Systems Thinking: From narrow goals to stewardship

Let's inquire a little deeper into our dominant paradigm of progress and problem solving.

Countries measure and direct their progress or level of development using a handful of metrics, the dominant ones by far being economic, such as GDP and GNI. In business, the measure of success is profit. In a scientific career, it's the number of papers and citations. In school, it's grades with respect to some defined universal standard. When a problem arises, the typical approach within this paradigm is trying to fix the metrics that are "problematic". Low grades -> improve them. Falling profits -> boost sales and cut costs. Global warming -> energy efficiency. Headache -> painkiller.

This linear and narrow-focused approach is reliable and works very well in straightforward, simple scenarios where the system in question is more or less mechanical, well understood and not dependent on many external factors. But are the aforementioned cases like this?

A person is a living complex open system in a dynamic interaction and interdependence with the surrounding environment, including other living beings and abstract structures like the economy, culture, ideologies, collective psyche, etc. The economy is a complex open system of interactions between billions of individuals and organizations, in a dynamic interplay with culture, society, politics, other species, ecosystems, biosphere, etc. We live in a world of interconnected and interdependent complex systems of which there's a lot we still don't even understand. Do the same approaches work with such systems?

The consequences of narrow goals in a complex world

In simple systems, the relationship between cause and effect or inputs and outputs can be easily defined and managed. When operating a machine - say, a kitchen oven - we know exactly what its function and the relationship between inputs and outputs are. We can set a clear goal - e.g. bake a cake - and, even though it may take a bit of experimentation to get it right, together with the process of preparing it, we'll eventually come up with a recipe that we can reliably replicate over and over and it will work every time. Regulation of the process is easy - if we see that the result is an under-baked cake, we try raising the temperature or increasing the time. The process is also contained, without much interaction with the rest of the world (not really, as the ingredients and electricity have to come from somewhere, but for the purpose of the example, we can assume those are given and there is not much outside the kitchen this process is interacting with).

In complex dynamic systems, such as most systems we live with, this is not the case. We call them complex not just because we can never fully understand them, but because they are dynamically interconnected through a number of feedback loops, which produces emergent behaviours that in turn keep changing the system itself. The system is thus in a state of constant evolution and "dance" with often inherently unpredictable outcomes, which means a static set of metrics and goals is not appropriate to be able to work with it. Not only this - such systems are open. They work because of constant fluxes of energy, matter and information between them and their surroundings (i.e. other systems). There are systems within systems within systems...

To see what happens when we use narrow goal-oriented thinking in complex systems, we will look at the economy. The economy is a complex system, living inside many other complex systems - the biosphere, communities, countries, etc... However, economics as a discipline looking to not only understand, but direct the economy, is mainly concerned with building mathematical models as if it were a mechanical system. One unit - money - is used to express more or less all value in the system, through things like GDP, profits, price of goods, costs, etc. Production functions connect clearly defined inputs (labor, capital, resources) to outputs (products). It's all so elegant and easy to define, measure and put on a chart that this became the dominant way we think about progress. Economic metrics such as GDP, productivity, profits and wealth became almost the sole measures of success for countries, firms and individuals. Not surprisingly, maximizing economic growth became the #1 goal of countries worldwide. Not only this - we built all sort of other systems around this metric - banking, pension funds, governments, public funding, etc. - all of it now depends on perpetual economic growth.

The system looks very nice in a textbook or in an upward-sloping chart on a boardroom presentation. And, to be fair, economic growth did bring a lot of benefits, as we all of course know. There are genuine cases where pursuing growth of economic output is a very reasonable goal - such as when a society is struggling to meet basic needs.

Here's the problem with this narrow-minded economic thinking though.

There's nothing in the economic equations that reflects the nature of the systems of which economy is a part. Resources like lumber, animals, minerals, fuels, people etc. are treated as quantities going into the equation without any regard for the roles they play outside the economy. A forest has no value for economics until it's cut down and sold. A person is worth nothing more than the labor they provide or the purchases they make. Animals aren't parts of living ecosystems - they're production factors. A piece of land is worth nothing more than the resources it contains and/or its potential for extracting economic value (like housing, infrastructure or agriculture), no matter if it's important for biodiversity, animal transits, common spaces, etc.

The economy is a part of and depends on all these other systems - biosphere and societies with all their subsystems - but these systems aren't accounted for anywhere in the measures of economic success or in economic equations. Because of this, economic success is free to happen at the expense of externalizing all the harm to them. We can destroy ecosystems, exploit human psychology to sell them more stuff, pollute, launch products causing massive health issues, exploit people wherever the laws allow it, steal land from indigenous communities that treat is as a part of commons, privatize the most basic resources such as water... We can inflate wars because selling weapons boosts economic growth, we can create addictions because it will make people buy more stuff, and we can enslave communities by taking away their self-sufficiency and making them pay for basic survival. All of those things break down social and environmental systems on a planetary scale, but we call it progress because economics only looks at economic output. In fact, such explosive economic growth could not have happened without externalizing all those harms. If we were to account for these externalities, most industry would be unprofitable. Thus, economic growth relies on almost literally eating through environmental and social systems and since the 1970s, we have started seeing the global economy bump against limits of planetary capacity. By continuing to pursue such a goal, we're eroding the very substrate upon which economy and society depends and ultimately the economy itself will collapse and/or lose any kind of useful function to humans (for example, becoming a bunch of AI systems trading with each other, begging the question what would even be the use of that).

Reading this, one might ask: What if we tried to include all those things in the economic equation - estimate the financial damage of, say erasing a hectare of rainforest, and include it in the cost of lumber or food or whatever the land is then used for. As mentioned in the previous paragraph, doing this for all externalities would collapse the economy, but let's forget about that for the moment and think if this would bring us any closer to an economic framework that appreciates the complex systems it's a part of. Many such attempts have been done - such as carbon pricing, ecosystem services valuation, or, one of the more perverse ones, value of a statistical life. While these can have some benefits in incentivizing policy and purchasing decisions toward protecting these systems, they still exhibit the same narrow economic paradigm. Once you assign monetary value to everything, everything becomes tradable. But the main feature of complex systems is diversity and interactions between the parts which you can't describe with any static set of parameters, let alone one single parameter. How can you possible compare value of a forest and value of a human life? Or value of mental health and value of carbon dioxide in the air? All of those play critical roles in life, but cannot possibly be compared along the same axis. If we do that, we've again collapsed all the complexity into one metric to be optimized and inevitably we will start making tradeoffs and sacrificing the integrity of the supporting systems in pursuit of the goal.

We can learn two things from this:

• Optimizing for one metric and pursuing a constant goal (such as economic growth) in a complex dynamic system is bound to create externalities that impact the entire system and eventually lead to collapse (which is the ultimate regulatory feedback mechanism that "resets" the system after strain has become too strong).

• Parts of the system and their interactions are non-tradable and trying to express them along the same axis will lead to thinking in terms of tradeoffs, again leading to a linear optimization approach of one metric, feeding externalities and collapse.

We're now in a situation where countries are pushing with all their might to pursue growth and competitiveness, in order to solve problems that growth caused in the first place. The role model for healthy economic development could instead be the development of a person; a person grows physically until reaching body maturity, and then growth stops. This doesn't mean the person is no longer developing, but development takes place along completely different dimensions. Physical growth is no longer appropriate and would actually damage the person's capabilities if it continued - the body would get fragile and eventually couldn't support itself any longer. Development instead continues in the direction of becoming more wise (in reality, often not), contributing to society, new ideas, raising children, etc. The economy is not much different - by continuing to pursue growth even when it is no longer appropriate, the whole society and its environment is becoming more fragile, begging for collapse. Why not stop growth and start developing more wisdom and contribution to the world instead?

Thinking in systems: a different paradigm

Learning collectively to think and act in systems is a good step towards developing this wisdom. This intuition is not new at all. Some of the millenia-old traditions, practices, philosophies and bodies of knowledge. such as Buddhism or Vedas, embody these principles - just not in the scientific, mechanic language we're used to. Only in the 20th century have such concepts started to emerge also in the western scientific arena, mainly as a consequence of new discoveries in fields such as physics, neuroscience and biology, and trying to model the increasingly complex world and its problems, such as wars and global supply chains, using the newly available power of computers. The work of pioneers in many disciplines ranging from physics and computer science to ecology and biology, such as Humberto Maturana, Francisco Varela, Donella Meadows, Heinz von Foerster and many more, started forming an interdisciplinary field of systems thinking, cybernetics and complexity science.

There are many useful resources we encourage the readers to look at, such as Donella Meadows project (and especially her piece on leverage points - places in the system where a small change can have huge impacts) or The Systems Thinker (one of the best resources to our knowledge for learning about the frameworks and tools of Systems Thinking - you will learn all about feedback loops, delays, common systemic patterns, etc.). Because these excellent resources already exist, we will avoid going into detail here. What we will do is mention a few basic differences between the linear optimization paradigm and the holistic systems paradigm.

1. Thinking about relationships, not parts. All of our brain functions, cognitive abilities, memory, etc. are determined not by neurons themselves, but by connections between them. In the same way, the behaviour of complex systems results from the interconnectedness of parts and structure much more than the parts themselves. Changing a player on a football team probably won't significantly change the game, while changing a strategy or rules of the game will. Changing a CEO of a corporation is very unlikely to lead to any significant changes, while changing the structure or economic incentives will. A single tree or animal in an ecosystem doesn't change much, but changing the way species interact through the food network could produce an entirely different ecosystem. We're so focused on parts - my goals as an individual, our goals as a company - that we neglect the fact that it's the structures defining our interactions that really matter. We try to fix or replace parts when we have a problem, while we should be changing the structures. That's why solutions to social and environmental problems caused by the economy don't lie in more efficient companies or different politicians, but in changing the structure of the economy itself - different legal entities, different financial system, different ownership models and different economics fundamentals.

2. Feedback loops, not cause - effect. We're used to thinking in terms of cause-efect relationships. But in complex systems, these relationships are not linear. Causal relationships form numerous feedback loops where A influences B, but B in turn influences A (directly or through various other parts of the system). These feedback loops can be very tricky, because in many of them, the feedback gets back with a delay, making regulation difficult. Addictive behaviour is one such example. It need not be substance abuse - addiction in terms of systemic structure is any dynamic where we repeatedly use a solution that instantly improves our symptoms, but the symptoms keep coming back and the solution has destructive consequences down the line, at which point it makes abandoning it very difficult (read about Systems Archetypes, specifically "Shifting the burden"). Say you're parenting and trying to teach a child to clean after herself when done eating. It's easier to just clean up yourself because when she does it, she will not do it well at first and might make an even bigger mess and it will take more of your time. So if you keep doing these tasks instead of her because it's easier than teaching her to do it, she will grow more and more incompetent and you will just end up with more burden having to do everything yourself and dealing with a fussy kid who is not used to doing anything. By reducing your burden short-term, you increased your burden long-term (in addition to raising a less competent child). That's delayed feedback.

3. Dynamic equilibrium, not static goal-seeking. As iterated many times before, static goal-seeking is one of the things to be abandoned. But this doesn't mean goals and optimization have no place. They are obviously very useful for individual tasks or certain periods of time where we need to coordinate action to achieve certain changes. But the point is they need to be continually reevaluated and adapted to the dynamics of the system. We need to be able to abandon a goal when we see it no longer serves the system. An obese person, for example, may get very motivated by setting monthly weight loss goals - but if they continue pursuing the same goals after they have achieved a healthy state, they will hurt themselves. The problems arise when we set narrow goals as the main drivers of systems, thus sacrificing resilience. This is the situation with economic growth - we tied so much of our society to the necessity for continuous growth that it's now very difficult to abandon this goal without collapsing the entire economy. Long-term successful systems are neither perfectly optimal nor completely versatile - they are in a dynamic equilibrium where both - efficiency and resilience - are balanced. The rule of T is a useful guideline.

4. Seeing patterns of behaviour and structures, not individual events. When thinking in systems, we see events only as tips of the iceberg. Surely sometimes these events also need our attention - there may be something that urgently needs to be fixed to avoid further damage or other unwanted consequences. But more often, we're interested in looking below the surface and seeing what these events are the symptoms of. In the economy, for example, a systemic focus would not be on an individual problem - like inequality, AI risk, CO2 emissions, waste, etc. - but on seeing how the economic structure is causing all these problems to arise. This makes us ask completely different (and better) questions, like: How does the financial economy incentivize environmental and social externalities and how can we change the incentives? How do the competitive dynamics produce races to the bottom like the AI race?

5. Fixing root causes, not managing symptoms. This is a natural consequence of point 4 above. The events we observe are usually just symptoms caused by systemic structures and dynamics. We can treat the symptoms (sometimes we have to, if they're acute), but if the underlying structure generating those symptoms stays the same, they will reappear. All the problems mentioned in the previous point are consequences of the structure of the economy, which is in turn a consequence of deeply embedded power games. Coming up with less carbon intensive technologies or doing philanthropy to redistribute some money from the rich to the poor is nice, but these are all symptomatic solutions that can't (and evidently don't) have any lasting effects. Inequality and externalities are structurally embedded into the economy. Economic growth depends on capitalizing more of the commons (natural resources, information, land, etc.) and dispersing waste outputs back into the commons (atmosphere, waters, broader society, unregulated countries...), while the financial structure (money breeds more money) inevitably causes inequality gaps to expand. Only by changing the financial backbone to not run on debt anymore and by restructuring the economy and society to remove the perpetual growth obligation could these problems really be solved.

6. Continuously questioning mental models, not constructing absolutist narratives. Our mental models - the views, interpretations and assumptions we hold about the world - determine our space of possibilities and generate the systems we build. Every single big breakthrough, be it scientific or cultural, came with changing a mental model to allow seeing a different possibility. That's why absolutist narratives are so dangerous - we think we know how things are and at that time we stop questioning their validity and close our minds for thinking of better ways. That's why thinking in systems and trying to model systems is not a process of describing how the system is, but a process of deeper inquiry into how systems behave and what hidden assumptions we have about them that may or may not be appropriate.