Thinking in Systems

What is a system?

I am glad you ask. The understanding of systems is critical to navigate in a complex world. But let's start by a definition. A system is an integrated assembly of matter and living beings that are interconnected and have one or multiple purposes, goals or functions. 

Our climate is a system, our forest is a system and our economy is a system too. Although the isolated analysis of one part can be helpful for some type of questions, a realistic measure of the impact of a shock/new scenario/variable change, requires the understanding of the relationship between the components of the system.

Please do not forget the purpose/s. It is particularly unnoticed the real goal that each systems follow. Oceans systems are critical for the live of corals and fish, for the moderation of the climate, for the generation of clouds and drinkable water... Forests are critical for the soil, for the availability of oxygen, local rain patterns, the wildlife that inhabit these spaces... 

Knowing the system purpose/s is a fundamental fact to understand its importance. What is the purpose of the economy? Before answering that question, keep in mind that purpose is defined by actions/behavior, not by rhetoric.

Last but not least, let me emphasize this, a system is much more than the sum of its parts. A forest is not a sum of trees, a agriculture field is not a sum of hectares of land, a society is not the sum of its individuals... Keep in mind that we, as all the systems around us, are relationship.

Stocks and Flows

To understand the dynamics of the system, two elements describe the status and behavior of the system in time. 

The stock is based on the accumulation or memory of matter, condition from past flows for the current state, and by future flows to define future realizations. The stock change over time through actions is a flow.

To define the health of a system, or simply to assess an stable amount of a stock, such as drinkable water, money or food...One must understand the input and output flows. This becomes particularly important for complex systems, which are most of the systems we care about.

Consider the example of the economy GDP, which shows the output flow coming from a given stock of materials, energy sources, physical and human capital. Looking only at GDP growth or even the absolute value on a historical basis will be flawed to define the sustainability of the system without considering the available stocks of materials and energy.

The fact that a system, such as the economy, has managed to give a continuous and growing amount of output flows of food, housing, services, transport...says nothing about the likelihood of permanence and stability of the system. One should look at stocks and flows, if is willing to assess the future behaviour of the system.

Humans tend to solve system stock decay but increasing the input flow only. In there is an energy crisis that makes energy more expensive, instead of reducing the demand of energy (output flow) we tend to desperately make our way out using very expensive and inefficient additional sources of energy (increasing the input flow).

Many economists approach poverty, climate change, and inequality in the same way. Instead of reducing the outflow rate, redistribute better, reduce unnecessary expenses and share prosperity with less pressure on Earth, more GDP growth is shown as the inevitable path for poverty eradication, climate change mitigation and inequality moderation.

It is important to note the delaying and smoothing effect on stocks on showing potential disequilibrium of the system. Imagine that we have oil for five years, considering that there are 100 million tones and every year we consumer 20 tones. At a constant extraction rate of 20 tones, the flows can go on till suddenly after year 5 there is no more oil. Economists will likely be right at saying that the price mechanism will ensure that the output flow will go down as prices goes up, due to scarcity. But, can they ensure that the transition will be smooth, and that the market will give price signals with enough time for a transition without oil? Before you read ahead, think of the oil crashes from the 70's and how fast the affect the economy, or simply look at the speed at which alternative energy sources are available for such inevitable transition.

Feedback loops and reinforcement feedbacks

Many systems regulate via feedback loops. Flowers and seasonal trees changes the behaviour based on temperature and rain patterns, investment decisions change based on interest rates and expectations. Systems adjust to such discrepancies from the goal, adjusting one or more of its elements, in order to pursue their objective.

This feedback mechanism does not ensure that the system will operate well in the long run. The information may arrive too late, the capacity of the system to react may be too limited and hence decay becomes inevitable. A fast information loop, significant levers for adjustment, and ideally slow shocks are the best conditions for a successful adjustment. The problem again is that many times, economies,soils, forests, oceans and glaciers cannot react sufficiently fast to the information feedback, since it comes too late. 

This is why many scientists advocate to slow down the economies, so the capacity of our systems to react to energy, climate, food, employment and other shocks will be much more likely. Very few system thinkers will disagree with that.

If information processing and adjustment are limited, they become even more with the existence of reinforcement loops. This loops create even stronger or faster feedbacks on the system making very unlikely adaptation. 

Taking the example of industrial farming, the pursue of lower prices and profitability leads to monoculture and the use of pesticides that erode the soil and its capacity to be resilient to pests or keep up with productivity growth. The farmer start a reinforcement loop of reducing yields and poor soil, with an increase used of chemicals that further damage the soil and condemned the farmer to a collapsed soil, and ultimately debt with the pesticides suppliers. As you may notice, the solution to build resilience and avoid collapse is a diversified agriculture with little or no use of pesticides, taking advantage of natural species that can regulate plagues while keeping the soil rich and diverse. As we will see later, there are attributes such as diversity, that make all systems more resilient, soil health is not an exception.

Another example that make reinforcement loops critical for any model of complex systems is the Co2 emissions and temperature rise. There is almost no one that puts under question such relationship, but the amount of uncertainty and discrepancy of how temperature would evolve considering feedback loop is remarkable. Here is the catch: there is a delayed effect on the long lasting (~100 years) stock of Co2 in the atmosphere and the average temperature rise and variability. This effect is likely going to be greater as warming reduce ice and cloud coverage, releases methane from permafrost melting, which further increases warning in a reinforcement loop. Long story short, an apparent little warming such as 1.5-2 could trigger feedback loops that easily lead to 6 degrees of warming. We simply don't know if live on Earth is possible with this warming, as warming beyond 2 degrees is out of human historical records.

Not sure how do you feel by reinforcement feedbacks, particularly the negative ones could be very scary and have to be avoided by all means.

Properties of healthy systems

After sharing the behaviour of negative reinforcement of feedback loops, let's learn how systems can stay healthy.

A system purpose is to satisfy one or multiple functions, and doing it over time. In order to achieve so, it must be managed not for productivity, profitability only...being the most important attribute its resilience. A resilient system is not an invincible system, as there are limits of how much resilient a system could be to shocks, but the capacity of the system to adjust to feedbacks depends on their resilience.

A resilient system is highly diversified and elegantly embrace complexity. It gather million of self regulated entities, that learn, iterate, diversify to build resilience from the bottom up. Self regulation does not mean that there is not hierarchy, but it is a hierarchy at the service of the units at the bottom of the system, being aligned in functions and goals. The bottom gathers information that is further synthesize to scale thanks to the hierarchical systems.

You can think of trees and cells giving feedback to the system forest, businesses and citizens giving feedback to politicians and decision makers...If the upper layers of the system truly serve the common goals of the bottom, resilience and stability are much more likely. You may wonder that forests exists for million years before human existence, and human organization has created wealth but also is likely to create devastation for our own system. Are the goals and behavior of politicians really in line with the common good? Are we organizing for resilience, or growth and profit?

Modeling Systems

Everything we think we know about the world is a model. Our brains ensemble brilliantly our partial knowledge of the laws and structure of the systems in front of us to give us the impression that we have all figured it out. As it all makes sense in our model, it must be true.

Something more close to the truth is that we do know little and probably not enough of the most important system structures of life. We do not know with certainly the damages that will come with significant global warming, or how are we going to develop in a post fossil fuels civilization without collapse...

The complexity of such events could give the impression that trying to build models for this questions is too pretentious and a rather useless pursue, being the contrary true, and here is why.

The best tool we have to manage systems towards resilience and harmony is the study, observation and modeling on their hypothetical structure. Given the fact that most, if not all structures are nonlinear in their connections, we need systems modeling to be able to simulate and test scenarios that are likely, with the most accurate system structure we can come up to.

The greater the amount of uncertainty we have in those structures, the more cautious our assessment should be. We must consider likely scenarios whether they show exciting or stressful futures. It is not very likely that absolute decoupling from material and fossil fuels took place, so our models should consider such scenario too and not permanent technological progress as it is assumed in most economic models. It is very likely that the damages for our life supporting systems after 1.5 degrees of warming are much worse that a linear extrapolation of the current damages, due to the forces explained before. Be aware that there is an inherent bias in the scenarios and the likelihood presented on those, and we should be sceptic towards those economists who never present scenarios without exponential growth, or that do not give adequate risk measures of the future with business as usual trajectories. Not to mention those who discount our grandchildren wellbeing to almost zero.

Given that the systems interact with each other, it is very important to leverage tools such as Integrated Assessment Models, or Computable General Equilibrium models to have an estimate, as accurate as possible, of the breakdown effects of a change in one system towards the others.

In other words, use partial equilibrium models wisely, or you will, more likely than not, miss the nonlinearities between systems structures.

A word of caution must be placed, as many models assume perfect rationality of agents, or perfect information, making the trajectories path of this models appear very smooth, and they hardly showed tipping points or collapses. This is a fault in design, and whenever possible, we must consider the limited capacity of agents to adapt to shocks, and the delays in responses from the system, which could lead to far from optimal transition paths.

Creating Change in a systems world

After some deep research on the system/s of interest, we may realized that the likely trajectory, or the current state of stocks and flows are sub-optimal for the goal at hand. The next question to ask is: How do we create change in a system?

Change the goal, design toward the goal

We need first to think of the goal/s of the system, to really evaluate whether the goal is aligned with what we think is better for the system to become resilient. 

Taking again the example of the soil. As long as the only goal is short term productivity and profit, the soil will erode due to the lack of recovery, diversity and time required for regeneration, even in the unlikely case that all chemicals used are removed. The very first step is to ensure that soil health and longevity is a critical goal from which any profitability goal cannot be decoupled. Our food security depends on that.

The same apply to the economy. If the economy is growing by design, any forced reduction in GDP growth or total value, will lead to crisis and collapse. If we instead design the economy for resilience, and not growth, it is much more likely that sustainable levels of throughput and moderate footprint can be achieved without a traumatic transition.

Keep in mind that for the system to be resilient, diversity and self organization is critical. This is why many movements towards resilience want to create stronger democratic structures and empower local governments for effective policy design. It is simply in line with good system design.

Leveraging the important inputs

If the goal of the system is aligned with its components, then our understanding of the system structure should lead us to the most important inputs, and within those, to the scarce ones.

Take our development as an example. Our current level of complexity is only possible with cheap energy, material abundance and specialisation. If we are not willing to go back to the caverns, nor lead to more resource wars, we should consider adjusting the layers and flows, such as:

• Complexity: a solution could be to focus less on superfluous industrial products and more on service intensive sectors: child and elder care, culture, sports...
• Energy: reducing the demand of energy: insulation, efficiency regulations, sharing services, public transport, local mobility...
•  Food: going from a cheap industrial model with eroding soils and poor resilience to a diversified and soil caring practice
• Specialization: Incentivize research and careers that make the system more resilient such as preventive medicine, organic farming, agroforestry, clean energy engineering...

This can be directed in multiple ways. While education and social norms take time and may not reach required levels of change, regulations and tax reform can rapidly ensure high levels of adherence to the desired goals. Remember, our system required resilience, not necessarily more complexity.

It is too frequent that in the medicine, politics and even in sport fields, that actions are taken as a reaction of a strong feedback, and in the worse cases in an advanced stage of a reinforcement loop such as cancer, bankrupt or debt, or permanent injury. The most effective, cheap and painless treatment in almost all systems is the preventive one. We should use our knowledge of the structure of systems such as our bodies and minds, our economies and our athletes, to apply interventions that avoid negative loops to escalate against the main goal. Remember, the goal of those systems cannot only be performance, social status, or growth...resilience have to be at the core of preventive interventions.

Conclusion

This text would no be possible with a wonderful mind such as \citep{Meadows2001}. Although she will not be able to read that modest summary of her research, I hope I am making a contribution to expand the knowledge and recognition that system thinking have in creating a better world.

System thinkers embraces the complexity of the world, creating humble yet rigorous models that tries to explain systems structures, while adding the most important stocks and flows.

We know that feedback loops are normally late, and the reaction to them are limited, particularly when there is reinforcement loop in place. It is important to ensure that the system is design for resilience, and that information can catch up with a flow rate that do not compromise the future achievement of the meaningful goals of the system.

Resilience is to diversification, what good governance is to self regulation. A bottom up flow of information, with a hierarchy that services the resilience of the system maximizing entities well being, is going to lead to good outcomes, in the absence of big negative shocks.

The complexity and nonlineratity of the systems is not a reason to avoid its modeling, and making mathematical or intellectual constructions to make the most sensible interventions to our system goals. While every model is wrong, models that are stock-flow consistent and can capture the most relevant structures from our reality can be very useful.

By all means, to ensure the health of our systems, we must understand if the goal is linked to the observed behavior. The behavior of the system must be in consonance with the goals it claims to serve.

Most of our challenges and painful situations can be better deal with a preventive application of system thinking. If the current state is negative but there is chance for recovery, on top of the right interventions we must allow the time adjustment requires.

We do not know enough of complex systems such as the economy, the climate,the soil, the carbon cycle...but we know enough to ask for clear adjustment on its goals, and apply preventive interventions for a resilient planet, society and economy.

 

References:

Donella Meadows 2008, Thinking in Systems. Chelsea Green