Incorporating Shock Events into Aviation Demand Forecasting and Airport Planning (2024)
Chapter: 6 Defining and Identifying Shock Events
CHAPTER 6
Defining and Identifying Shock Events
Defining Shock Events
Much of the literature in this area uses a variety of terminology and definitions regarding shock events. However, in general, there is broad agreement that shock events are those whose likelihood, timing, and magnitude are hard to predict or even anticipate but which when they occur, impact business in significant or profound ways (positive or negative). These have also been referred to as “high-impact, low-probability events” (Lee, Preston, and Green 2012), “unknown unknowns” (a term popularized by former Secretary of Defense Donald Rumsfeld during a press conference in 2002), and “black swans” (Taleb 2007). Taleb codified this conceptualization of a black swan as an event or world state that cannot be (or is not) imagined and therefore is not considered even as a possibility. Feduzi and Runde (2014) contrast this with known unknowns, which is an uncertainty that has some real possibility of occurring. They provide an important caveat that “an event experienced as a black swan by one person may not come even as a mild surprise to the next person” based largely on the information available to an individual (Feduzi and Runde 2014, 274).
There is broad agreement that shock events are those whose likelihood, timing, and magnitude are hard to predict or even anticipate but which, when they occur, impact business in significant or profound ways (positive or negative).
In fact, Taleb (2007) argues that often we are dealing with grey swans—events for which there is little information, but at some level, we can anticipate or be aware of them. Taleb has argued that COVID-19 was not a black swan event and was entirely predictable (Avishai 2020).
As an example of the difference in information, and by some measure imagination, in whether an event that unfolds is a black or a grey swan (a known unknown), Feduzi and Runde (2014) discuss the 9/11 terrorism event. They argue that for a great many people, this was a black swan surprise event, but to North American Aerospace Defense Command (NORAD) this was a known unknown as they had simulated the potential impacts of an attack like 9/11 on targets including the World Trade Center and the Pentagon. This is an important distinction between what may be considered a shock or surprise event, particularly deeming something an unknown unknown, as the ability to contemplate the possibility of uncertain events can transform it into a known unknown. The director of the National Commission on the Terrorist Attacks upon the United States (Kean and Hamilton 2004) called the misreading of known terrorist group activities (previous hijackings, flying lessons, etc.), “a failure of imagination.”
Other authors have argued that true black swans may not exist in the sense that the market or scenario in which a surprise event occurs may be created by the setting in which it occurred. Walter (2020) posits that financial black swans, such as market crashes or recessions, are “illusory” as they are a construction of the regulations of the market themselves, among other factors. Walter (2020) goes on to suggest that how the financial sector measures and quantifies risk, with regulations intended to minimize that risk, creates opportunities for shock events but that
those should not be considered as black swan events as they arise from the design of the market itself. This distinction is useful as it reminds the reader that shock events can be from known or imagined sources, and it is through the deeper examination of risk and risk events can the unknown unknowns be revealed as potential events.
Wucker (2016) takes this one step further, referring to “grey rhinos”—high-impact but high-probability events that we are blinded to or actively choose to ignore. Wucker suggests that the 2007–2008 Global Financial Crisis was one such example, where there were plenty of warnings (and plenty of people warning) that the financial system was at high risk, but policymakers chose not to address them until the crisis hit (the rhino charged, in Wucker’s analogy). Wucker argues that some shock events are missed or downplayed not because of a lack of information but due to an unwillingness to consider uncomfortable outcomes.
Aggarwal and Bohinc (2012) suggest that the underlying frequency and impacts of shock events are increasing over time due to such factors as the internet and information technology, globalization, and climate. The scale and scope of shock events can be linked to how increasingly interconnected the global society and the global economy have become. With less interconnection, the spread or speed of the consequences of a shock event may be less and more local. Aggarwal and Bohinc propose looking at indirect as well as direct impacts from black swan events and as an example, put forth terrorism-related supply chain disruptions. Therefore, individuals and organizations dealing with risk must project trends beyond simply looking at historical data and how the future developments of major trends will affect the probability and magnitude of future potential shock events. A document by Munich Re (2018), a reinsurance company, discusses trends in shock events, particularly how the frequency and impact of natural catastrophes appear to be on the rise due to the impacts of climate change. Through analysis of loss statistics, this insurer has “demonstrated the plausibility of climate change already influencing some types of events in certain regions” (Munich Re 2018, 23) in its work to reduce the potential risk loss from catastrophic events by better informing their actuarial models about future risk potential. In its continued research, Munich Re has looked at how the probability of catastrophic natural events, as well as the number of claims and value of damage, has grown as climate change progresses and included trend projections in its risk analysis and management.
Shock events may be becoming more frequent due to such factors as the internet and information technology, globalization, and climate change.
Risk researchers and actuaries are always trying to move events that initially seemed like shocks or even black swan events into standard risk calculations. Once a shock or surprise event occurs, ex post evaluation of its impacts and causes can allow what may have been previously considered an unknown unknown to become a known unknown. It is argued in the literature that greater awareness of risks and uncertainties could improve an individual’s or organization’s ability to postulate new risks and shocks and potentially convert what would otherwise be considered a black swan surprise event into a known uncertainty, even where information is limited.
Identifying Shock Events
There is an abundance of empirical research that indicates humans are too often surprised. Studies have shown that when asked to estimate the 10th and 90th percentile ranges (or some other extreme) of some distribution, the true range turns out to be considerably outside of the estimated range (Morgan, Henrion, and Small 1990). Individuals and organizations tend to err on the side of overconfidence, particularly when projecting future activities. This has been explored in the realm of airport traffic forecasting
Individuals and organizations tend to err on the side of overconfidence, particularly when projecting future activities.
in the United States (Suh and Ryerson 2019) from a variety of causes, including the under-representation or appreciation for the potential of significant contractions in aviation traffic. An optimism bias, or overconfidence in the ability to estimate probabilities, will tend to lead to more positive outlooks and generally downplay the probability or impacts of negative events.
The gambler’s fallacy, also known as the Monte Carlo fallacy, is another psychological phenomenon that creates practical challenges when attempting to model and predict shock events. In this psychological phenomenon, people may be led to believe that if a certain event or outcome occurs more frequently than in the past, then that outcome is less likely to occur in the future, despite that the outcome of the event is not conditional on previous outcomes (i.e., that the probability of any given outcome of the event is statistically independent). Psychologists have analyzed this phenomenon, which frequently uses gambling as an example, using games involving independent probabilities like dice rolls or spins of a roulette wheel, and have extended the examination of the phenomenon to behavior such as observing an athlete’s performance and predicting future outcomes based on whether they have happened recently or not. As a form of cognitive bias, the gambler’s fallacy can be difficult to overcome, especially when evaluating the likelihood of a future event or outcome when it is not fully understood whether or not the event’s probability is independent or conditional on past outcomes or some other force.
Methods for eliciting or uncovering the potential risks that seek to avoid or minimize human biases are raised in much of the literature on risk analysis and strategic planning under uncertainty. ACRP Report 76 summarizes several approaches—such as Delphi, statistical groups, nominal groups, and unstructured interacting groups—for eliciting such information from experts and stakeholders (Kincaid et al. 2012). The Delphi method, also known as estimate-talk-estimate technique, is one of the more commonly used methods for forecasting future outcomes and identifying emerging trends. Developed by the RAND Corporation in the 1950s, the Delphi method is a systematic and qualitative method of forecasting by collecting opinions from a group of experts through several rounds of questions. It consists of providing a questionnaire to relevant selected experts to give their opinion on the topic at hand (in this case, potential shock events that could impact an airport and the likely results of those shock events). After the experts answer the questionnaires, the facilitator collects all the answers and hands out a summary report of the answers to each expert. The experts then review the summary report and either agree or disagree with the other experts’ answers. Another questionnaire is completed that allows the experts to update their opinions and challenge the opinions of others. The Delphi method becomes complete when a consensus of forecasts is achieved via this iterative process.
Goodwin and Wright (2010) argue that Delphi and similar methods are less effective in identifying shock or rare events, as they remain heavily influenced by past experience and cognitive biases (availability and linearity biases). These methods rely on consensus views, while the authors suggest in the field of rare events, more weight should be given to minority opinions that challenge the consensus. They argue for the use of scenario planning that is less focused on predicting the future and allows for less mainstream but still uses plausible outlooks for the future to be considered and played out. The authors acknowledge that even these methods are problematic for aiding the anticipation of rare events but may aid in the development of protective strategies. The authors propose the use of “devil’s advocacy and dialectical inquiry,” whereby the Delphi approach is modified to include groups who are challenged with offering contradictory views, and there is less focus on consensus.
In a similar vein, Bonabeau (2008) proposes the use of “augmented paranoia.” Writing in the context of predicting terrorism and related activities, he argues for a mix of methodologies: outreach using nontraditional experts from other fields, additive aggregation; exploiting the “wisdom of crowds” and self-organization; and working outside of the organizational hierarchy
to avoid groupthink and thought hijacking. [“Wisdom of crowds” is the idea that large groups of people are collectively smarter and better informed than individual experts (Surowiecki 2004).] Walker et al. (2019) propose a number of methodologies for decision-making under deep uncertainty that do not focus on predicting the future but rather on considering the range of possible futures and ensuring that decision-making is robust enough to support that range of possible futures.
Identifying shock events requires being open minded and a creative viewpoint to considering the risks (and shock events) facing an organization, including uncomfortable and extreme scenarios.
The goal of all these approaches is to ensure an open-minded and creative viewpoint when considering risks (and shock events) facing an organization, including uncomfortable and extreme scenarios. Paté-Cornell (2012) argues that many previous supposed black swans were no such thing but rather failures of analysis and imagination, citing examples such as the Fukushima tsunami and nuclear accident, the Columbia space shuttle explosion in 2003, and the 9/11 terrorist attack.
Risk Identification in Aviation
For most airports, air traffic forecasts play a key role in airport planning and development. The standard master plan approach can be characterized as follows:
- Determination of the forecast, and
- Selection of a plan that best suits this forecast.
Even broader airport strategic planning is guided by air traffic forecasts, although in this case, it can be argued that the relationship is more two-way as the airport strategic plan can impact future traffic through marketing and business planning.
Therefore, air traffic forecasts can play a key role in the identification of risk factors. As documented in ACRP Synthesis 2: Airport Aviation Activity Forecasting (Spitz and Golaszewski 2007) and master planning guides such as the Airport Development Reference Manual, edition 12 (IATA 2022), forecasting generally involves the identification of key drivers or factors affecting traffic development, such as economic growth, fuel prices, airline behavior, and government policy. Uncertainty about the future development of these factors can be considered by changing their assumed values in the forecasting process to produce low- and high-scenario forecasts, as documented later. Arguably, this remains the most common approach to addressing uncertainty in air traffic forecasts. However, this approach does not formalize the identification of risk and may miss out-of-model factors such as shock events.
ACRP Report 76 (Kincaid et al. 2012) proposed the use of a risk register, whereby information is elicited from airport management and other stakeholders on the risks and uncertainty facing the airport. Using a combination of judgment-based and data-based methodologies, risks (including shock events) are identified and evaluated in terms of their likelihood (or probability) of occurring and the anticipated magnitude of impacts (positive or negative). This information is cataloged in a risk register that can feed into subsequent tasks (such as forecasting) and ideally be updated regularly. These risks can be represented in a graphical form, as shown in Figure 20. Note that the register includes a shock events category with “Pandemic” indicated as one of the risks. The probability and scale of impact are relatively low, likely reflecting experience with past pandemics such as SARS in 2013, which had minimal and short-lived impacts on U.S. airports.
Examples of this approach to strategic risks and shock events in the aviation industry appear fairly uncommon, certainly in the public domain. One example is the “Opportunities and Risks Report” produced by Munich Airport as part of its Integrated Annual Reports (covering finance, environment, social, employees, and infrastructure) (Munich Airport 2018, 2019). The risk
assessment covers various categories of risk: force majeure (shock events such as terrorist attacks, pandemics, and natural disasters), market risk (carrier loss and economic cycle), operating risks, legal risks, and financial risks. Figure 21 shows the risk assessments done in 2018 and 2019. The 2019 assessment was started in late 2019 and completed by April 2020 and so it incorporates the emerging COVID-19 pandemic. As a result, “market slump from epidemics/illness” changes from a medium likelihood/medium risk factor in 2018 to a very high-impact/high-likelihood in 2019.
For each risk factor, the report indicates the countermeasures the airport has in place and evaluates how these countermeasures change the risk outlook facing the airport (the “net risk”), with some risk factors being removed from the matrix. In 2018, the countermeasure for “market slump from epidemic/illness” indicated that “Due to a relatively high fixed cost ratio, Munich Airport’s ability to react to market downturns is limited.”
The focus of this research has been on strategic risks and shock events rather than those related to safety, security, or operations. Airports and the aviation sector more generally have sophisticated and well-established approaches to safety and operations, as documented in ACRP Report 131: A Guidebook for Safety Risk Management for Airports (Neubauer, Fleet, and Ayres 2015) and ACRP Synthesis 71: Airport Safety Risk Management Panel Activities and Outcomes (Ayres and Parra 2016).
Figure 21. Overview of gross risks at Munich Airport, 2018 and 2019 (Munich Airport 2018, 2019).
Similarly, there are risk management processes for capital and maintenance projects as documented in ACRP Report 116: Guidebook for Successfully Assessing and Managing Risks for Airport Capital and Maintenance Projects (Price 2014).
Many U.S. airports, particularly large airports, do have dedicated risk managers and departments and have adapted enterprise risk management (ERM) practices. Airports Council International-North America has a Risk Management Committee with representatives from a large number of airports. The focus of this work is fairly broad, ranging from loss control and prevention, airport insurance agreements, regulatory requirements, and various safety issues to more strategic aspects related to the air market and the economy. For example, ACRP Synthesis 115 (Murphy et al. 2021) documents emergency plan practices at airports, including a case study of Charlotte Douglas International Airport (CLT) that developed a Communicable Disease Response Plan in 2018 through planning meetings with various stakeholders, including the Centers for Disease Control and Prevention (CDC). The result was a plan that could be activated or triggered by sick passengers at CLT who meet specific characteristics. The aim of the plan was containment of a communicable disease rather than the wider implications of a pandemic outbreak. Some of the airports interviewed for Part I indicated that they had developed business contingency plans or emergency plans to address how the airport would continue to operate given potential disruptive situations. While these plans did not consider a pandemic specifically, some did consider declines in traffic volume. However, in the context of long-term forecasting and planning, the evaluation and incorporation of shock events are generally not addressed in a fairly formalized manner.
Airports typically have well-established procedures to identify and devise mitigation strategies for operational risks. Airport operations manuals will typically include response strategies and mitigation activities for operational risks such as fires, aircraft accidents, security incidents, power losses, and so forth. ACRP has published two guidebooks on this topic including ACRP Report 131 (Neubauer, Fleet, and Ayres 2015) on risk management for airports’ safety management systems, and ACRP Report 74 (Marsh Risk Consulting 2012) on the topic of applying ERM at airports. Similarly, there are risk management processes for capital and maintenance projects as documented in ACRP Report 116 (Price 2014). These guidebooks provide valuable information to airports on how to identify and manage their risks, though they do not directly address the issue of accounting for risks associated with shock events. While ACRP Report 74 (Marsh Risk Consulting 2012) does provide a framework for identifying and managing both operational (e.g., fires, security, accident response) and strategic risks (e.g., loss of a carrier, recession, challenges to workforce) its focus is on how to start and integrate an ERM process into an airport’s business activities but not, unsurprisingly, how to specifically integrate shock events into processes like forecasting or planning. It has also been argued that ERM generally focuses on compliance and known or familiar risks and tends to ignore or give little weight to rare or shock events (Le Merle 2011).
Airports typically have well-established procedures to identify and devise mitigation strategy for operational risks.
