had
and among suppliers in 2010. When asked specifically
18
been considered during planning.19

Box 6: Business responses to the 2010 ash cloud

In total there were 36 responses from businesses to the Chatham House survey on the impact of the ash-cloud event
in April 2010 (See Appendix 3). A large proportion came from the ‘air travel/transport’, ‘government/regulation’ and
‘food’ sectors. For the majority, the event appears to have had a negative effect on profit/revenue. However, a small
percentage responded positively in this area. Respondents (47%) also reported that it took less than a week for
organizations to return to normal after the event. However, a large proportion also estimated that had the disruptions
continued for a few more days it would have taken over a month to return to normal. In the main, organizations
appeared to have coped positively with the disruption. Areas of exception were insurance cover and profits/revenue.
It appears that, where possible, organizations utilized other forms of transport or transferred business locally in
order to function. Regular communication was used to aid public image, supplier and customer relations. With regard
to information or advice about the event, it appears that in the majority of cases respondents received this either
immediately or within 24 hours. However, 40% claimed that information from regulators, government and international
bodies was not received until at least two to three days later and in some cases over a week. The majority claimed
not to have received targeted information from the government or regulators at all. Those that did, through websites,
described it as inconsistent or too late to have a beneficial impact.
Company websites were the main tool used by organizations to relay key messages to consumers and the general
public. Other methods (not specified) were mainly used to relay messages to the government, but social networking sites
and advertising were barely used. With regard to stakeholder performance, decision-making by regulators, governments
(home and other national) and the European Commission, as well as scientific bodies, was rated poorly (by over 60% of
respondents). In contrast, NGOs, trade unions and private enterprise were generally rated positively on decision-making
and communication (around 80% of respondents deemed decision-making ‘Good’ or ‘Very Good’ in relation to these
actors). Generally, communication was rated lower than decision-making across all stakeholders surveyed.
The majority of respondents claimed that organizations had not made changes to existing strategies. However, a
small number in ‘air travel/transport’, ‘food’, ‘government/regulation’, ‘tourism’ and ‘other’ sectors claimed changes had
occurred or were being considered.
In the main, respondents commented that a unified decision-making process across Europe and further afield
should be adopted. Many said that inconsistencies in the stated risk of flying increased disruption, with some airports
opening where others remained closed. Better preparation, scientific research and communication strategies in future
were also called for.

17 Business Continuity Institute, ‘Supply Chain Resilience 2010: BCI Survey of Resilience Professionals’, October 2010.
18 Ibid.
16 19 Business Continuity Institute, ‘A Report on the Disruption Caused by the Closure of European Airspace Due to Volcanic Ash in April 2010 as Reported by
BCI Members’, June 2010.

www.chathamhouse.org
 In any crisis, stakeholders bring a range of economic,
political and public interests to the table. Scientific and
technical information is often negotiated and contested,
reflecting the core interests of key players. The complica-
tions introduced by scientific uncertainty can become
3. The Contest over more pronounced where events are completely unforeseen,
which further highlights the need for effective procedures.
Science, Certainty Applying evidence-based practice to crisis or risk
management is not always possible as a time-sensitive
and Legitimacy crisis unfolds. Application of the precautionary principle
may mean shifting the burden of proof so that those
taking action have an obligation to demonstrate that it is
not harmful. It also implies an approach – in the presence
of scientific uncertainty – that focuses on weighing up the
costs and benefits of different actions, including the option
Key messages in this chapter of doing nothing. But in some risk environments, the
z There cannot be adequate planning for all even- price both of taking action and of inaction is potentially
tualities, especially when it comes to ‘black swan’ high, making it difficult to form decisions on the basis of
events. Governments and stakeholders must identify either a precautionary approach or traditional cost-benefit
robust but not necessarily ‘threat-specific’ processes analysis. In such circumstances, identification of least- (or
to mitigate a disruption. where possible no-) regret options is likely to be critical.
z The competing interests of key stakeholders are often Navigating the interrelated challenges of uncertainty
played out in a public manner – through, for example, and the competing interests of stakeholders is crucial in
the very different ways in which scientific uncertainty is preparing for and managing the consequences of a high-
interpreted and communicated to the public. Mapping impact, low-probability event. Scientific uncertainties are
and managing these conflicting interests is a critical a common feature of both emergencies and slow-motion
component when preparing for and managing the crises and may be very difficult to model. Such events may
consequences of an HILP event. even challenge the very basis of current understanding,
z Addressing investment gaps in scientific capacity and undermining the relevance of predominant assessment
institutions following an HILP event is an important techniques. The financial crisis is a case in point – however
part of the evolutionary response for key agencies. This advanced the financial tools were at the time, they could
should be combined with strong and coordinated action not explain a world in which the whole system broke
by governments to introduce and enforce transparency down. At the same time, the treatment of uncertainty can
and effective risk-management frameworks. play a central role in the decision-making process.
z Scientific uncertainty remains a key part of the
decision-making landscape. While sound in principle,
an evidence-based approach may not be possible 3.1 Scientific uncertainty in high-impact,
during a crisis. This makes the transparency of low-probability events
risk-based decision-making vital to maintain public
confidence in crisis management. The application of A recent UK House of Commons report stressed the
the precautionary principle is also not straightforward importance of integrating science in planning processes,
where the price of action and inaction are both especially for known events, before an emergency
potentially vast. occurs: 17

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

we have been left with the impression that while science The catastrophe in Japan in March 2011, caused by a
is used effectively to aid the response to emergencies, massive earthquake leading to a tsunami and meltdowns at
the Government’s attitude to scientific advice is that it is the Fukushima nuclear plant, also revealed the influence of
something to reach for once an emergency happens, not a conflicting interests during an emergency. According to a
key factor for consideration from the start of the planning former executive of the power company Tepco who is also
process.20
a member of the Japan Atomic Energy Commission, Tepco
delayed the use of seawater in an attempt to protect its
While scientific and technical input is certainly important assets, acting only when eventually ordered to by the prime
for evidence-based decision-making in preparing for and minister.22 This illustrates how the choice of emergency
responding to a crisis, focusing on improved science as measures taken to deal with an HILP event cannot be
a solution can also be misleading. Scientists can inform reduced to a question of scientific evidence. A conflict
us about the most probable worst-case scenario for each between competing stakeholder interests and scientific
event, but not always that one disruptive event can lead uncertainty can emerge as alternative interpretations of
to another. Later in this chapter, an analysis of the key evidence and approaches to communication converge.
areas of disagreement between the major stakeholders in
the ash-cloud crisis shows that competing interests and
information asymmetry were as important as scientific 3.2 Understanding risk and decision-
uncertainties. These issues were confirmed by a number of making during crises
other HILP events.
One such event was the 2002–03 outbreak of Severe Government, the private sector and civil society are
Acute Respiratory Syndrome (SARS). The outbreak constantly making decisions based on cost-benefit risk
was not brought to the attention of the World Health assessments. A government may choose to deregulate a
Organization until late November 2002, nearly a month sector after weighing up the political gain and loss for the
after the first reported cases, when Canada’s Global Public main governing party. A business may choose to locate a
Health Intelligence Network informed it of unconfirmed facility in a hurricane zone if it can secure insurance to
reports of ‘flu outbreaks’ in China which it had picked defray potential costs if damage occurs. A civil-society
up on the internet. Even then, institutional limitations group may advocate building more low-cost housing if the
meant that the reports were only presented in English and policy appeals to more of its members and funders than it
French, limiting the spread of information on what had the alienates.
potential to be a truly global catastrophe. It was only much Each decision may be rational within its own narrow
later that the disease received significant public attention. cost-benefit parameters. However, given the intercon-
The lack of information contributed to widespread uncer- nected nature of our globalized systems, these decisions
tainty about the nature and properties of the disease, do not exist in isolation and can leak vulnerabilities into
leading to unnecessary, indirect impacts on economies other sectors. Government deregulation in one sector can
and societies far from the centre of the outbreak. For weaken the global financial system. A hurricane off the US
example, despite not having a single case of SARS, business Gulf coast can (as Katrina did) destroy over 100 oil and gas
in New York’s Chinatown went into rapid decline and had platforms, creating a spike in global oil prices and affecting
still not recovered a year after SARS had been contained. 21
business worldwide. Low-cost housing can end up being

20 House of Commons, Science and Technology Committee, ‘Scientific Advice and Evidence in Emergencies’, Third Report of Session 2010–11, Volume I:
Report, together with formal minutes, oral and written evidence, 2011.
21 Laura Eichelberger, ‘SARS and New York's Chinatown: The Politics of Risk and Blame During an Epidemic of Fear’, Social Science & Medicine, Vol. 65,
18 Issue 6, 2007, pp. 1284–95.
22 Norihiko Shirouzu, Phred Dvorak, Yuka Hayashi and Andrew Morse, ‘Bid to “protect assets” slowed reactor fight’, Wall Street Journal, 21 March 2011.

www.chathamhouse.org
 The Contest over Science, Certainty and Legitimacy

built on flood plains, requiring military intervention to Some forms of insurance can also distort the evalu-
evacuate residents and secure critical installations. ation of risk. For instance, government will for political
One reason for this discounting of risk is the common reasons back or directly insure activities in some sectors
practice of ignoring ‘externalities’ when calculating cost. that private-sector insurers have deemed too risky to
Externalities can be costs expected to be borne by others take on. The result is an undervaluation of the real risk
(for example insurers, or the government), or costs that involved. One clear example of this is the US National
will become due past the timeframe under consideration Flood Insurance Program, which allows for government-
(for example, during another administration or in a future backed insurance on properties in areas that private
corporate bonus period). insurers consider unprofitable. It can essentially act like

Box 7: Translating early warning of drought risk into timely action

All crises are different – playing out over different periods of time, in different political, cultural and economic contexts
and with different stakeholders and competing interest groups. Yet a common theme remains the problem of
translating scientific analysis into timely action.
Drought, and the food crises that may result, are a classic example of slow-motion crises in which failed rains lead
to a failed harvest, after which food reserves and asset bases are eroded as people struggle to feed themselves. Food
availability declines and prices rise, leaving the vulnerable populations facing months of severe food insecurity until
another harvest is possible.
A crucial difference, however, is that in certain regions of the world, most notably West Africa and East Africa,
drought is not a high-impact, low-probability event; it is a high-probability, high-impact event. Droughts happen every
few years. Governments and international agencies often have a very good idea about how a food crisis is likely to
evolve and are able to see it coming a long way in advance. Yet time and again the international response is inadequate.
The 2011 East African food crisis was correctly forecast in August 2010 by the Famine Early Warning Systems
Network (FEWS NET) an entire year before the critical ‘hunger gap’ period in which the crisis peaked. Once it became
clear that the drought was a certainty – when the 2010 October rains failed – FEWS NET began issuing regular
warnings about the deteriorating situation in the region. However, the response of the international community was
inadequate. With the hunger gap upon it in July, and regions of Somalia being declared famine zones, the UN’s appeals
were less than half-funded. What is more, the failure to mobilize resources early meant that agencies were unable to
build adequate logistical capacity in anticipation of the crisis peak.
This problem is not confined to the 2011 food crisis – it is a familiar pattern, noted most recently in the 2005 and
2010 Sahel food crises. The early provision of scientific warnings of drought and crisis, inevitably including uncertain-
ties, is failing to precipitate adequate early action, despite the relative lack of conflicting interests between stakeholders,
and their familiarity with these kinds of event. A simplistic explanation that donors are reluctant to mobilize resources
until uncomfortable images of starving people are broadcast around the world offers at best only a partial explanation
and little in the way of solutions.
Attention must also be turned to how information is provided to decision-makers so that it is easier for them to take
early action and be held to account. This means translating early warnings into recommended steps, with the clear
identification of least- (and no-) regret options able to deliver a minimum level of early action that can be built upon
as information improves. Other innovative mechanisms to automate certain elements of the response could also be
explored – for example a catastrophe bond or insurance mechanism that pays out when drought happens – to reduce
the burden on decision-makers.
19

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

Box 8: Risk assessment within the nuclear industry

‘If constructing in a seismic zone that hasn't seen an earthquake above a magnitude of 6.5 in 100 years, do you build to withstand 7?
Or put in extra millions upfront to protect against a magnitude of 8? Or do you simply choose not to build a nuclear power station in an
earthquake zone at all?’
Cleo Paskal, 2011a

Since nuclear power plants use hazardous material that needs to be contained in all circumstances, the industry has
developed a risk assessment tool called a Probabilistic Safety Assessment (PSA) to enable regulatory bodies and
operators to gain an insight into the risks that the plant faces, in particular as they relate to the key safety systems
such as reactor shutdown, reactor containment, emergency core cooling and emergency power supply. The information
obtained by the PSA is used to improve the components, systems, structures and management for ensuring safety,
and indicate numerically the risks of serious accidents, particularly those relating to damage to the core and/or the
significant release of radioactivity.
PSAs are also used to communicate the risk of significant or serious accidents to the public and are expressed as the
chance of an accident in a given number of years of operation of a particular reactor or fleet of reactors. For example, the
probability of an accident that results in damage to the core of the European pressurized water reactor, the most modern
design being constructed worldwide, is described as 5.2 x10-07 – which means that there is a chance of such an accident
just over five times in every ten million years of operation of this particular design of reactor – while the anticipated risk of
core damage with the failure of containment (and therefore release of radiation) is less at 3.9 x 10-08 or almost four times
in a hundred million years of operation.b For older reactors, which have operated for longer – and hence are subject to
material ageing – and which have fewer safety systems, the risks are perceived as greater, often by an order of magnitude.
In the case of the Japanese nuclear fleet as a whole, an assessment in 2004 undertaken by the Nuclear and Industrial
Safety Agency reviewed the PSA reports performed by the electric utilities for the 52 existing nuclear power plants. It
concluded that the risk of a core-damage accident, presumably in the oldest reactors such as those at Fukushima, was
less than 10-6/reactor year, and the containment failure frequency was less than 10-7/reactor year.c To put these risk esti-
mations in context, the Japanese nuclear industry stated: ‘The individual risk, 10-6 per year, is equivalent to several occur-
rences in a long history of human beings (about 4 or 5 million years after the first human appeared on the earth).’d
However, there is a significant difference between these theoretical assessments and what has been observed. To
date there have been three major civil nuclear accidents that have resulted in damage to the core, at a minimum: Three
Mile Island in the US in 1979, Chernobyl in Ukraine in 1986 and now Fukushima. Since the start of the civil nuclear
industry the cumulative operating years of the world’s reactors is less than 15,000 reactor operating years. Therefore on
average there has been accident every 5,000 reactor years, which is twenty times greater than the probability calculated.
The significant difference between the expected and observed risk raises two sets of concerns. First, is the method-
ology of PSA correct and is it considering all the relevant factors sufficiently – in the case of Fukushima the seismic and
tsunami risk? Secondly, are the inherent risks of nuclear power, in particular human engagement, adequately considered
and how might this change over time, i.e. taking account of changing maintenance, regulation and operating regimes?

a Chatham House Expert Comment, http://www.chathamhouse.org/media/comment/view/163419.

b UK-EPR. ‘Fundamental Safety Overview, sub-chapter R.2 Probabilistic Safety Assessment’, www.epr-reactor.co.uk/ssmod/liblocal/docs/V3/
Volume%202%20-%20Design%20and%20Safety/2.R%20-%20Probabilistic%20Safety%20Assessment/2.R.2%20-%20Level%202%20
Probabilistic%20Safety%20Assessment%20(PSA)%20-%20v2.pdf.
c JNESO, ‘Probabilistic Safety Assessment (PSA), Incorporated Administration Agency’, Japanese Nuclear Energy Safety Organization, December 2007.
d JNESO, ‘Severe Accident and Accident Management, Incorporated Administration Agency’, Japanese Nuclear Energy Safety Organization, 1 July 2009.
20

www.chathamhouse.org
 The Contest over Science, Certainty and Legitimacy

Box 9: Conflicting stakeholder interests during the 2010 ash-cloud event

At the beginning of the event regulators across Europe took a safety-first approach, following the advice of aviation
safety regulators. Many airlines were unhappy that the risk decision-making was taken out of their hands – the regulators
essentially closed airspace instead of allowing them to evaluate safety on a case-by-case basis. As the economic costs
grew, pressure mounted on regulators to implement a more nuanced set of measures for dealing with the ash cloud.
This brought up another vital issue on which adequate information was necessary in the process of decision-making: the
level of ash in which aircraft could operate with a reasonable level of safety.
The importance of defining a safe threshold had been raised in industry discussions for many years but little progress
had been made.a Stakeholders have put forward a variety of potential reasons for the slow progress, including the diffi-
culties of setting standards given modelling and engineering uncertainties; limited participation by the private sector in
planning and response mechanisms; and poor communication between industry and other stakeholders. Some stake-
holders also suggested that it may not have been in the economic interest of the engine and aircraft manufacturers to set
safe ash thresholds, given the higher costs of flying in ash. However, the balance of economic incentives is far from clear.
Information asymmetry between the private sector and the regulators became a serious obstacle to resolving the crisis.
To the extent that analysis of safe levels was available, it was held by the manufacturers of aircraft engines and the airlines.
During the event data on ash concentration thresholds were shared between the leading engine and aircraft manufacturers
for the first time, but details of the arrangement or the results were not made public or released to regulators.b
As economic pressures mounted, engine and aircraft manufacturers came under intense pressure to provide
further information that could lead to 'safe' ash thresholds, which would provide a clear basis for regulators to relax the
rules. Yet although the European Council of Ministers emphasized the necessity of evidence-based decision-making,
the decision to moderate the flight ban was eventually made regardless of the absence of a clear statement from the
engine and aircraft manufacturers.

a Z. J. Przedpelski and T. J. Casadevall, ‘Impact of Volcanic Ash from 15 December 1989 Redoubt Volcano Eruption on GE CF6-80C2 Turbofan
Engines’, in T.J. Casadevall (ed.), Volcanic Ash and Aviation Safety: Proceedings of the First International Symposium on Volcanic Ash and Aviation
Safety: U.S. Geological Survey Bulletin 2047, 1994, pp. 129–35; First Symposium on Volcanic Ash and Aviation Safety, held in Seattle, Washington,
United States, 1991; International Volcanic Ash Task Force (IVATF), First Meeting. Montréal, 27–30 July 2010, www2.icao.int/en/anb/met-aim/met/
ivatf/Documents/DP%205.pdf.
b Interview with engine manufacturer.

a government subsidy to put people in harm’s way and is extremely difficult, especially as the timeline seems
result in properties being rebuilt in the same location after so extended. Does one plan coastal infrastructure for a
having been repeatedly destroyed by flooding. 15cm or 50m sea-level rise by 2050? That date may seem
Even the most comprehensive risk assessments, however, far in the future, but it is well within the lifetime of new
face another complex issue, the increasing variability of the infrastructure builds. And what happens after 2050? Often
three ‘geos’ (geopolitics, geoeconomics and geophysics). these new variables are simply ignored, thereby increasing
Predicting the behaviour of the physical world in the likelihood of future low-probability (according to past
particular is a growing challenge. With environmental geophysics), high-impact events.
change, past indices are no longer reliable guideposts Additionally, risk assessments themselves are primarily
for future events. The science is good, and getting better, designed to predict the likelihood of a specific event, whereas
but already it is often marginalized in calculations. what is of primary concern is the impact of that event. It is not
Ensuring that government, the private sector and civil the hurricane in itself that is the problem; it is the impact of
society include the new range of science-based variables that hurricane on infrastructure, governance, etc. 21

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

The ultimate impact of an event depends both on sector is dominant, it can engage with civil society over
pre-existing levels of vulnerability and on the response. potential regional risks, allowing them to become involved
Effects can be mitigated at three stages of a crisis: before (by in contingency planning. In areas with strong commu-
reducing vulnerability), during (crisis response) and after nities, civil society can organize grassroots paragovern-
(post-crisis response). Large events, such as ash clouds, mental emergency units.
hurricanes and earthquakes, involve all three key sectors – Transparency and clarity over why and how decisions
government, the private sector and civil society – with the are taken in a time of crisis are a critical issue for
interaction between them often affected, for better or worse, governments and the private sector alike. During a crisis,
by the media. Each sector has a role to play in managing flexible and transparent processes are required to avoid the
impacts, but which one takes the lead during each phase emergence of ad hoc structures with limited legitimacy,
can vary by country. For example, during Japan’s Fukushima and to ensure that regulatory structures engage all key
nuclear crisis, the private sector took the lead. During stakeholders while maintaining resistance to lobbying
earthquakes in China, the government takes the lead. In the pressures. Post-crisis impact assessments could later help
Mumbai floods of 2005, it was civil society that led the way. to establish – in particular for crises with cross-border
When trying to reduce the impact of a disruptive event, implications – how risk uncertainty was handled; the
vulnerabilities can be identified and resilience improved, basis of risk decision-making processes; and the conse-
by examining how the three main sectors are projected to quences (positive and negative) for the environment,
engage with risk before, during and after the crisis, and society and economy. This would help ensure continuous
then working to address deficiencies. In places where the improvement in responses to future crises and, impor-
private sector is weak, governments can give support for tantly for public confidence, would also enhance the
business continuity planning. In areas where the private accountability of risk-based decisions.

22

www.chathamhouse.org
 but one of maintaining credibility and legitimacy, which
may in turn determine the effectiveness of actions taken
to manage any current crisis and prepare for future ones.
Political pressure can be used to hinder necessary public
measures if authorities do not manage the communications
4. The Battle for the battle. As discussed earlier, one consequence of globaliza-
tion and interdependence is that crises, risk and uncertainty
Airwaves are far more likely to have international and transnational
dimensions than would have been the case in the past.
Uncertainties of some kind are very likely to feature in
any high-impact event, presenting communication activities
throughout a crisis situation with both opportunities and
serious challenges. The media attention brought about by
any crisis provides an opportunity to inform the public
Key messages in this chapter
on key scientific issues. However, risks and probability are
z Communicating risk and uncertainty during
notoriously difficult to communicate, whether on climate
emergency situations remains a critical challenge.
change, bird flu or terrorism threat levels. This will be
z Traditional media remain crucial in shaping public
particularly pronounced for genuine black swan events –
perceptions during a crisis.
but even well-known threats can pose major challenges for
z More scientists and technical experts need to be
communication strategies.
given a voice during a crisis to provide in-depth
Scientists deal with uncertainty every day, but for
assessments and analysis that can help decision-
journalists and the general public it can be difficult to
making and risk management.
understand. Reporters do not often provide all the infor-
z In terms of the media, there is a limited window in
mation required for the reader to assess a risk. Writing
which the authorities can engage critical audiences.
on the Columbia Journalism Review website, risk commu-
Effective communication is paramount in the
nication consultant David Ropeik says of his time as a
hours and days immediately after a crisis breaks.
journalist, ‘Our reporting was inherently deficient because
Communication style can also affect level of influence.
we just didn’t know that there are important details
z Social media can provide an effective vehicle for
without which a story about a risk is simply incomplete.’24
real-time communication to a broad audience
In-depth explanations of the science and technology
during a crisis, particularly if stakeholders are
relevant to an event can help people assess the levels of
engaged in the relevant social networks in advance
uncertainty and risk involved in a situation, and what
of an emergency.
it means for them. But as the 2010 ash-cloud event
revealed, scientists, weather forecasters, engineers and
‘The Government must communicate risk effectively to the public other experts need to be given a greater voice in the tradi-
in an emergency; this is vital to prevent mistrust and anxiety.’ tional media. There is a strong case for the establishment
23
UK Parliamentary Science and Technology Committee of independent national or regional web-based services
for up-to-date risk and science information in a crisis.
Effective messaging and communication are highly These could be established by government and connected
important in the management of any national and global to critical scientific institutions that can be expected to
emergency or crisis. This is not a matter of public relations, play a role in future crises.

23 House of Commons, Science and Technology Committee, ‘Scientific Advice and Evidence in Emergencies’. 23
24 See Columbia Journalism Review website: http://www.cjr.org/the_observatory/risk_reporting_101.php?page=all.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

Communication serves multiple roles in a crisis analysis and opinion from a myriad of non-mainstream
situation and different stakeholders will have different sources.
priorities, including: The internet cuts both ways, however, providing an
opportunity for organizations and experts to commu-
• coordinating responses among key stakeholders; nicate directly with the public, in full and without
• informing the general public and those directly affected mediation by journalists. This new media landscape
about the latest situation; changes the way public perception is shaped, allowing
• explaining and building confidence in decisions; people to fact-check news sources against primary data
• avoiding unnecessary public concern; or opposing opinions found online and use traditional
• protecting brand image and customer or investor media stories as jumping-off points for the discussion of
confidence; an event or phenomenon.
• generating interest in a related policy issue (and in Media narratives are often based on conflict – including
some cases, promoting a product). false conflict. In theory, social media can provide a means
to explore the issues in more detail and with greater clarity.
If all the stakeholders involved in a situation are using social
4.1 Crisis management in the age of media, not only will they all be able to give their own point
social media25 of view to the public, they will also have an opportunity
to understand one another’s organizations more clearly,
In Crisis Communications for the Social Media Age, written building their presence, reputation and network. Especially
for the technological website GigaOm, Aliza Sherman during crisis situations, social media offer a means to
commented: gauge the public reaction for all the stakeholders, from the
government to the private sector. Such intelligence could
People want to see a human response to a crisis, not an overpro- help organizations to understand, in near-real time, what
duced, formulaic or canned reply. When there is a crisis, there type of communication is effective, what additional infor-
are people involved. Pretending there are no emotions mixed mation could be provided and which policy responses are
in the mess is a sure-fire way to lose credibility with others. 26
likely to be effective.
The shift to a more open communication style can be
This goes to the heart of a social media response to challenging for many organizations, bringing both potential
crises. This requires a different style of communication, benefits of talking directly to their constituents but also new
which is often at odds with traditional communications risks. Traditional methods of communication can be too
or public relations techniques. For example, social media slow to counter rumour and misinformation effectively.
users expect an immediate acknowledgment of a problem, Many of the organizations playing a key role in providing
even if no solution is imminent. public information were not active participants in social
The rise of the internet has changed not just the way in media before the 2010 ash-cloud event. Stakeholders cannot
which the general public seeks and receives information expect to be granted trust purely through the strength
but also its expectations of how official bodies should of their existing brand. The use of social media space is
communicate. When big stories break, the traditional more effective when the organization has a pre-established
media are no longer the only source of information; position within it, including linking with other organiza-
the public can now easily supplement their reports with tions that will be important during a crisis. An International
information directly from primary sources, as well as SOS analysis found that during the 2010 crisis, pre-estab-

25 The term ‘social media’ refers to the suite of tools and services that provide users with the ability to interact socially around a topic, activity or object.
24 Synonyms for ‘social media’ include ‘social tools’, ‘social software’ and ‘Web 2.0’.
26 Article available at http://gigaom.com/collaboration/crisis-communications-for-the-social-media-age/.

www.chathamhouse.org
 The Battle for the Airwaves

Table 3: Organizations on Twitter and their network size/activity

Twitter account Date joined Followers Tweets

Britishairways 19 Oct 2007 99,533 2,223

Schipholairport 17 Jan 2009 2,321 17

Metoffice 21 Jan 2009 25,362 3,623

Heathrowairport 14 Feb 2009 45,464 7,491

Easyjet 8 May 2009 13,025 4,477

Ryanairnews 27 July 2009 3,248 245

Eurocontrol 11 Aug 2009 16,848 5,124

NATS_UK 18 April 2010 92 47

UK_CAA 21 April 2010 319 46

Dublinairport 17 Aug 2010 11,099 308

Source: Cambridge IP

lished internal and external communication plans and This is not because they do not trust official sources – if
channels were essential; and that in future they should anything, the 2011 Edelman Trust Barometer29 shows that
be implemented as part of an organization’s overall travel trust in academics and experts has risen to 70%, up from 62%
policy and crisis management plans. It is possible to benefit
27
in 2009, and trust in NGOs stands at 47%, up from 41% – but
from using social media in a crisis even with a new account because they tend not to seek out official sources such as press
but greater benefit is derived from established networks. releases. The medium of social media is, for many people,
Time is needed to build a network and to earn the trust of opportunistic and serendipitous. Instead of actively searching
the community, but the level of activity is also important, as for information, its users wait to receive notifications and
illustrated by Table 3. messages direct from those people in their networks.
One of the reasons why social media can be so powerful
is that a service such as Twitter is not a single network, but
a network of networks. This means that important infor- 4.2 Lessons for stakeholders
mation can ripple through the system, even if the origi-
nator has a relatively limited network. Larger networks are, There is a window of opportunity to engage effectively
of course, more effective, but the critical mass required to during a crisis situation. Reacting slowly can cede control
get information out is lower in a group of interconnected of the message to other stakeholders who have quite
networks than in a hub-and-spoke network such as a different interests. However, acting hastily without a clear
mailing list. KLM have estimated that customer-service strategy can affect credibility.
agents working in the social media department are 75% In many crises, not every type of organization is given
more productive than those in call centres. 28
equal space. When a crisis occurs, stakeholders have a
Bloggers are likely to rely on information they find via the number of potential audiences with whom they need to
media or recommended to them by their social networks. communicate:

27 International SOS, ‘Every Ash Cloud Has a Silver Lining’, 2010, http://internationalsos.fr/en/files/AshCloudBriefing.pdf.
28 Chris Heath, ‘Twitter vs Icelandic ash cloud: How one airline transformed its use of social media’, 2011, www.silicon.com/management/sales-and-
marketing/2011/06/23/twitter-vs-icelandic-ash-cloud-how-one-airline-transformed-its-use-of-social-media-39747616/#comments.
29 PR firm Edelman produces an annual survey which gauges attitudes about the state of trust in business, government, NGOs and media across 23 25
countries. The trust barometer is accessible from the company website: http://www.edelman.com/trust/2011/.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

• the general public; to be effective during a crisis. A shared vocabulary is

• other stakeholders; needed for communications to be effective, and all parties
• the media; need to understand fully the new tools and techniques, and
• volunteers and workers on the ground; how to implement them. These conversations not only give
• volunteers and workers in remote locations. the public a chance to find out what is happening, but also
provide them with an opportunity to weigh what they are
The traditional media remain critical in shaping the reading in the media against what they have learnt from
public perception in any major crisis. In the case of the their own interactions with an organization. This can be
2010 ash cloud, the Chatham House analysis showed that very useful when a stakeholder organization comes under
the travel industry as a whole dominated the traditional attack in the media from other agencies.30
media conversation, swamping the voice of scientists and In the case of the earthquake in Haiti in 2010, it was
air traffic control. The preference for quotes from the found that it was individuals within larger organizations
travel industry was evident across all publication types, who were knowledgeable about and implemented the use
but was especially apparent in the tabloids, in which nearly of cutting-edge disaster communications.
half of all quotes came from travel industry sources. The For example, Craig Clarke, a civilian analyst for the US
tabloids quoted regulators such as the CAA only 5% of Marines, used social media to support marines on the
the time, but were more likely to quote science sources ground. According to the Lessons from Haiti report:
(12%). Broadsheets and UK national news outlets put less
emphasis on weather sources (9% for both), and broad- There were no systematic technology-based connections
sheet and regional outlets quoted fewer air traffic control between the media activists and the military or the large
sources (6% and 8% respectively). humanitarian organizations. Contacts between the media
Despite the dominance of the travel industry in tradi- and the Coast Guard and the Marines took place on an
tional media coverage, bloggers chose not to repeat those isolated, ad hoc basis, through personal connections.31
quotes, but to focus on primary data from weather sources,
FlightRadar24.com, or to quote scientists and regulators. While stakeholders need a plan for communications
Although the sample used in our analysis is small, this that includes the full spectrum of media options, they also
indicates an opportunity for the producers of primary data need a plan for times when communications systems are
to engage directly with the public, and shows that such compromised.
data will find a receptive audience. Recent crises have shown that modern communications
Stakeholders who engage with social media have the networks can be fragile and lack redundancy. For example,
opportunity to forge relationships not just with their own massive inoperability – failed, destroyed, or incompatible
communities but also with each other. These relationships, communications systems – was identified by the US
which begin online, can prove invaluable in times of crisis. Congress inquiry as the biggest communications problem
While challenges are presented by the culture of top-down, in the response to Hurricane Katrina.32 In serious disasters,
closed communication that is common to many organiza- older communications systems may be needed as backup.
tions such as government or the military, openness and the In the early stages of a crisis, social media may be a strong
liberal sharing of information are required for collaboration tool for disseminating information. Social media are not,

30 This was the experience of EUROCONTROL during the ash-cloud crisis which led it to reinforce its capabilities by the time the 2011 ash cloud
occurred. See, for example, Marc Wright, ‘How Twitter helped EUROCONTROL dispel the ash cloud travel crisis’, http://www.simply-communicate.com/
case-studies/company-profile/how-twitter-helped-eurocontrol-dispel-ash-cloud-travel-crisis.
31 The Knight Foundation, ‘Media, Information Systems and Communities: Lesson from Haiti’, 2010, http://reliefweb.int/sites/reliefweb.int/files/resources/
F156DD1E2F9D2D0085257815005DD82F-Full_Report.pdf.
26 32 Tom Davis et al., ‘A Failure of Initiative: Final Report of the Select Bipartisan Committee to Investigate the Preparation for and Response to Hurricane
Katrina’, Congressional Reports: H. Rpt. 109-377, US Government Printing Office, 2006, p. 163.

www.chathamhouse.org
 The Battle for the Airwaves

however, a guaranteed recipe for communications success. (in a non-crisis environment) as well as during and after
Information quality and consistency can vary substantially, a crisis. For example, it is important to know which social
particularly when the number of experts, government media tools proved effective, whether information assets
agencies, businesses and other key stakeholders engaged and were reused, and if so where. Most critically, it is important
active remains unbalanced. to know how the communications were relayed by others.
As with other forms of communication, organizations This knowledge can be used to further refine the crisis
can learn a lot from measurements taken regularly before communication strategy.

Box 10: Shaping the debate during the 2010 ash-cloud eventa

In the case of Eyjafjallajökull, the travel industry as a whole dominated the traditional media conversation – or the
public narratives – with 37.5% of the quotes swamping the voice of scientists (6.9%) and air traffic control (9.2%).
This was due to a wider range of travel industry organizations being quoted – 62 in total, as opposed to 13 science
organizations and five air traffic control organizations. Even though many of the travel industry representatives were
quoted only once, collectively they overwhelmed other voices.
The traditional media’s bias towards covering the story primarily from a travel industry perspective diminished the
opportunity for the public to form a well-rounded understanding of events. Not only were scientists’ voices marginalized,
but the focus on consumer travel obscured other disruptions such as the ability to deliver organs for transplants and
supply-chain issues for companies reliant on air freight. This made it harder for businesses and members of the public
to assess levels of risk and plan accordingly.
Throughout the eruption, the Icelandic Met Office (IMO) provided copious data in English, including earthquake
maps,b tremor datac and regular updates on the eruption, but these went largely unused by traditional media.d This
stands in contrast to the blogging communities. For example, Erik Klemetti, assistant professor of geosciences at
Denison University, Ohio, and writer of the blog Eruptions, used such data along with information from many other
sources to help explain the event to his readership.e Klemetti’s community was itself very active in his blog’s comments,
locating primary sources and answering one another’s questions.
Other volcanologists and geologists took part in the online discussion on the eruptions, alongside enthusiastic
amateurs and people with no background knowledge at all. The result was a very well-informed community capable
of understanding some of the more obscure information produced by the primary sources. Many community members
learnt a great deal of science by engaging with the event via blogs and communities of this type. Indeed, Klemetti’s
blog is an exemplary model of science communication.
Despite the opportunities afforded by easy access to both data and expertise, the traditional media did not often
examine in depth the more complex issues that the eruption brought to light. The traditional media also did not use
visual aids such as graphics, diagrams or visualizations to help explain what was happening and why. Part of this may
be a simple lack of scientific understanding in the major news outlets, or it could be the assumption that the public
has no interest in looking ‘under the hood’ of scientific events. The popularity of quality science blogs indicates,
however, that there is an appetite for such analysis.

a Analysis and data in this box supplied by Suw Charman-Anderson, 2010.

b Live data available at the IMO website: http://en.vedur.is/earthquakes-and-volcanism/earthquakes/myrdalsjokull/.
c Live data available at the IMO website: http://hraun.vedur.is/ja/englishweb/tremor.html.
d Rolling updates available at the IMO website: http://en.vedur.is/earthquakes-and-volcanism/articles/nr/1884.
e Klemetti’s biography available at: http://personal.denison.edu/~klemettie/Welcome.html. His blog is now at http://bigthink.com/blogs/eruptions, but
during the eruption was hosted at http://scienceblogs.com/eruptions/.
27

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

Mainstream media (MSM) also help to shape messages in It should be noted that there is very little information
the social media sphere. However, although the Chatham
33
available about how the public uses the internet, social
House analysis shows that MSM were a significant player media and email. In the UK, for example, both Ofcom and
in the ash-cloud case, others were more dominant online. the Office for National Statistics produce general infor-
The influence of MSM is likely to be more subtle: they mation, but the more focused research needed to inform
may influence their audience’s ideas, which are then inter- the kinds of online communications strategies that are
preted and translated into social media. At the same time, required is simply not being carried out. The creation of a
news from the ground (e.g. by tweets from friends) can global Internet Institute to carry out original quantitative
challenge the MSM picture. It would also be beneficial and qualitative research has been recommended previously,
for stakeholders, as a group, to benchmark their metrics, in Carnegie UK Trust’s report Making the Connection: Civil
agreeing a suite of tools for web and social media analysis Society and Social Media.34 This would make even more
and sharing data, so that each organization understands sense at EU level: an independent body, focusing not just
more clearly the online landscape in which it is working on issues of the moment but also on longitudinal studies
and the nature of the community within which it is partici- providing data on trends and variations, would be inval-
pating under normal conditions. uable to government, business and NGOs alike.

33 In this report, the terms ‘mainstream media’ and ‘traditional media’ refer collectively to newspapers, TV broadcasters, radio stations, their respective
websites, and the websites of newer news organizations that follow a traditional model.

28 34 Suw Charman-Anderson, ‘Making the Connection: Civil Society and Social Media’, Commission of Inquiry into the Future of Civil Society, Carnegie UK
Trust, 2010, available at: http://suw.charman-anderson.com/files/Published-Makingtheconnection.pdf.

www.chathamhouse.org
 will be essential to enhance preparedness in coping with the
unexpected.
Although uncertainties will remain a common feature of
HILP events, decision-making should as far as possible be
based on scientific and technical evidence. Links between
5. Conclusions and national scientific advisers and national civil contingency
agencies – before, during and after an event – should be
Policy Implications strengthened, and the transparency of risk-based decisions
should be enhanced. Mandatory post-crisis impact assess-
ments would be a step in the right direction – to consider
how crisis decisions were taken; the basis of risk decision-
making processes; and the consequences (positive and
negative) for society, the economy and the environment,
Strengthening processes and transparency with a particular focus on vulnerable communities.
Evidence from many HILP events of the last decade
Recent events, such as Hurricane Katrina, Deepwater indicates that impacts are felt unevenly across the world,
Horizon and the nuclear reactor meltdowns at Fukushima, with the most vulnerable and poorest communities dispro-
have called into question whether policies and regulations portionately at risk. Governments need to work with and
currently reflect the real risks (and the real cost) associated alongside civil society organizations to ensure that the
with high-impact, low-probability events. With growing needs of vulnerable communities are identified and are
environmental and resource pressures, incorporating costs properly considered across the full crisis planning and
of externalities should be a priority in preparing and response cycle (before, during and after). Pre-emptive
responding to crises, factoring in potential costs related to strategies to reduce vulnerability to shocks before they
accidents and disasters. Moving forward, industry bodies occur are likely to include social protection, disaster risk
and safety regulators need to work in coordination with reduction and climate adaptation. During crises, arrange-
governments, business and the science community to ments should be made for the real-time identification of
review risk-assessment practices in critical infrastructure particularly affected social groups or countries, and for
sectors in the light of worst-case scenarios. the rapid disbursement of financial and practical support
Although the precise nature of HILP events cannot always where national organizations are unable to cope or where
be known in advance, there are a limited number of cross- the consequences are cross-border in nature. After crises,
cutting responses to the consequences of a crisis (mass policies to help the worst-affected groups recover and
evacuation and treating sick people, for example) compared build back assets should be prioritized.
with hundreds of potential risks. It can therefore be counter-
productive to focus only on event or sector specifics –
although for some types of risk, highly specific technical Stepping up communications in a crisis
and engineering capabilities remain essential. Robust but
adaptable structures for coordinated decision-making Communicating during a crisis poses a dilemma for many
are crucial because when stakeholders make different stakeholders, particularly when uncertainty abounds,
judgments about risk during a crisis public confidence can and especially for scientific agencies and government
be rapidly eroded. Sharing best practice and, where relevant, departments. There is a limited window in which stake-
capacity, especially across sectors, and red-teaming HILP holders can engage effectively during a crisis – delay
scenarios with key decision-makers – focusing in particular risks ceding the public discussion to noisier voices –
on critical sectors such as transport and communications – but inaccurate or poorly handled communications can 29

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

undermine legitimacy during the event and long into the the communication of scientific and risk-related infor-
future. For example, the Japanese government came under mation for media and science institutions during a crisis,
intense criticism for a perceived slow and non-transparent reflecting the new opportunities and challenges presented
response to the nuclear disaster, particularly in terms of by social media.
communicating with the public.
The availability of consistent, accurate and reliable
information is one of the cornerstones of effective crisis Enhancing business resilience and
management. During crises, information changes quickly responses to shocks
and different types of information are needed at different
stages as events unfold. To facilitate effective information Greater globalization and a lack of sufficient preparation
flows, there should be independent, high-quality hubs for HILP events create vulnerabilities for both the global
(national or regional) for up-to-date risk notification and economy and individual businesses. An important first
provision of scientific information in a crisis, supported step for businesses is to review the vulnerability of their
by governments, businesses and industry. These one-stop business models to high-impact, low-probability events.
centres should be created to aggregate information and This includes cost-benefit analysis of options such as
advice from official sources with information provided by shifting to regional hubs and storage centres for non-
individuals via social media networks. This would become perishable goods to avoid urgent intercontinental trans-
known in advance as the go-to place in a crisis for stake- portation. As the ash-cloud incident indicated, transport
holders, with enhanced capacity to meet huge increases in risks are often more difficult to overcome for perishable
traffic during a crisis. goods trade. But in some instances, different packaging
Social media and other information and communi- and storage methods may permit delivery by land and
cation tools will continue to play even larger roles in sea instead of air. Regional airfreight hubs might reduce
crisis management and preparedness. Microblogging sites some risks but would raise distribution costs. Indicators of
like Twitter, blogs and social networking sites allow the business resilience should be developed that can actually
transmission of critical information to a broad audience be audited or reported on and passed on to stakeholders
in real time, speeding up communication (as well as or the stock market, to bolster incentives for investing in
misinformation) and increasing awareness across the resilience.
board. Tools are needed to enhance and streamline the use However, increasing business resilience goes beyond
of social media in ways that are tailored to major events ensuring continuity despite significant disruption, or
– for example, to aggregate social network messages in quick recovery from a crisis. It is also about protecting
a manner relevant to different stakeholders or to solve profit margins and being better placed than competitors
analytical challenges rapidly with crowdsourcing. to capitalize on game-changers, such as disruption to
Individual organizations, such as regulators and transport networks. The private sector, supported by
government bodies, should also step up planning for governments, needs to invest additional resources in
communications in a crisis, including establishing a training and investment in holistic ‘business resilience’
robust website (for example, a ‘dark site’ prepared in planning, especially for small businesses. Rapid and broad
advance but only made available to the public when a take-up of a new international standard for preparedness
crisis hits). Research for this report shows that cultivating and continuity management systems for organizations
networks on social media long in advance of a crisis can (ISO 22301), due to be published in spring 2012, is also
be decisive in ensuring the successful communication of essential, particularly among small and medium-sized
key messages. enterprises (SMEs). Governments should support the
National science institutions should work together to creation of networks of SMEs to share best practice. In
30 develop, strengthen and promote effective guidelines for parallel, governments could also promote the ISO mark

www.chathamhouse.org
 Conclusions and Policy Recommendations

across industry and the public. This would help ensure Creating a multi-disciplinary reference library around
that adopters achieve a competitive advantage by demon- HILP events and their impacts would allow analysts to
strating a commitment to robust business continuity systematically build up a record of observations that
management. can help quantify the impacts and, by analogy, similar
Although state aid can fulfil a vital role in alleviating risks that might arise in the future. Mistakes made in
paralysis during and immediately following an event, impact studies can be used to improve future assess-
concerns remain around issues of anti-competitiveness ments, creating a more reliable reference system to
legislation and market distortion. Governments should provide faster and more accurate analysis when faced
work with the insurance industry to set up a global pooling with recurring events, and to improve policy planning.
system for reinsurance to address future disruptive events Arrangements for government support (state aid) to
and review existing arrangements regarding the provision businesses during high-impact, low-probability events
of state support to businesses during HILP events. should also be reviewed.

31

www.chathamhouse.org
 Appendix 1: Acronyms

ANSP Air Navigation Service Provider

CAA Civil Aviation Authority
CANSO Civil Air Navigation Services Organisation
CCS Civil Contingencies Secretariat
COBRA Cabinet Office Briefing Room (Alpha)
DFS Deutsche Flugsicherung (German Air Navigation Service Provider)
DGAC La Direction Générale de l’Aviation Civile (French Directorate General for Civil Aviation)
DLR Deutsches Zentrum für Luft- und Raumfahrt (German Aerospace Centre)
DSNA La Direction des Services de la Navigation Aérienne (French Air Navigation Services Branch)
DWD Deutscher Wetterdienst (German Met Service)
EASA European Aviation Safety Agency
FAA Federal Aviation Administration
FCO Foreign & Commonwealth Office
FEWS NET Famine Early Warning Systems Network
GDACS Global Disaster Alert and Coordination System
HILP High-Impact, Low-Probability
IATA International Air Transport Association
IAVW International Airways Volcano Watch
ICAO International Civil Aviation Organization
IFALPA International Federation of Airline Pilots' Associations
IMO Icelandic Met Office
ISAVIA Icelandic CAA and Airport Operator
ISO International Organization for Standardization
IVATF International Volcanic Ash Task Force
LBA Luftfahrt-Bundesamt (German Civil Aviation Authority)
MSM Mainstream Media
NACME National Airspace Crisis Management Executive
NAT North Atlantic (referring to region of North Atlantic airspace)
NATS National Air Traffic Services
NFZ No-Fly Zone
OCHA Office for the Coordination of Humanitarian Affairs (UN)
OEM Original Equipment Manufacturer
OSOCC On-Site Operations Coordination Centre
PSA Probabilistic Safety Assessment
SAGE Scientific Advisory Group in Emergencies
SARS Severe Acute Respiratory Syndrome
UNISDR UN International Strategy for Disaster Reduction
USGS US Geological Survey
USHHS US Department of Health and Human Services
VAAC Volcano Ash Advisory Centre
VATF Volcanic Ash Task Force
WHO World Health Organization
WMO World Meteorological Organization

32

www.chathamhouse.org
 Appendix 2: Methodology

This report examines the impacts of the ash cloud and the response of stakeholders. It seeks to draw out the lessons from the Eyjafjallajökull eruption – a
relatively ‘benign’ high-impact event – for other global threats such as pandemics, terrorist attacks, a radiological event or fuel crises. While different
types of events do of course have unique characteristics, they share three fundamental aspects. First, they have the potential to disrupt the international
economy and the ability of national governments to respond. Second, in preparing for and responding to these threats, governments, businesses and
other stakeholders face significant scientific and other uncertainties. Third, the potential impacts are greatly amplified by the nature of our interde-
pendent world. Additionally, in most instances worst-case scenarios can only be addressed through action at the transnational level, be it via govern-
ments or along critical supply chains.

Table A1: Methodological components

Research Summary of research activity Risk Contest over Mapping the Battle for the
component governance and science, impacts airwaves
preparedness uncertainty
and
legitimacy

Formal Interviews with high-level decision-makers in key stakeholder

interviews organizations and other informal discussions. This included
regulators, civil servants, scientific bodies and businesses.

Survey of Over 40 organizations responded to our survey of the impacts,

businesses responses and lessons from the ash cloud. The survey was
circulated to over 300 firms (see Appendix 2 for respondents).

Economic and An assessment of the economic consequences of high-impact

social analysis events – and specific economic analysis of the cost of the
ash cloud to key sectors (see Appendix 4). A related piece of
research considered the social impacts of the ash-cloud event.

Trade and supply Systematic review of EU trade data to identify goods that
chain analysis are largely air-freighted (and therefore vulnerable to a
disruption to aviation) and goods that were severely affected
in the ash-cloud event.

Institutions and Desk-based mapping of institutions, their linkages and

decision-making decision-making processes (see Appendix 5). This was tested
mapping and refined following the interviews. This component also
explored approaches to civil contingencies planning and risk
management by governments and businesses.

Communications Comprehensive mapping exercise to analyse the media

and public and public response to the ash-cloud event, based on an
perception assessment of public communications and the role of (social and
traditional) media in shaping public perception and responses.

Research was conducted between June 2010 and April 2011. The research methodology is reflected in the four analytical chapters of the report: risk
governance and preparedness; mapping the impacts of the event; the contest over science, certainty and legitimacy; and the battle for the airwaves.
Each of these analytical dimensions was explored using four or five separate but overlapping research components, ensuring that information could
be triangulated, cross-referenced and reinforced and discrepancies identified. Table A1 shows the links between the key research components and the
analytical chapters.
The methodology for the project combined desk-based research with interviews, a survey and systematic data analysis. The mapping of the key insti-
tutions at European and national level and their decision-making processes continued throughout the project, drawing on the interviews and the survey.
High-level decision-makers from key organizations and institutions involved in the ash-cloud event were interviewed between August 2010 and
January 2011. Questions covered scenario and contingency planning before the event; how the event unfolded; key institutional linkages; communica-
tions strategies; and priorities for the future. Semi-structured interviews were held with representatives of the UK Met Office, the UK Cabinet Office
(Civil Contingencies Secretariat), the European Commission, the UK Civil Aviation Authority, Eurocontrol, the Confederation of British Industry and
33 a prominent global engine manufacturing company. Informal discussions were also held with a range of companies and other stakeholders throughout 33
the project. Two major airlines and two national air traffic control agencies including UK NATS fed information to the project via our survey.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

In October 2010 a confidential survey was sent to over 300 small and medium-sized enterprises (SMEs), global corporations and trade associa-
tions (see Appendix 3). The recipients were asked to outline the impact of the volcanic disruption on their operations – both negative and positive
– and their respective responses. Out of the 40 responses a large proportion came from aviation, transport and food sectors as well as government and
regulatory bodies. Critical questions included what could be learned from the ash-cloud episode in terms of vulnerability and resilience; how critical
supply chains were affected; what would have been the consequences if the ash cloud had persisted; and how information was communicated.
A detailed review of previous economic studies of high-impact events (such as avian flu, SARS, Hurricane Katrina) was conducted to explore how
the economic impacts of a high-impact, low-probability event can be measured (and what cannot be measured) and how impacts tend to flow across
the economy via critical sectors. The research team also assessed the specific economic impacts of the ash-cloud event and what would have happened
if it had persisted. This focused on the key industries (aviation, other transport, perishables, pharmaceuticals, high-value electronics, service industries etc.).
This analysis was reinforced with a separate exercise drilling down into EU import and export data to identify which products were affected by the ash
cloud and the extent to which high-value intermediate goods are air-freighted.
In order to understand how information was communicated by key stakeholders during the crisis – and the response of the public – a comprehensive
analysis of public responses and communications during the event was undertaken by Cambridge IP. This was based on an assessment of public commu-
nications and the role of (social and traditional) media in shaping public perception and responses. A bespoke web search and indexer was constructed
to intelligently mine information from blogs, news sources, twitter and other sources. This resulted in a database of 87,202 records from which additional
analysis was undertaken. In parallel, a separate piece of analysis was supplied by social media expert Suw Charman-Anderson. This examines the role
of the media and lessons for communications strategy related to the ash cloud and other events, and systematically considers the flow of information
between key media outlets and key stakeholders.

34

www.chathamhouse.org
Appendix 3: Chatham House Survey – Questions and Responses

A short confidential survey aimed at businesses – from SMEs to global corporations – trade associations and unions was circulated to over 300 contacts.This
provided an opportunity for companies to explain both negative and positive impacts of the volcanic disruption on their operations, and how they responded. In
total there were 40 responses to the survey, with a large proportion coming from aviation, transport, and food sectors as well as government and regulatory bodies.

Chatham House Survey

The eruption of Icelandic volcano Eyjafjallajökull in April 2010 resulted in an unprecedented disruption to air transport, especially in Europe. A major new
Chatham House project is analysing the impact of the event and asking what we can learn about managing future shocks to international supply chains.
This short (around 5 minutes) confidential survey is an opportunity for companies to explain the impact on their operations – both negative and
positive – and how they responded. The survey results will be fed into high-level discussion with policy-makers and business leaders during early 2011.
The survey is aimed at senior representatives from businesses – from SMEs to global corporations. We also welcome the participation of other
organisations (for example, trade associations, trade unions and NGOs).
Critical questions include: what can we learn from this episode in terms of business flexibility and resilience? How robust are critical supply chains?
How well did decision-makers perform as events unfolded? How effective were the lines of communication between businesses, policy-makers and
others? How were key decisions taken and were they correct?
All submissions will be handled sensitively and securely – and you can choose to exclude all details specific to your company from any published
analysis (see next page).
For further information on the project please see www.chathamhouse.org/ashcloud.

Q1: Please complete the following details to allow us to validate your …… Manufacturing …… Transport/logistics
response. An email address is required so that we can confirm the …… Retail …… Other (please specify)
submission is authentic. …… Tourism
Your personal details will be held confidentially and will never be
supplied to third parties. Q5: How would you describe the organisation you represent?

Your name: _______________________________ …… Private company …… Not-for-profit

Company: _______________________________ …… Public limited company …… Other (please specify)
Job Title: _______________________________ …… Trade association
City/Town: _______________________________ …… Public sector/state-owned
Email Address: _______________________________ enterprise

Q2: A concise summary of the survey results will be prepared for partic- Q6: What is your primary focus within the organisation?
ipants. Would you like to receive this? ________________________________________________________
________________________________________________________
…… Yes …… No
Q7: Please specify how the event affected your organisation, positively
Q3: Please tell us how to handle your submission (answer required): or negatively, in the following areas.
++++ indicates an extremely positive impact,
 Full anonymity – no specific reference to my company/organisa- – – – – indicates an extremely negative impact.
tion will be made in any publications/project output
 Please check with me before making any specific references about …… Profits/revenue …… Health and safety
my organisation (no reference will be made without prior permission) …… Public image/reputation …… Human Resources
 Feel free to refer to my company/organisation and job role in …… Supplier relations …… Insurance cover
publications and project outputs based on my survey responses …… Customer relations
…… Demand for goods and
Q4: Which sector(s) does your company/organisation operate in? services

…… Air travel/transport …… Heavy industry Q8: How long did it take for you to return to business-as-usual?
…… Finance …… Information &
…… Food Communications …… Less than one week …… More than one month
…… Government/Regulation Technology …… Less than two weeks …… Other (please specify)
35
…… Health …… Insurance …… Less than one month

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

Q9: If the disruption to air travel had persisted, how much longer would Q15: Did you receive targeted information for your industry from regu-
your operations have been able to continue without serious disruption? lators or government? Please describe:
________________________________________________________
…… 1–2 days …… 2–4 weeks ________________________________________________________
…… 3–7 days …… more than one month
…… 1–2 weeks …… N/A Q16: How well did these different types of stakeholders perform during
the event from the perspective of your industry – in terms of a) deci-
Q10: What actions did your organisation take during the event in the sion-making and b) communications
following areas? (please describe)
Decision-making performance:

…… Profits/revenue …… Health and safety 1 2 3 4 N/A
…… Public image/reputation …… Human Resources (1 = very poor performance, 4 = very good performance)
…… Supplier relations …… Insurance cover
…… Customer relations …… Scientific bodies …… European Commission
…… Demand for goods and …… Regulators …… Media
services …… Government – home …… NGOs
…… Government – other …… Trade unions
Q11: How effective were your organisation’s systems in managing the impact national …… Private enterprise
on the following areas?
1 = not effective, 4 = highly effective Communications performance:
1 2 3 4 N/A
…… Profits/revenue …… Health and safety (1 = very poor performance, 4 = very good performance)
…… Public image/reputation …… Human Resources
…… Supplier relations …… Insurance cover …… Scientific bodies …… European Commission
…… Customer relations …… Regulators …… Media
…… Demand for goods and …… Government – home …… NGOs
services …… Government – other …… Trade unions
national …… Private enterprise
Q12: Has the event caused you to make changes to your business
strategy? (Please describe) Q17: What improvements, if any, could be made to the approach of
regulators/policy-makers for a high-impact event such as the volcano?
Q13: After the initial eruption, roughly how long was it before your ________________________________________________________
organisation received targeted information/advice from the following: ________________________________________________________
________________________________________________________
…… Regulators …… International bodies ________________________________________________________
…… Government …… Other (please specify)
…… Industry organisations Q18: Anything else you want to tell us?
________________________________________________________
Q14: During the event, how did your organisation communicate key ________________________________________________________
messages to the following? (you can select more than one) ________________________________________________________
________________________________________________________
Company website | Bulk email | Advertising | Social networks |
Comment pieces/articles | Other Q19: Would you like us to contact you again/invite you to a workshop
about this?
…… Consumers …… Industry ________________________________________________________
…… Government/regulators …… Other (please specify) ________________________________________________________
…… General public ________________________________________________________

36

www.chathamhouse.org
Appendix 4: Measuring the Economic Costs of a High-impact Event: the Stylized GDP Model

This appendix sets out a methodology for estimating the economic impact of extreme events, as developed and applied by Vanessa Rossi in Chapter 2
of the report.
We can derive an estimate of the overall economic impact of an event risk by summing up the varying losses across the various expenditure compo-
nents of GDP. Adopting a simplified form of ready reckoner and grouping terms by their typical responsiveness to general shocks, the overall effect on
GDP for each period can be summarized as:

(1) GDP loss = a (XS) + b(CDIS) + c (XG) + d(I) + e (C-CDIS) + f(GC) – g(MS) – h(MG)
(Fast response (Slower response (Very slow (Policy reaction – (Correlated with
variables) variables) response variable) may be 0) other variables)

where CDIS is the discretionary part of consumption (and may be further disaggregated according to the shock being examined, say into travel
and tourism versus other) while C is total consumption. XS and XG are exports of services and goods, MS and MG are imports of services and goods.
I is total economy investment, which could be further subdivided into business (IB), housing (IH) and infrastructure (II) investment depending on the
nature of the shock. GC is government consumption spending (which tends to be relatively static except in cases where political changes are part of the
shock). All these variables are the ex-ante values of the expenditure components (i.e. either the current data or, preferably, the forecast values prior to
the shock for the period of the shock).
The coefficients (a, b, c etc.) are all functions of a suitable scalar (S) for the specific shock, which may be the percentage of the population, companies
or transport routes affected by the shock (scale factors therefore pick up the intensity and spread of the risk) and other factors such as sentiment, which
may be high in the case of the international disruption of transport services or infectious diseases but low if the threat is highly localized or very slow
developing, such as erosion of coastline or skin cancers).
For example a = function (S, sentiment, other factors)
The coefficients increase as S rises.
This form of simplified loss function separates the ‘fast response’ (typically volatile) sectors from the slow to change – we would generally expect to
see rapid large losses in XS and CDIS, for example, but very little response from non-discretionary consumption (C-CDIS) unless the shock were to be
specific to the non-discretionary sector (e.g. in the case of food shortages). We might expect coefficient f to be close to zero except for policy responses
– in some cases, government spending GC could increase as an offset to a negative shock (f would be negative). Some coefficients may be similar (e.g.
coefficients a and b) but we expect coefficients a and b to be most often larger than c and d and considerably larger than e.
The most rapid and large impacts come from CDIS, the most critical factor along with the estimate of b, which is a scale factor between zero (no
losses) and one (all existing CDIS is lost as consumers cut back spending). The other terms will typically be relatively small over one year but losses
would mount in following years under a prolonged shock.
We may reasonably assume that total CDIS is at least 10% of world GDP (if only based on the scale of luxury goods sales and the travel and tourism
sector), possibly as much as 30% of GDP for the wealthiest countries. A serious threat from an event such as a pandemic could rapidly decay this component
of GDP – at the limit all of the international component of such spending could disappear during a prolonged incident such as a dangerous infectious
disease outbreak or severe security threat. The statistics highlight just how substantial is the risk to the global economy from shocks impacting on CDIS.
The equation also highlights the offsetting influence of imports on GDP. In theory, if imports change faster than exports, GDP might rise in response
to a negative shock – this would not really happen, however, as job losses and consumer reactions would still create cuts in CDIS and other GDP compo-
nents. Nevertheless this point does highlight the fact that the equations above are only valid for country analysis (and reveal the relative importance of
net trade and the proportions of each component in each country’s GDP). They are not valid at the level of global GDP (as export and import losses
have to be equal by definition, cancelling each other out).
At the global level with trade netted out, and assuming GC is broadly unchanged, then we can simplify (1) as the sum of the fast-response, slower
and very slow-moving expenditure functions:

(2) Global GDP loss = { b(CDIS) + d(I) + e(C-CDIS) }

The schematic equation outlined above crudely illustrates the way in which GDP losses are built up and the importance of the structure of GDP in
any analysis – the scale of the incident (S, influencing b, d and e) is crucial but so is the proportion of CDIS in GDP (as well as XS for individual country
estimates of GDP losses). Using the existing proportions of each expenditure component prior to the shock, then (2) can be rewritten as (3) below:

(3) Global GDP loss = { b(s1 GDP) + d(s2 GDP) +e(s3 GDP) } = (b*s1 + d*s2 + e*s3) GDP

Thus the impact on GDP becomes a weighted average (weighted by GDP shares) of the reaction speeds of each of the three types of expenditure b,
d and e (all between zero and 1): b is high, possibly above 0.5, d is low and e is virtually zero in the first year. These parameters vary over time and are
all functions of the type and scale of shock. 37

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

For a threat such as a short-lived weather problem or an infectious disease outbreak, it is the impact effect on GDP over the first quarter or year that
will dominate most calculations (unless death rates were to be very sizeable, causing a marked drop in population and long-tail GDP losses). The impact
effect is certainly a critical variable for short-term policy responses and economic planning.

Figure A1: Illustration of the impact of a persistent shock such as the ash-cloud incident, assuming that
mitigating action eventually starts to reduce the negative effects, summing to the overall loss in GDP in
each period

Varying impacts of a
shock and mitigation
in the long run
Total effects
GDP loss

Medium-term impacts

Fast-response impacts

Slow build
01

02

03

04

01

02

03

04

01

02

03

04

01

02

03

04

01

02

03

04
Q

Q
1

5
Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0

Y0
A more complete treatment of all the possible economic losses over the short and long term would mean that long-tail impacts would have to be
cumulated over the years including and following that of the initial incident. Alternatively, these losses can be summed as the present discounted value.
Measuring potential long-term costs in this way is important in the cost-benefit analysis for preventative or mitigating action and the costs of related
investment spending (also, as highlighted in other studies, raising the issue of the appropriate discount rate to be used in assessing net present values).
There is an equivalent loss to the expenditure accounts across industry sectors:

(4) GDP loss = x (services) + y1 (manufacturing) + y2 (construction) + z1 (utilities) + z2 (agriculture)

(Fast response) (Slower response) (Very slow response )

This breaks out the losses on a sectoral basis but the overall results should be consistent with the expenditure method shown above: much of the
short-term impact of the ash-cloud incident was in the fast-moving services sector. There were only minor effects in manufacturing and virtually no
impacts on construction, utilities and agriculture. Services are a very large sector, however (about 75% of the UK economy). Therefore it would be
important to break this out further into foreign and domestic travel and tourism, other leisure and the rest in order to assess losses – and gains – from
any incident more accurately.

38

www.chathamhouse.org
Appendix 5: Agencies and Institutions Involved in an Ash-cloud Crisis (Selected Countries)
and Decision-making Linkages during the Ash-cloud Event

Phase I: Sticking to the international protocols (the IAVW)

In Phase I the focus was on the application of internationally agreed procedures for monitoring and reporting ash. Rather than allowing airlines to
make the call on safety, aviation regulators chose to close air space where scientific modelling indicated ash was present at any concentration. Airspace was
closed rapidly across northern Europe following the major ejection of ash into the atmosphere on 14 April 2010. Countries and agencies would have been
justified in claiming they were ‘aware and prepared’ for an ash cloud at this point – only as the event dragged on were the limits of the preparedness revealed.
The global system for monitoring and advising on the presence of atmospheric ash, the International Airways Volcano Watch (IAVW), is symbolized
within the first sphere of Figure A3. This critical infrastructure aided the exchange of intelligence between a number of stakeholders throughout the event.
The IAVW was established in 2002 by the International Civil Aviation Organization (ICAO – a United Nations Technical organization), in conjunction with
the World Meteorological Organization (WMO).
The London Volcanic Ash Advisory Centre (housed within the UK Met Office) took a lead role, issuing advice on the location of ash every six hours.
National civil aviation authorities (responsible for aviation safety) and air traffic control services (which have the operational responsibility for controlling
air space) used advice from the London VAAC to inform decisions on whether airspace should be restricted. Initially, the approach taken by regulators was
to restrict airspace wherever analysis from the London VAAC predicted the presence of ash in a given location (at any altitude).
The airlines were informed of air space restrictions by air traffic control. An alternative approach, preferred by some airlines, would have been to allow
airlines to take the decision on when it was safe to fly, avoiding ash when it was visible. The regulators chose not to take this approach on the basis of the
crowded nature of European air space as well as – in their view – the potential risks to passenger safety.

Phase II: Searching for a solution

The second phase began after the widespread closure of air space on 15 April and lasted until 19 April. During this time, governments came under increasing
pressure from businesses and stranded passengers to open up the skies. The appropriate locus of decision-making was not always clear, and the lack of a
harmonized EU approach raised further questions about the legitimacy of rule-making procedures.
The second phase comprised the actions of the parties during the peak of the crisis. Airlines, engine and aircraft manufacturers, governments and
regulators held a series of international teleconferences to discuss their approach and possible rule changes that could help get planes flying again. The
London VAAC and other European Met authorities drew on the input from a range of other scientific agencies to continually improve and verify their
volcanic ash advisories, and reported this information back to the teleconferences.
International teleconferences were initially hosted by the UK CAA, with the European Commission initiating its own teleconference dialogue as the
disruption worsened. Although responsibility for aviation safety within the EU ultimately lies with its member states, pressure grew during the event to
agree and coordinate an alternative approach at EU level. Specifically, the questions centred on whether to allow flights in areas of low ash concentration
– issues addressed in Chapter 3.

Phase III: The European Commission intervention

The impact of the restrictions reached its peak on Sunday 18 April, with commercial air traffic having ceased in 23 member states. Under the growing
political pressure and with continuing uncertainties, changes to the rules were rapidly drawn up and agreed upon. Decision-making was apparently influ-
enced by economic and political interests as well as the scientific evidence. In addition, the speed and ad hoc nature of the new rule-making process led to
widespread confusion among key operational and regulatory stakeholders.
The search for an alternative risk management system – on the rules on flying in ash – created a parallel process with the existing international proce-
dures. This was an ad hoc approach to finding a solution, rather than a pre-arranged protocol. It became clear that a prolonged event had not been foreseen
by governments; countries were therefore unaware and unprepared.
On the morning of Monday 19 April, an Extraordinary Meeting was co-chaired by Eurocontrol and the European Commission in Brussels with the
National CAAs, ANSPs, airlines and the Spanish EU Presidency to discuss three options moving forward: maintain the status quo – avoiding flying wherever
ash is present; follow the ‘US approach’ – closing the area immediately around the volcano but allowing airlines to make the decision on where was safe to
fly; or a hybrid approach – this would allow flights in area of lower ash concentration, subject to the airlines having undertaken a risk assessment, but no-fly
zones could still be established by individual EU member states. The participants agreed upon the third option (described as the three-band model), which
received approval from EU Transport Ministers at 16.00 that afternoon, in an additional Extraordinary Meeting.
The option included:
zz A no-fly zone (NFZ), consisting of a high ash density area (Black Zone) and additional 60 nautical mile buffer zone surrounding this;
zz A second zone where flights would in principle not be impeded (subject to agreed risk assessments and measures) even though some ash was
present, and where the decisions about operations would be taken by national authorities;
zz A third zone which was not affected. 39

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

Six days after air space had initially closed, regulation moved from an ‘ash/no ash’ distinction (as indicated by ICAO guidance) to a more differen-
tiated system, marking contaminated airspace according to ‘how much ash’ was present. However, at this point there was no clear public statement
from engine and aircraft manufacturers about where it was safe for aircraft engines to operate in this partially contaminated zone. A meeting on the
morning of Tuesday 20 April between the UK Secretary of State and airline industry representatives concluded that ‘while it had previously been
hoped that engine manufacturers would support such a revision, consensus had not yet been achieved’. It was therefore necessary to consider whether
the industry could temporarily adopt the proposed revised no-fly zone prior to approval from manufacturers; it is important to note that this was
after many EU countries had opened their airspace and introduced the new concentration limits.
Moreover, it was at the discretion of national regulators whether to permit flights on the basis of the revised approach. Therefore, while the
European Commission Vice-President Siim Kallas was keen to stress that ‘all decisions must be based on scientific evidence and expert analysis’,
it was not necessarily clear whether this was undertaken consistently among states. In Germany there was subsequent criticism of the decision to
allow flights.
On the afternoon of 20 April, the Fourth International Teleconference took place. During this consultation confirmation came from key engine
manufacturers that operating in ash concentrations of less than 2x10-3 was possible within reasonable parameters of safety. Following this decision,
European Aviation Authorities reopened airspace more rapidly, reaching near to 80% normal capacity by 21 April and a return to normal operation
across Europe (apart from airspace in southern Finland and minor restrictions in the UK) by 22 April.
It also proved challenging to translate the change in rules into practices that key stakeholders, especially airlines, could easily adopt. Initially
Eurocontrol took responsibility for drawing up revised Volcanic Ash Advisory charts, detailing the no-fly zones and associated buffer region.
However, owing to the time constraints imposed upon Eurocontrol cartographers, an ad hoc method for plotting data had to be adopted. So rapid
was the change in approach that just six hours before it came into force British Airways expressed its confusion. In an email to CAA and UK
Department for Transport executives, the company stated that CEO Willie Walsh ‘remains concerned that there is still a significant level of uncer-
tainty about tomorrow’s operation’, also highlighting that ‘we are now well past the go/no go stage’.
Further eruptions the following month threatened to close airspace once again. On 11 May, following a series of test flights into the contaminated
zones and a revised safety assessment, the UK CAA concluded that the retention of the 60 nautical mile buffer zone was no longer necessary. This
decision further eased restrictions and was progressively adopted across other EU states. Additional dialogue between some aircraft OEMs, the CAA
and the UK Met Office led to a further revision of the Volcanic Ash Advisory Charts on 17 May. These charts essentially increased by a factor of two
the acceptable ash concentration that defined the boundary of the current no-fly zone.

Phase IV: Preparing for a future event

Governments and agencies are currently focused on testing and implementing the new rules, but preparations for a prolonged disruption to aviation
have so far been limited. Following the crisis the search for a more comprehensive agreement on rules and procedures for flying in volcanic ash has
continued, using the three-band model as a foundation. Many stakeholders have expressed scepticism that new rules will be agreed internationally,
in part because in the absence of crisis conditions the pressure to complete the process is reduced.
Under the new European rules the responsibility for risk assessment and management, and for the safe operation of the aircraft, rests firmly with
the operator of the aircraft. This clarification is in line with the requests of airlines during the crisis. However, operation through or within any area
where volcanic ash is forecast continues to be subject to the decisions of national regulators and governments.
ICAO has established a European/North Atlantic Volcanic Ash Task Force (EUR/NAT VATF) to move forward on the development of a harmo-
nized safety risk management framework. This aims to make it possible to determine routinely the safe levels of operation in airspace contaminated
by volcanic ash. The multidisciplinary team of experts from member states and the aviation industry prepared revised Volcanic Ash Contingency
Plans for both regions on 10 June 2010; these gained endorsement from the North Atlantic Systems Planning Group (NAT SPG) and the European
Air Navigation Planning Group (EANPG) in July.
In parallel with its European efforts, ICAO established an International Volcanic Ash Task Force (IVATF) in May 2010 to develop a global
safety risk management framework which would make it possible to determine safe levels of volcanic ash concentrations for operations in airspace
contaminated by volcanic ash.
Yet despite these efforts, there remain questions around whether this approach will be adopted at international level. Current guidance based
on the use of ash concentration thresholds has been adopted as guidance material in the EUR/NAT region but this ‘living document’ is still under
validation and assessment by the IVATF and it is not clear that such an approach could be agreed and applied globally within the International
Airways Volcano Watch (IAVW) system of ICAO. One issue raised by stakeholders is that a new European approach to risk management cannot
easily be enforced on non-European countries. Although they could be required to adopt the risk-management approach proposed for Europe, it
would not be straightforward to verify their compliance.
One year on from the ash cloud, Volcex held a major scenario exercise on 13 and 14 April 2011 with broad stakeholder participation,
including 77 airlines. This exercise tested the implementation of the revised approach to flying in ash at European level. However, it did not
consider the implications of a prolonged event. One major stakeholder told us that if an event were to last longer than a week, there is still
40 ‘no plan – A or B’.

www.chathamhouse.org
 Appendix 5: Agencies and Institutions Involved in an Ash-cloud Crisis (Selected Countries)

Table A1: Methodological components

UK France Germany Iceland EC/EU US/other Global

Civil contingencies/ Cabinet Office Direction de la Bundesamt für Department of Monitoring and Federal Emergency UN OCHA,
emergency Sécurité Civile Bevölkerungs- Civil Protection Information Centre Management GDACS, OSOCC,
preparednessa (DSC) schutz und and Emergency (MIC) Agency (FEMA) ReliefWeb, UNISDR
Katastrophenhilfe Management
(BBK) (DCPEM)b

Aviation NATS (National Air La Direction Luftfahrt-Bundesamt ISAVIA Eurocontrol Federal Aviation ICAO, IAVW, IVATF,
(Air Navigation Traffic Services) générale de (LBA) Administration (FAA) IATA, CANSO
Service Provider and l’Aviation civile Icelandic European Aviation IFALPA
National Aviation Civil Aviation (DGAC) Deutsche Civil Aviation Safety Agency
Authority, Airport Authority (CAA) Flugsicherung (DFS) Administration (EASA)
Coordinator) Direction des Flughafen-
Services de la koordination
Navigation Aérienne Deutschland
(DSNA) (FHKD)

Other transport DfT, Highways Ministère de Federal Ministry Ministry of Transport, DG MOVE US Department of
Finding alternative Agency, UK Port l’Equipement des of Transport, Communications and Transportation
transport options Authority Transports et du Building and Urban Local Gov.
for passengers and Logement Development
goods (BMVBS),
Eisenbahn-
Bundesamt (EBA)

Meteorological UK Met Office, Météo-France, Deutscher Icelandic Met Office EUMETSAT NOAA WMO
information Wetterdienst
Providing advice London VAAC Toulouse VAAC EUMETNET
to other decision-
makers on ash
dispersion and
deposition

Food and farming Defra,c FSA Ministère de BMELV Ministry of Fisheries EFSA,d DG Envt, USDA, FDA FAO
Potential for l’Agriculture et la and Agriculture DG Agri
contamination of Pêche
food supply chain

Environment Defra, EA/SEPA,e Ministère de BMU, Ministry for the EEA EPA, DOE UNEP, UNISDR
Potential Centre for Ecology l’Écologie, de Umweltbundesamt Environment
contamination of soil and Hydrology l’Énergie, du (UBA)
and water with e.g. (CEH) f Développement
fluorine durable et de la Mer

Health MoH, NHS, HSA Ministère de la Bundesministerium Ministry of Health European USHHS WHO
Impact on HPA Santé et des Sports für Gesundheits- Emergency Data-
respiratory health, wesen based Syndromic
food and water Surveillance System
contamination, (SIDARTHa)g
disruption to
healthcare provision
including blood and
transplant services

a http://www.euromedcp.eu/index.php?option=com_weblinks&view=category&id=954%3Aeu-countries-civil-protection-authorities&Itemid=967&lang=en.
b Almannavarnir, Online New Article, 23.4.2010; President of Iceland meets with senior officials from the CPEM. Available at: http://www.almanna-
varnir.is/displayer.asp?cat_id=413&module_id=220&element_id=2323.
c UK Parliament, 14 July 2010. Evidence heard by Lords EU Sub-Committee B on the volcanic ash crisis. Available at: http://www.parliament.uk/
business/committees/committees-a-z/lords-select/eu---internal-market-sub-committee-b/news/evidence-session/.
d EFSA Journal 2010; 8(1):1593 [16 pp.] Available at: http://www.efsa.europa.eu/en/scdocs/scdoc/1593.htm.
e Scottish Environment Protection Agency (SEPA); Volcanic ash cloud – the latest news from SEPA. Available at: http://www.sepa.org.uk/about_us/
news/2010/volcanic_ash_cloud_%E2%80%93_the_lat-3.aspx.
f Centre for Ecology and Hydrology, Online News Achive. Available at: http://www.ceh.ac.uk/news/news_archive/2010_news_item_12.html.
g European Emergency Data-based Syndromic Surveillance System (SIDARTHa), 15 May 2010; SIDARTHa Volcanic Ash Cloud Rapid Public Health 41
Impact Assessment. Available at: http://ec.europa.eu/eahc/documents/health/SIDARTHa_rapid_assessment_volcanic_ash_cloud_15MAY2010.pdf.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

UK France Germany Iceland EC/EU US/other Global

Defence/security MOD, UKHO, UK Ministère de la Bundesministerium No standing army EDA US DOD NATO
Impact on air force Borders Défense der Verteidigung
capability during
the event, potential
role in repatriation
and support to civil
contingencies

Foreign relations Foreign and Ministre des Affaires Auswärtiges Amt Ministry of Foreign RELEX US Department of
Repatriation of Commonwealth étrangères (MAEE) (AA) Affairs State
stranded passengers Office
and wider diplomatic
services

Health and safety Health Protection Les Agences Bundesanstalt für Ministry of Health DG for Health and Occupational
Agency, Health and françaises de Arbeitsschutz und Consumers, European Safety and Health
Safety Executive sécurité sanitaire Arbeitsmedizin Centre for Disease Administration
(AFSSAPS) (BauA) Prevention and (OSHA)
Control

University International CEV Deutsches Zentrum Institute of Earth International

departments/ Volcanic Health für Luft- und Sciences Association of
networks Hazard Networkh Raumfahrt; Volcanology and
Chemistry of the
VMSG Earth’s Interior
(IAVCEI)

Scientific bodies British Geophysical French Aerospace German Aerospace Joint Research US Geological
Association Lab (ONERA) Center (DLR) Council Survey, US National
Research Council
Royal Society

NERC – British
Geological Survey
– NCAS

Scientific
Advisory Group
for Emergencies
(SAGE)

Business Confederation of Bundesverband der National Federation

organizations British Industry, Deutschen Industrie, of Independent
Federation of Small Bundesministerium Business, United
Businesses für Wirtschaft und States Federation of
Technologie Small Businesses

Government House of Commons L’ Office US House of

enquiries Science and Tech Parlementaire Representatives
Committee d’Evaluation des Committee on
choix Scientifiques Science and
et Technologique Technology
(OPECST) Subcommittee
on Space and
Aeronautics

Airlines BA, Virgin Air France-KLM Lufthansa, German Icelandair United Airlines, RyanAir, EasyJet
Wings Delta, South-West,
American Airlines

Engine Rolls-Royce Plc CFM International GE Aviation, Pratt &

manufacturers Whitney

h C.J. Horwell et al., 2010, Respiratory health hazard assessment of ash from the 2010 eruption of Eyjafjallajökull volcano, Iceland. A summary of
42 initial findings from a multi-centre laboratory study. International Volcanic Health Hazard Network. Available at: http://www.ivhhn.org/images/pdf/
iceland_ash_health_report.pdf.

www.chathamhouse.org
 Appendix 5: Agencies and Institutions Involved in an Ash-cloud Crisis (Selected Countries)

Figure A2: Institutions and decision-making during the ash-cloud crisis

International Airways Volcano Watch

Eruption
zzUpon the detection of the eruption, the Icelandic
Met Office contacts the UK Met Office and the
14 April London VAAC.
zzOutputs of the volcanic ash modelling from London
Icelandic
Met Office VAAC are disseminated to aviation stakeholders in
accordance with the established IAVW system.
zzFollowing the ICAD guidance on Volcanic Ash that
VAAC aircraft avoid any kind of ash concentration, NATS
UK Met
Montreal/
Office (after consulting the CAA) begins pre-emptive
Toulouse
IAVW UK VAAC
restriction airspace by issuing NOTAMs.
Met Watch zzUpdated VAAC charts the following day predict
Offices increasingly larger parts of European airspace
ANSPs Civil Aviation
and Authorities will be affected. One by one, European states
Eurocontrol implement similar flight restrictions but the overall
NOTAM approach is uncoordinated.
NOTAM
zzThe UK Civil Aviation Authority upholds that ICAO
Airspace 15 April guidance must be followed in order to maintain
Aircraft in Restrictions
flight (All UK airspace passenger safety.
unavailable by
1200)
Searching for a solution (ad hoc policy
making in a crisis scenario)
Grounded
London VAAC and flights
other Met zzLondon VAAC updates forecast every six hours
Authorities seek – continually acting on input from both national
additional scientific and international scientific authorities and
support UK CAA hosts
series of meteorological competencies.
teleconferences to zzPressure from airlines, representative bodies and
NOAA pool expertise and Engine manufacturers the media mounts upon Civil Aviation Authorities
DLR
DWD reassess CAO 16 April
MWO guidance Airlines National Aviation to review their zero-tolerance position.
Met, Research Authorities
CEV zzUK-CAA hosts the first in a series of International
JRC Councils, EASA
Universities, etc ANSPs Representative
bodies Teleconferences, gathering together airframe
FAA Weather Airframe
USGS and engine manufacturers, service providers,
Météo Group EC teleconference manufacturers
France co-hosted by operators, research and geological agencies and
Eurocontrol
European and International Safety Regulators.
zzWith airspace restrictions at their peak, the
European Commission intervenes, hosting its
Extraordinary meeting
of EU transport own teleconference on the morning of 19 April in
which the 3-band model is put forward.
zzProposals easing restrictions are adopted before
clear confirmation from engine manufacturers
Proposals signed into force by
EU transport ministers over safe to operate limits in volcanic ash.
19 April

European Commission intervention

Adoption of zzNew zoning system effective 8.00 CET Tuesday

‘three-band model’
20 April.
zzEUROCONTROL to provide revised Volcanic Ash
Ash coverage Brief period of forecast, detailing the 'no-fly zone' and update
effectively uncertainty in Test flights every six hours.
decreases by application of new ongoing
approximately zzInitially engine manufacturers still not clear on
zoning system
80% ash threshold and new zoning system fails to
Engine manufacturers specify the exact ash concentration level at which
endorse new zoning system flights could operate.
4th international 22 April zzPermission for flights remains the responsibility
teleconference of the National Aviation Authority.
zzAfternoon of Tuesday 20 April, the Fourth
International Teleconference brings confirmation
Rapid opening of Return to near from key engine manufacturers that operating
airspace normal operation
in concentrations less than 2x10-3 is permissible.
zzAirspace in the EUR/NAT region opens up more
Formation of the EACC Formation of the IVATF May rapidly, reaching near to 80% capacity by 21 April
and review of operations and ongoing review of and a return to normal operation across Europe
in EUR/NAT region operations in volcanic by 22 April.
43

www.chathamhouse.org
 Appendix 6: National Approaches to Decision-making during the Ash-cloud Event

This appendix sets out in detail the institutional responses by three European member states: the UK, Germany and France.
The ash cloud raised a series of public policy issues including those related to health (the respiratory impact of ash and medical supply chains), environmental
risks (food and water quality and related health implications), and the repatriation of stranded passengers. Such challenges are, of course, not specific to an ash-cloud
crisis – they might arise from a range of high-impact, low-probability events. Furthermore previous events have helped shape the institutional structures that handle
civil contingencies. For example, avian and swine flu, SARS and other global threats have informed the national and international structures for responding to a health
crisis. The process and response of UK health agencies during the ash-cloud event closely reflected the structure put in place for pandemic flu.35

The UK
The UK was one of the most affected countries in terms of ash dispersion and deposition, and many UK citizens were stranded overseas during the event. The
UK Met Office also had a central role in forecasting the distribution of ash across Europe.
Less than 24 hours after Eyjafjallajökull went into its explosive phase, the UK’s National Airspace Crisis Management Executive (NACME) convened.
This was attended by officials from the CAA, Ministry of Defence, NATS and the Department for Transport (DfT). NACME met three times daily there-
after, and served as the principal forum for executive decision-makers in UK aviation regulation. Taking its cue from these meetings, the Department

Figure A3: UK decision-making structure during the ash-cloud crisis

Cabinet Office
Civil Cont.
Committee/
Geology and COBR
Volcanology Civil Cont.
SAGE
Group
Methodology Secretariat
Chair: GCSA
subgroup FCO
Aviation
Engineering Department of Health
subgroup DfT Collaborating
HPA with WHO
Input into BADC
EUFAR DEFRA
database MOD
NERC NACME
CAA
Scottish Gov.
ARSF NATS
SEPA
CEH DARD
NCAS Welsh Associate
FAAM AFBI
BGS UK Met UK Flights in UK
Office VACC Devolved
Airspace Administrations

Working in
collaboration
with Icelandic
Met Met Office
Iceland

ERUPTION
AFBI: Agri-food and Bioscience Institute Central Government
ARSF: Airborne Research and Survey Facility and Departments
BGS: British Geological Survey
CEH: Centre for Ecology and Hydrology Health Agencies
COBR: Cabinet Office Briefing Room
DARD: Department of Agriculture and Rural Development
Defra: Department of Food and Rural Affairs IAVW, Operators and
DfT: Department for Transport Regulators
EMARC: Emergency Monitoring and Response Centre
FAAM: Facility for Airborne Atmospheric Measurement Scientific Agencies
FCO: Foreign and Commonwealth Office
GCSA: Government Chief Science Advisor
HPA: Health Protection Agency
MOD: Ministry of Defence
NACME: National Airspace Crisis Management Executive
NCAS: National Clinical Assessment Service
NERC: National Environmental Research Council
SAGE: Scientific Advisory Group in Emergencies
SEPA: Scottish Environmental Protection Agency

44
35 See Health Protection Agency, ‘Pandemic Influenza Contingency Plan: The HPA’s strategic roles and actions for preparation and response to an influenza
pandemic’, 2009.

www.chathamhouse.org
 Appendix 6: National Approaches to Decision-making during the Ash-cloud Event

for Transport coordinated loosely with the Cabinet Office to disseminate information to other government departments, also briefing ministers and the
public as and when new information became available.
However, it was not until 19 April, around the time when changes in the rules for aviation were being finalized, that the UK government assembled its
crisis response committee (known as COBRA after its normal meeting place in the Cabinet Office Briefing Room) and formalized the roles of two lead
government departments: the Department for Transport, for managing transport contingency measures, and the Foreign and Commonwealth Office
(FCO) for managing the repatriation of UK nationals. The importance of pre-identification of lead government departments is one of the key findings
from the post-crisis UK Parliamentary Inquiry.
The UK Parliamentary Inquiry also concluded more generally that the government’s attitude to scientific advice is that it is something to reach for
once an emergency happens, not a key factor for consideration from the start of the planning process. Despite early meetings between the Chief Scientific
Adviser and the Cabinet Office Civil Contingencies Secretariat (CCS) on 15 April, the full competency of the Scientific Advisory Group for Emergencies
(SAGE) was not activated until after the revised European approach to volcanic ash had been implemented. Additionally, the lack of transparency and
openness in the work of SAGE has attracted criticism from those who see this as restricting the efficacy of its actions.
A number of separate scientific institutions in the UK contributed to scientific assessments and advice during the event. Initially responding to the calls from
the UK Met Office, the National Environmental Research Council (NERC) assisted in pooling data from a limited number of ground-based stations
and provided its own research aircraft to seek atmospheric measurements. NERC also drew upon European research ties, feeding into the ash-forecast
database EUFAR. This has subsequently been developed to provide scientists with a comprehensive research infrastructure that can be called upon in
other atmospheric pollution incidents.
The UK Department for Health and the Health Protection Agency (HPA) took the lead on evaluating potential health risks. HPA press releases were
issued from 15 April onwards and consistently stated that the concentration of particles was likely to be low and unlikely to cause serious harm.36 Some
confusion was caused on 16 April when the World Health Organization (WHO) suggested that families stay indoors as a precaution, but the WHO shifted
towards the HPA line in the following days.37
A separate area of work on environmental hazards was led by the UK Department of Environment, Food and Rural Affairs (Defra). Its analysis also
indicated that the impact on water, food and air quality would be limited, but a review after the event has highlighted weaknesses in communications
between different agencies on the assessment and reporting of environmental hazards during the ash cloud.38 It stressed that the boundaries of respon-
sibility between Defra and the relevant health agencies (including the HPA and the Food Standards Agency) needed to be identified and clear to
both parties. Defra already had a Volcanic Ash Network and was able to draw on a range of scientific agencies such as the British Geological Survey
and the efforts of the UK’s devolved administrations.

Figure A4: German decision-making structure during the ash-cloud crisis

DFD: German Remote Sensing Data Centre Federal Ministry of Transport, Strategic
DFS: German Air Navigation Service Provider Building and Urban
DLR: German Aerospace Centre Development (BMVBS)
DWD: Deutscher Wetterdienst (German Met Service)
EMARC: Emergency Monitoring and Response Centre
LaMP: Laboratoire de Météorologie Physique
LBA: German Civil Aviation Authority Luftfahrt-Bundesamt (LBA)
MWO: Met Watch Office

Deutsche Flugsicherung (DFS)

UK Met Office UK VACC DWD

EMARC

DLR
Met Office Iceland
DFD Operational

MWO (Iceland) Scientific

Community
LaMP
ERUPTION

36 HPA press releases www.hpa.org.uk/NewsCentre/NationalPressReleases/2010PressReleases/.

37 'Volcanic ash health warning issued as travel chaos continues’, The Times, http://www.timesonline.co.uk/tol/news/uk/article7099692.ece.
38 Science Advisory Council to Defra: 17 February 2011. Lessons learned from the volcanic ash incident. SAC (11) 09. http://sac.defra.gov.uk/wp-content/ 45
uploads/2011/04/SAC-11-09-Lessons-learned-volcanic-ash.pdf.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull

Germany
After receiving conformation of the eruption, the London VAAC further disseminated the message to European air navigation service providers
(ANSPs), other European meteorological services and Eurocontrol.
The German ANSP, Deutsche Flugsicherung (DFS), bases its air traffic management decisions on publications by the German meteorological
service Deutscher Wetterdienst (DWD), which in turn uses the information of the Volcanic Ash Advisory Centre (VAAC) in London. According
to the ICAO regulations, the VAAC must send an updated forecast to the national meteorological services every six hours.39
DWD issues its own warnings of significant meteorological activity, termed SIGMETs. Based on these warnings, DFS decides on the appropriate air
traffic management measures in the affected airspace. The decisions of DFS are formed by the national aviation authority Luftfahrt-Bundesamt (LBA),
which serves as an independent regulator for aviation safety.

France
On 14 April 2010, the London VAAC informed the French national meteorological service Météo-France and its national aviation authority, la
Direction Générale de l’Aviation Civile (DGAC), of the approaching volcanic ash-cloud. Météo-France used the Volcanic Ash Advisory issued by
the London VAAC in combination with its own analysis as the basis of its meteorological warning to the French ANSP, la Direction des Services de
la Navigation Aérienne (DSNA). In the interest of maintaining maximum safety for aircraft passengers the DGAC, operating through the DSNA,
implemented the progressive closure of French airspace from north to south, from Thursday 15 April.

Figure A5: French decision-making structure during the ash-cloud crisis

Council of Defence and

Président de la République National Security (CDSN)

Secretariat-General for
Premier Ministère National Defence and
Security (SGDSN)

MEEDDM Ministry of
Environment, Energy,
Sustainable Development Strategic
and Sea
All other government
departments (Health, Inter-Ministerial Crisis Cell
Defence etc.) Centre for Departmental Centre for
Crisis Operational Intelligence
Management and Warning (CMVOA)

Directorate General of Direction générale de la Directorate General for

Civil Aviation (DGAC) prévention des risques Infrastructure, Transport
(DGPR) and the Sea (DGITM)

Direction des Services Fédération Nationale

Scientific IPGP, LSCE, Aéroports Airline Operational
Community de la Navigation des Transports de
LMD, SAFIRE de Paris Operators
Aérienne (DSNA) Voyageurs

UK Met UK
Office VACC Metro-France
EMARC

Met Office
Iceland

MWO
(Iceland) IPGP: Institute of Geophysics of Paris
LMD: Laboratoire de la Météorologie Dynamique
ERUPTION LSCE: Science Laboratory of Climate and the Environment
SAFIRE: French Office of Aircraft Instrumented for Environmental Research

46
39 ICAO (2007); Annex 3 – Meteorological Service for International Air Navigation, Chapters 3, 4, 7 and 9 and Appendices.

www.chathamhouse.org
 Appendix 6: National Approaches to Decision-making during the Ash-cloud Event

The Ministry of Ecology (Ministère de l’Écologie, de l’Énergie, du Développement durable et de la Mer, MEEDDM) took lead agency status during
the crisis, overseeing operation of transport as well as coordination between other government departments. A crisis unit was officially established
within the MEEDDM on Saturday 17 April, chiefly to centralize coordination of key stakeholders. With the assistance of transport operators such as the
National Federation of Passenger Transport (NFTV), Aeroports de Paris (ADP) and the crisis-cell SNCF were able to coordinate the return of passengers
stranded abroad via those airports that remained operational, and also to coordinate management of inland transport of passengers by exceptional
means via the SNCF and bus carriers.
A number of institutions assisted the MEEDDM in the timely provision of scientific advice and input: the Institut de Physique du Globe de Paris
(IPGP), Météo-France, Laboratoire des Sciences du Climat et de l’Environnement (LSCE), the Laboratory of Dynamic Meteorology (LMD) and the engine
manufacturers of Air France. The IPGP was almost immediately called upon to coordinate a crisis-cell, drawing together the relevant scientific expertise.40 The
IPGP also set up a teleconferencing system which was utilized twice daily to exchange information between scientists and the MEEDDM and which also
disseminated daily reports, updating departments on the latest volcanic activity. Additionally, Météo-France worked on simulating the path of the ash cloud,
and organized several test flights – but was limited by the preparedness of research aircraft and their inability to gain permission to fly.

40 Report No. 28 (2010–2011), Jean-Claude Etienne, Senator and Christian Kert, MP, on behalf of the Parliamentary Office for Scientific and Technological 47
Advice, filed 12 October 2010.

www.chathamhouse.org
 Preparing for High-impact, Low-probability Events: Lessons from Eyjafjallajökull
Preparing for High-impact,

Bernice Lee and Felix Preston, with Gemma Green

Low-probability Events
Lessons from Eyjafjallajökull
A Chatham House Report
Bernice Lee and Felix Preston, with Gemma Green

ISBN 9781862032538

Chatham House, 10 St James’s Square, London SW1Y 4LE

T: +44 (0)20 7957 5700 E: [email protected]
F: +44 (0)20 7957 5710 www.chathamhouse.org
www.chathamhouse.org
Charity Registration Number: 208223
9 781862 032538