Crises, resilience, and complex systems
In this article for the National Preparedness Commission, systems thinker and business strategist Miguel Pantaleon, Ideas Infinitas, explores how systems theory can help us understand the characteristics of complex systems and their behaviour to discern what is important and what is urgent. By breaking through complexity to reveal the way our systems are built and connected, can we design systems to be better able to respond, adapt and recover?
Covid-19 has caused over two hundred thousand deaths in the UK since March 2020. The pandemic left a deficit of £312 billion in the government budget. The end of the Brexit transition period in January 2021 means UK businesses are prevented from freely accessing the EU single market, reducing export revenues and labour force, and disrupting the supply chain. In 2022, Russia invaded Ukraine, displacing more than 8 million people, and exacerbating the cost of living in the UK. Kwasi Kwarteng’s September 2022 mini budget forced the Bank of England to spend £65b to avoid pension funds collapsing, and the extreme summer drought in Europe imposed severe water restrictions. The Silicon Valley Bank collapsed in March 2023 shaking the global financial system.
In the last three years, we have suffered at least seven major crises, all different in nature. Some of them are political, others economic, violent conflicts, epidemics, etc. Crises are disturbances that a complex system cannot absorb and respond appropriately to. Therefore, it is necessary to understand how our complex systems interact to improve our resilience.
Complex systems are formed by a large number of elements that are interdependent and interact, producing non-linear behaviour. These constant interactions and interdependence make them unpredictable. For example, our financial system is formed by institutions, banks, regulators, businesses, stock markets, traders, IT systems, etc. that are interdependent and interact producing growth, inequality, stagflation, and financial cracks simultaneously for different groups of people.
Over the years, we have developed an analytical mindset and a wide range of tools to make sense of this complexity, giving us the false impression that the more data we accumulate the more accurately we can predict and control the future. This idea conditions our approach to managing crises and their intrinsic risks. We calculate risk based on how likely an event is to happen and what its impact could be. However, occurrence and impact are not linear phenomena, but emergent consequences of the constant interaction between the elements in a system. This makes it unpredictable when a key event will unfold, as well as its intensity or its duration. Additionally, this constant interdependence can make effects appear far away in time and space, making it extremely difficult to assess real impact.
From this perspective, how risky a crisis can be for us does not seem to be a good way of deciding what is important and what is urgent for us. To respond better and for design systems to be able to respond, adapt and recover, we need to break through complexity to reveal the way our systems are built and connected.
These interactions make complex systems unique. As Edward Lorenz discovered, minimal differences in the initial condition of a complex system can produce massive differences in its later stages, popularly known as the “Butterfly Effect”. This uniqueness makes “best practices” ill-equipped to respond to crises that may look alike but require completely different approaches and strategies to cope with them.
Analytical mindset and command and control
Is there a better way of managing risk and crises?
We can improve how we manage risks and crises by challenging our analytical mental model and command-and-control mindset. The analysis focuses on breaking the whole into its constituent parts to understand them. Once we know how the parts work, we put them back together to understand how the whole works. This way of thinking allowed scientific and technical development. However, it is not very effective when the parts of the whole can make their own decisions when interacting together, as human beings do.
Russell Ackoff said: “The whole is more than the sum of its parts; it is the product of its interactions”. Therefore, relationships and interactions make the systems behave the way it does, and this behaviour can have exponential consequences.
We cannot control complex systems. Control requires predictability, it means to know exactly what the outcome is when you intervene in a system. In the case of social systems, people have the autonomy to decide what to do, from actively collaborating in the pursuit of a goal, to violently resisting change. People’s autonomy to decide their goals and their means adds an extra layer of complexity to control. There are two usual ineffective strategies to deal with this extra complexity. These strategies are ineffective because they may provide some internal stability to make decisions and allocate resources but do not respond to the real needs of those involved.
The first strategy is to limit the options available at the input level. For example, reducing the options available in the 2016 referendum to exit the EU to only two options, regardless of how complex the decision was and the multi-generational impact of its consequences. The second is to limit the options available at the output. As Henry Ford claimed: “You can choose any colour as long as it is black”.
The army is usually an example of a complex system governed by a clear and strict chain of command. Rules, orders, processes, protocols, punishment, signs, and rituals are constantly reinforced to make soldiers behave like a unified machine when it is required. This can only be achieved by reducing the autonomy of the soldiers at a minimum, and optimising communication with a centralised control system to make information and orders flow correctly. However, this design has multiple vulnerabilities. For example, the centralisation of power can challenge troops’ effectiveness if the communication channels are interrupted, or the centralised authority is compromised. This control by limiting the options available brings the army close to being a complicated system where a different set of principles apply.
Dave Snowden introduced the Cynefin framework to help us better manage complexity. For me, the most interesting part is the difference he makes between complicated and complex problems. Complicated problems are those that have a large number of interdependent elements but that have one known solution. A plane is a classic example of this. It has, engines, fuselage, hydraulic, navigation, electric, and many others, but they have a specific configuration in which they need to be assembled to work. On the other hand, complex problems do not have a specific solution, and we need to act in different parts of the system to make it develop in the direction we want.
Misleading concept of risk
What are the implications of looking at crises as complex systems for managing risks and improving our response to them?
As mentioned above, crises in complex systems are unpredictable. This means that we cannot know the probability of occurrence or the impact a crisis would have. This makes the current risk assessment based on these two factors ineffective in understanding what is important or urgent to prioritise our response. On the other hand, artificially reducing the complexity of the system to transform it into a complicated system can have severe consequences.
We were not able to predict the Russian invasion of Ukraine, nor anticipate the impact the war will have on the UK economy and society in the long term. For example, the war in Ukraine increased the price of energy in the UK by 130%, rising food inflation by 16.7%. This had a direct impact on people’s disposable income, especially on low-income families, increasing the number of those depending on food banks to cover their basic needs. Public workers initiated a series of industrial actions to demand an increase in their salaries to cope with the increasing cost of living. To manage inflation, the Bank of England increased interest rates to 4.25%, which increased the number of households struggling with mortgage payments.
All sectors of the economy have been affected, and all layers of society are impacted, especially those with fewer resources. However, the real consequences on public services may appear in ten or fifteen years, when a significant proportion of the population becomes vulnerable, local areas become deprived, or people get discouraged to be teachers, doctors, nurses or firefighters, which will deteriorate the public sector even more. If we leave these dynamics to dominate us, our future resilience will be much more difficult.
Most of the impacts that crises have is conditioned by the vulnerabilities we create in our systems. Systems, where power and resources are too centralised, have difficulties addressing local problems effectively. On the other hand, systems, where power and resources are too decentralised, have problems coordinating and developing nation-scale capabilities. Finding the balance between the two of them is key to improving our resilience.
Our healthcare system is not designed to keep us healthy, but to keep us under treatment. This design creates an incentive for the pharmaceuticals industry to improve their current drugs just residually to maximise their profit, rather than researching more creative and effective treatments. During the Covid-19 pandemic, the UK government allocated significant funding to the pharmaceuticals industry to develop a vaccine in record time in a breathtaking collaborative effort. This national emergency probably justifies the investment. However, national food systems are systematically making people develop high-risk illnesses like diabetes, cardiovascular diseases, and obesity. The quick fix for these problems is medical treatment, while the fundamental solution is to transform the food system and people’s eating habits.
Complex systems principles to better understand crises.
We cannot know what the next threat will be, but systems theory can help us understand the characteristics of complex systems and their behaviour. Some of the greatest systems-thinking experts have summarised their learning into rules, principles and aphorisms that can help us to better interact with them.
The structure of the system conditions its behaviour: the structure of the system is the network of relationships and interactions among the elements of that system. These interactions allow the flow of information, resources, people, etc. through it. Systems are governed by their feedback loop structures. There are two types: reinforcing feedback loops, which are present where a system grows or decays exponentially (like in the financial bubble and burst or the spread of a pandemic); and balancing feedback loops, which are present when a system is stable or stagnated. The combination of both can explain more complex behaviour such as shape growth, oscillations, escalation, etc.
Silicon Valley Bank collapsed in March 2023, spreading losses and fear of a new financial meltdown in the global financial market. One of the key factors that enable this kind of behaviour is the capacity of banks to lend money relying just on a fraction of its value to cover a possible withdrawal. This measure paves the path for speculation, interlinking banks in chains of borrowing and lending. This structure is responsible for the economy’s regular bubble-and-burst cyclical dynamic. In this example, the best way of improving our resilience to an economic collapse would not be to improve the bank’s stress test to assess their capacity to respond to an adverse situation but to change the basic rules that allow speculation and encourage interdependence between banks.
Today’s problems come from yesterday’s solutions: when a crisis ends, it usually results in a set of new policies and strategies that will help prevent future crises of the same nature. Generally, these solutions take time and effort to implement and, once they are in play, they improve our capacity to respond and recover from past crises; but not necessarily from future ones. In this situation, we are affected by path dependency, which means that our capacity and options are limited by the decisions made in the past. These solutions are now a part of our system, regardless of their effectiveness, reinforcing our vulnerabilities and making our systems prone to suffer new crises.
Improving parts of the system in isolation will not improve the system as a whole: we design systems to effectively respond to crises. However, these systems will be as effective as their weakest part. For example, when designing an emergency service to respond to life-threatening situations, we can bring together police and firefighter services, hospitals, ambulance services, local councils, etc. and provide them with the necessary resources to perform appropriately. However, improving each element of the system separately will not improve the effectiveness of the service if the communication and coordination between them does not improve. Without seamless coordination and communication, capacity and resources can be wasted and our response to the crises is counterproductive. It is necessary to understand the system and its interaction holistically.
Complex systems have emergent properties that the parts of the systems do not separately have: the financial bubble and burst dynamic is an emergent property of the financial system, not a property of any individual separate part. If we take the money apart it does not produce financial bubbles, nor does the interest rate in isolation or the corporation market capitalisation. It is the interaction between them that produces this emergent behaviour. However, sometimes emergent properties become complex problems themselves. The financial crash of 1929 brought an abrupt end to the prosperity of the 1920s. The economic instability of the following years led to political instability which favoured the emergence of dictatorial and totalitarian regimes around the world, creating the context that ended in the second world war.
Improving resilience by developing our systemic capacity
Looking at crises as complex systems that we are a part of has important implications for the way we design and articulate our responses. The first one is:
We need an appropriate way of understanding interdependence and complexity: a crisis affects different parts of the system in different ways and with different levels of severity. We need to bring all these perspectives together to capture the real nature of the threat. In a complex system, causes and effects are not closely related in time and space. This means that the way we respond to a crisis today can be one of the causes of the next crisis. For example, the Covid-19 pandemic had a direct impact on people’s physical health and the economy. However, it may have mental health consequences on children, young adults and the elderly. These consequences have been buried in favour of the need to recover our normal lives and restart the economy to confront the massive national debt. However, this can be one of the underlying factors to unleash the next national crisis. Therefore, we can find clues to the next crisis in the way we respond to the current one.
To be better prepared, we need to reveal the feedback loops that make a crisis behave in the way it does. To do this, we need to bring a wide range of perspectives around the system, not only from experts and policymakers, but from people that can be affected by the problem today and in the future.
We need a way of collaborating and coordinating our strategies and actions: complex systems tend to self-organise with no central authority. For example, during the pandemic lockdowns, community groups emerged spontaneously to help vulnerable, self-isolating people that do their shopping, take them to the GP or just provide them with some company and emotional support. These community groups were critical for the well-being of completely unattended people. Local councils made efforts to keep these community groups alive when lockdowns ended, but some disappeared once lockdowns ended.
Improving our resilience needs ways to coordinate the centralised responses to the decentralised community-based organisations and people in general, providing them with the knowledge, information and resources needed to respond to the crisis at a more granular level. In April 2023, the government introduced a mobile phone alert system to inform the population of life-threatening situations, which created a unidirectional informative channel, but additional work will be needed to make this channel bidirectional and coordinate actions effectively.
We need to improve our resilience system as a whole: we have a governance system able to mobilise resources and laws to support the bank system in a financial meltdown. However, we do not count on a robust financial support system to cover people’s income when they are unemployed or have a health condition that prevents them from working, as many of the EU’s country members do. This structure, combined with low salaries and the high cost of living, make people vulnerable. Families, nurses, and teachers among others have to use bank food to cover their basic needs. Salaries, cost of living, or inflation are regular disturbances that our complex system should be able to absorb and respond to. The actual existence of a food bank is proof of the failure of our basic systems, that force people into poverty. Without improving people’s capacity to respond to their more basic problems, we will not be able to improve our resilience.
I propose to look at crises as disturbances that interact with our complex systems. This perspective gives us agency and allows us to focus on how we design and coordinate systems that can interact with the disturbances as they unfold. Complex adaptive systems are unpredictable, and we cannot forecast how they will behave in the future. However, we can anticipate their dynamics and take advantage of their structural characteristics to better understand their inherent complexity.
Approaching crises from the perspective of risk leads us to design fragmented solutions that reinforce our vulnerabilities. Similarly, designing policies and strategies for responding to the last emergency leads us to be better prepared for the past crisis, but not necessarily for the next one.
In my opinion, improving people’s awareness of interdependence and feedback is critically important if we want to design systems that adapt better to uncertainty and changes in the environment.
A Cybernetics law says: “For a system to be viable all its subsystems have to be viable too”. Viability here means the capability of independent existence by adapting effectively to the changes in the environment. To me, viability and resilience are very closely related, a resilient system must be viable. If we want to develop a resilient system at a national scale, we need to develop it at a local level, community level, family level and individual level.
- Putnam, R. (2001). Bowling alone. Simon & Schuster.
- Re-Creating the Corporation by Russell Ackoff, June 18, 1999, Oxford University Press
- Gharajedaghi, J. (1999) Systems Thinking—Managing Chaos and Complexity. Butterworth Heinmann.
- Ashby, W. R. 1958, Requisite Variety, and its implications for the control of complex systems, Cybernetica (Namur) Vol. 1, No. 2, 1958.
- Kadushin, C. (2012) Understanding Social Networks: Theories, Concepts and Findings. Oxford University Press, New York.
- Perrow, C. (1984) Normal Accidents: Living with High-Risk Technologies. Basic Books, New York.
- The Next Catastrophe Reducing Our Vulnerabilities to Natural, Industrial, and Terrorist Disasters. Charles Perrow.
- Upheaval: How Nations Cope with Crisis and Change 01 by Diamond, Jared
- Acemoglu, Daron, and James A Robinson. 2012. Why Nations Fail: The Origins of Power, Prosperity and Poverty.
- P. (2008). The Fractal Organization: Creating sustainable organizations with the Viable System Model.
- Castells, Manuel. (2000). The rise of the network society. Oxford ; Malden, MA : Blackwell Publishers
- Senge, P. (2006). The Fifth Discipline: The art and practice of the learning organization. Random House Books.