Economic Resilience

Updated: Dec 1, 2020

Economic resilience is the capacity for an economic organization to recover from, adjust to, or maintain functionality in the face of negative internal or external impacts.[1] The idea of economic and financial resilience has become a hot topic since the recent financial crisis. Research into the network analysis of financial systems has since shown a very high concentration and centralization of resources within a few core financial institutions, but the vulnerabilities go far beyond the financial system.[2] There is a strong correlation between increases in capital flows across borders and economic volatility.[3] As our global economy becomes more integrated with a freer flow of goods, people, capital, and services, small movements within global networks like commodity and money markets can have major implications for national economies. There remains an open question as to how do national economies integrate into these often-volatile global networks while ensuring their resilience? Economic resilience analysis involves trying to identify some of the key factors involved in the resilience of an economic or financial network. In this article, we discuss two very different strategies that can be taken when it comes to dealing with change effectively.


We can define resilience as the capacity of a system to maintain critical structure or functionality in the face of perturbation; more simply it is “the capacity to recover quickly from difficulties or toughness.”[4] This perturbation may be internally or externally derived. It may be small or large. It may be a short shock to the system or a prolonged incremental effect. Economic resiliency would then be the capacity of an economy to stay functioning given some internal or external alteration. These critical functions might include such things as basic supply chain operations like delivering food, but also institutional functions, such as maintaining a stable liquid currency. Perturbations to an economic system might include such events as major internal social instability, external political or military conflict, natural disasters, or major alterations within the supply of critical commodities such as oil.

An economy is vulnerable to both internal and external factors. When we are looking at external factors, we are asking: What is the system dependent upon within its environment? How many different input types does it require, such as social capital, political regulation, natural resources, external financial services etc? The system’s robustness will be contingent upon how many input types there are, how volatile these inputs are, and finally, we also need to ask, what is the critical range of these input values? Is it very narrow, reducing the system’s robustness or very broad increasing the system’s robustness. For example, how much can the price of oil change before it will significantly alter the state of the economy?[5]

Network Theory

An economic system will only have very limited capacity to alter these external factors. Some it will be able to influence, many it will not. No one country can stop a global financial crisis. No one country can stop climate change. What an economy can do though is alter its internal configuration to make it less dependent on these external factors. Internal to the economy, we need to consider both the structure to the network of interrelations and the properties of the components within that system. The structure to the interconnections and interdependencies can be modeled through network theory. The first factor we will want to analyze is the overall density of connections. At a low level of link density shocks will propagate slowly, cause and effect will have a more linear proportionality as limited connectivity reduces the capacity for positive feedback to take hold. The higher the density, the faster something can propagate across the system and the stronger the influence from positive feedback; meaning we can get the butterfly effect of a nonlinear disproportionality between cause and effect. Rapid failure propagation and the butterfly effect are the reality of heightened interconnectivity.

Another key metric to the internal network structure is its degree distribution and centralization. The economies of scale that has been a key driver in the development of industrial economies has created fragile networks that are dependent upon critical centralized nodes. This makes these centralized nodes “too big to fail” as their loss can potentially disintegrate the entire network.6 Just as these local and global hubs facilitate connectivity, they also facilitate failure spreading.

Within the industrial economic model these hierarchical centralized systems are the only real mechanisms we have for any form of advanced specialized organization, and thus whether we are talking about global cities or large corporations these large centralized systems typically correlate with a high level of specialization. Which is again another important parameter to the networks internal resilience. The criticality of a node is not just a function of its size but also how irreplaceable that node is, in more specialized networks where components are monofunctional – meaning we have more irreplaceable nodes – that will lead to a higher degree of overall network criticality. Distributed networks without centralized coordination mechanisms typically have to maintain themselves, components are often multi-functional, they often have a lower level of specialization giving the network a greater capacity to swap out any degraded components, small players can adapt and pick up the slack when the system receives a shock. There are many examples of this from peer-to-peer finance to community alert schemes and mesh networks of all kind.


Next, we will discuss the strategies that a network can adopt in order to achieve robustness or resilience. In terms of resilience strategies that we may adopt, there is fundamentally just two: resistance or adaptation. Resistance is a strategy that involves aligning all elements in the system towards preventing change. Here we are aiming to build up a store of resources, such as liquidity, fixed capital or any form of redundancy, so that we can weather out some perturbation or any failures can be identified at their source and quickly removed by superior force, typically through a centralized regulatory system that is monitoring the whole operation. Walter Bagehot in his book ‘Lombard Street’ put it like this: “In wild periods of alarm, one failure makes many, and the best way to prevent the derivative failure is to arrest the primary failure which causes them.”[7]

This resistance approach is aligned with optimization in that by becoming more efficient we can gain more resources and develop a greater capacity to resist change. Within such a strategy diversity is seen as a weakness. It reduces the system’s capacity to act in a coordinated fashion in removing any possible failures at their source. In such an approach, there is no distinction between the system and its parts. As the saying goes: “What is good for General Motors is good for America.” In such a model, major components such as corporations or financial institutions are the system and we simply can not let them fail.

This strategy of resistance will be most effective within relatively simple environments. The aim is to become the biggest fish in the pond so that nothing will affect you. This will work in both quite simple and complicated environments as long as the whole environment is relatively stable. In order to be able to resist change, the system needs to be in control all the time. In order for a centralized linear regulatory system to be in control it must remove disorder from the system. If it has a linear model of the system and environment it is regulating, it will try to linearize the system in order to have greater control and robustness. To the extent that the control process and the model that guides this process is accurate and reliable, control itself delivers stability, robustness and the capacity to resist change, which is one method of ensuring critical functionality and the system’s preservation.[8]


In more complex environments, where any one component is just a small fish in a big pond, this strategy will have its limitations. In such a context the whole landscape may change and no component within it will be able to resist a systemic shock.[9] Here we are going to need a different strategy, that of adaptation. Adaptation is the capacity of a system to change its state in response to some change within its environment.[10] Resilience through adaptation means being able to generate a variety of states depending on the context. Because the system needs to exhibit different states at different times, there needs to be some differentiation between the system’s core function – that will remain invariant – and the different states that it will generate depending on the context – which will have to change. This is different to our previous model where the system was simply the set of components, in particular, the large dominant nodes that were too big to fail, because they ran the organization. With adaptation in contrary, because the system has to change and adopt new states depending on the context, any component in the system has to be disposable. In this model, the system cannot be dependent upon any instance of its states but is capable of letting them go and generating new ones as the context changes.

As such, this adaptive approach to resilience is concerned with the maintenance of core functionality rather than the preservation of specific system’s components. Unlike in our previous model where we were trying to stamp out any small disturbance at its source, with adaptive resilience we are saying quite the contrary, that the system needs disturbances in order to maintain and develop its adaptive capabilities, which is ultimately the only thing that is going to ensure its long-term survival within a complex environment. Without disturbance, components become fixed into a specific configuration that will reduce the system’s capacity to let them go and generate different configurations as needed to adapt. A corollary to this is that when the system is left static or without significant perturbation for a prolonged period, components will become increasingly tightly coupled in order to reduce friction and cost of transactions. As the network becomes more tightly coupled, the important buffers that mitigate failure propagation are reduced. This is all summarized in Hyman Minsky’s main insight that “stability is destabilizing.”[11]


Because of all of this, the adaptive approach focuses its interventions more on adapting to major disturbances, whilst allowing small and moderate disturbances to run their course. This approach would avoid stamping out the disturbance at its origin and focuses its attention on reducing the wider impact of the disturbance on the macro economy. But equally, the system can only deal with so many shocks of such a large magnitude. If they become too great, they will damage the network’s critical functionality and its capacity for sustainable development.

This is most clearly expressed within ecology by what is called the intermediate disturbance hypothesis, IDH for short. IDH is a non-equilibrium model used to describe the relationship between disturbances and species diversity within an ecosystem where diversity is correlated to the system’s long-term resilience. IDH posits that local species ‘diversity is maximized when ecological disturbance is neither too rare nor too frequent. At high levels of disturbance, due to for example frequent forest fires or human impact, all species are at risk of going extinct. At low levels of disturbance, competitive exclusion increases leading to a decrease in species diversity. Thus, at intermediate levels of disturbance, diversity is maximized.’[12]

Order & Chaos

Adaptive resilience is seen to exist at an interplay between order and chaos, what is called the edge of chaos. Whereas the resistance strategy will work best in orderly and stable environments, adaptive systems are normalized for operating within volatile and semi-chaotic environments, what business management calls VUCA environments, standing for volatility, uncertainty, complexity, and ambiguity.[13] VUCA is a key characteristic of the 21st-century economic environment due to a number of factors, such as technology disruption and environmental change.

An adaptive strategy also comes with the cost of maintaining diversity. There is a core tradeoff between diversity and efficiency. Diversity requires components that occupy heterogeneous states. Efficiency requires synchronization and coordination between components. This is most easily achieved by homogenizing the states of the components. Thus, a resistance strategy will aim to reduce diversity and linearize the system, which may be a good strategy if it works. However, it also makes the system brittle, as we are creating a strong impermeable boundary condition to the system that once broken means the lack of diversity within the system will allow for some failure to easily propagate, making it more susceptible to positive feedback loops and systemic shocks. An adaptive strategy is predicated upon diversity because with diversity components are occupying heterogeneous states, meaning we will get negative feedback as different states balance each other out. This negative feedback makes the system less vulnerable to rapid failure propagation driven by positive feedback.[14]


The effectiveness of diversity as a strategy for resilience and long-term sustainability is not always a function of the sheer scale of diversity – that is to say the number of different states that the components occupy – but also a function of maintaining weak ties and suboptimal marginal components. For example, species within ecosystems that often have only a weak role to play on average, tend to have a strong and important function in maintaining resilience during times of stress. Thus, we need to focus not just on the keystone components which are important during normal functioning, but also on those components at the fringes of the system, and this may be even more the case when these perturbations are large nonlinear black swan events. Adaptive resilient systems then need to maintain what engineers would call degenerate components, components that are lacking some property, order, or distinctness of structure that is usually present. These degenerate components are typically the ones that are first removed when optimizing a system through linearization because they are very clearly suboptimal when the system is operating at its normal equilibrium state, but only become relevant during non-normal states when the system is being tested by a perturbation.


Within complex networks where agents have the capabilities to create connections at their own discretion, we do not always know all of the linkages. If the system is configured in such a fashion that it is in the agent’s interests to conceal their interactions  – as is often the case with financial networks and particularly the case with hedge funds and other forms of shadow banking that are not formally regulated – then this opacity can be a major issue in terms of managing resilience. Complex networks like our global financial system and global supply networks have many interdependencies that we are not aware of. These hidden interdependencies typically only get revealed by shocks and failures. The more opaque the system, the lower our capacity to receive critical information about it, thus debilitating our capacity to use a centralized regulatory system in order to manage it. At some critical level of opacity the system becomes unmanageable through a centralized regulatory mechanism because we can no longer sense it properly and receive the critical information required to manage it in a top-down fashion.

Information is central to enabling distributed adaptive regulation. An information and networked economy is one built on data and open platforms. This means developing technical solutions for scrambling and anonymizing sensitive data into aggregate volumes that can be mined through advanced analytics to identify and preempt critical patterns. This is essentially part of the success of the Internet as an open platform where transparency and sharing of information are an attractor.


  1. (2017). Retrieved 27 May 2017, from

  2. James B. Glattfelder: Who controls the world?. (2017). YouTube. Retrieved 27 May 2017, from

  3. (2017). Retrieved 27 May 2017, from–58-83.pdf

  4. resilience – definition of resilience in English | Oxford Dictionaries. (2017). Oxford Dictionaries | English. Retrieved 27 May 2017, from

  5. (2017). Retrieved 27 May 2017, from

  6. (2017). Retrieved 27 May 2017, from

  7. Wiley: Lombard Street: A Description of the Money Market – Walter Bagehot . (2017). Retrieved 27 May 2017, from

  8. Near-Optimal Resilience through distributed robustness. (2017). Snippets. Retrieved 27 May 2017, from

  9. (2017). Retrieved 27 May 2017, from

  10. (2017). Retrieved 27 May 2017, from

  11. Thomas Tan, C. (2008). Introducing the Minsky Theory – Stability Is Destabilizing. Seeking Alpha. Retrieved 27 May 2017, from

  12. Intermediate Disturbance Hypothesis | Wikiwand. (2017). Wikiwand. Retrieved 27 May 2017, from

  13. (2017). Retrieved 27 May 2017, from

  14. Gerry Marten | Human Ecology – Populations and Feedback Systems. (2017). Retrieved 27 May 2017, from

Systems Innovation

  • LinkedIn
  • YouTube
  • Twitter
  • Facebook