Systems - especially complex ones - do not always act the way that we would intuitively expect them too. A good starting resource to explore complex systems is Jay Forrester's paper titled Learning through System Dynamics as Preparation for the 21st Century. Barry Richmond, one of Jay Forrester’s students and the creator of STELLA®, helps to uncover the characteristics of complex systems in his Systems Thinking Skills.
By applying common characteristics to complex systems, understanding of these systems becomes easier. Learn more about these common characteristics below.
Complex systems are composed of many interacting feedback loops. They often contain long time delays. What may appear to be an obvious reason for a particular problem is often not the fundamental cause of the problem, but only a symptom.
From the time a pollutant enters the environment, it may take decades for any ill effects to be seen. The ill effects can appear in many forms and thousands of miles from the source, making it difficult to trace back to the underlying culprit.
Social and economic policies have effects that reach beyond election cycles. The elected officials in power when a crisis becomes apparent are rarely directly responsible for creating the mess. It may look as though recently-implemented policies have had immediate impact, but the reality is often that the seeds of the problem took root long before.
See examples at Population Dynamics
Complex systems contain balancing feedback loops that surround the various goals of the system. Low-leverage policies often seem to be the “obvious” solutions to the problem at hand, but they encounter resistance – the tendency for interventions to be defeated by the response of the system to the intervention. Low-leverage policies are unable to overpower the balancing loops in order to align the competing goals of the system.
The US has been fighting a “war on drugs” for decades. Low-leverage policies (seizing more drugs, arresting more addicts) are defeated by the compensating feedback that responds to the intervention. When drugs are taken from circulation, prices go up, along with the incentive to risk importing more drugs.
In areas where flooding occurs regularly, people to build dams and levees to reduce the risk of property damage. Relief costs from flooding have been rising for decades, however, even when population increases have been taken into account. The construction of dams and levees has impeded the natural ability of the ecosystems to shed excess water and has also encouraged further development in flood-prone area.
Complex systems contain areas of high leverage – places where a small push in the correct direction is likely to effect the desired change. In many cases, these high-leverage policies are difficult to identify and difficult to apply correctly. The “levers” for such policies may be pushed in the wrong direction, or not pushed at all.
In the case of improving the well-being of urban residents, Jay Forrester, by building and testing a system dynamics model, found that low-income housing is a high-leverage policy. (Urban Dynamics, Jay W. Forrester) Unfortunately, the change needed is to build fewer, not more low-income housing units.
In arid regions around the world, the natural water supply may vary greatly from year to year. Water that is collected in dams must be divided among competing users. In years of plentiful rainfall, pressure mounts for the “extra” water to be distributed to help boost the residents’ well-being. The most effective policy to maintain sufficient distributions through drought years, however, may be to consistently distribute the minimum.
Problems observed in complex systems are almost always internally generated. While it is easier and more comfortable to place blame on others, it is more productive to look within the system itself to understand and change undesirable behavior.
This complex system is commonly identified by the oscillations within the system.
A person may experience burnout repeatedly, all the while thinking that his workload is fluctuating over time and “driving” him to burn out. In reality, it could be that his own desire to achieve ever-higher standards is the real culprit.
Locally rational choices, such as when ranchers try to maximize their own use of a shared resource (a tragedy of the commons situation), can lead to problems for everyone when the resource disappears. If each actor in the system thinks his own behavior is reasonable but everyone else is taking too much, he will never see his own contribution to the problem.
Complex systems tend to drift to lower levels of performance over time. This can occur over a long timeframe, making the downward spiral both insidious and hard to combat. This situation occurs when individuals or institutions respond to failing to reach their goals by adjusting them downward in order to relieve the discomfort of failure.
A student accustomed to achieving the highest scores may accept lower grades as she moves up in grade level and tackles harder material. Instead of challenging herself to maintain high scores, she may set her sights on lower grades that seem more achievable.
A goal to pay off credit card purchases every month (incurring no interest charges) may gradually be replaced by a desire to at least pay more than the minimum charge, until eventually one is satisfied with only paying the minimum payment.
In complex systems there are tradeoffs between short-term and long-term goals. What is achievable or desirable within a short timeframe can reveal problematic consequences in the fullness of time. Conversely, concentrating on a future payoff almost always involves sacrifice in the present.
Indulging a sweet tooth provides immediate pleasure, but may detract from future health, particularly to teeth and waistlines.
Many years of living below one’s financial means can build wealth over time, leading to a higher future standard of living. The opposite situation can be illustrated by using credit cards to live beyond present financial means. The future standard of living is likely to be worse when the debt plus interest must be repaid.
This characteristic is often closely related to the tradeoff between short-term and long-term goals. Both play out over time, but the presence of an intervener usually means that a form of addiction or dependence is at work. The system's natural ability to fend for itself declines over time as the addiction/dependence becomes stronger.
The decision to smoke cigarettes or use illegal drugs for immediate pleasure can quickly give way to addiction; more is needed to achieve the same feeling. The person's ability to function normally without the substance declines as the addiction grows stronger.
Farmers on worn-out or marginal land may apply fertilizer to boost their yields. It may work well for a season or two, but the soil has not been improved for the long term. Without correcting the underlying issue, the application of fertilizer must continue (in even greater amounts) for yields to be sustained.
Road Maps is a self-study guide to learning system dynamics. It is organized as a series of chapters, and was developed by the System Dynamics in Education Project (SDEP) at MIT under the direction of Professor Jay Forrester. Ten chapters of Road Maps are available for download. Road Maps teaches the reader how to identify different kinds of systems all around us and how to model these systems. Road Maps can be a resource for both beginners and advanced system dynamics modelers, and requires no previous system dynamics knowledge and only basic math skills.
Road Maps is available from the System Dynamics Society website. It is free for members and there is a small cost for non-members.