Blog 14. Review of Thinking in Systems by Donella H. Meadows


The book’s origin.

In previous blogs, I’ve noted that economy, culture, and political governance are complex systems, displaying known system behaviors. Donella Meadows’ book, Thinking in Systems (Chelsea Green Publishing, 2008),  presents crucial insights regarding the complex systems inherent in society. Meadows was a lead author of the 1972 volume Limits to Growth. She drafted Thinking in Systems about 1992. The book was published posthumously with editing by her colleague, Diana Wright. I’ll outline some of Wright’s and Meadows’ thoughts here, including a few terms and examples of my own.


Wright’s introduction.

In her introduction, Wright notes that understanding the behavior of systems may be our best hope for making needed changes, because the trends of the day that frustrate us are often symptoms of underlying system structure. We should make changes to the rules and communications that underlie that structure, rather than attacking a particular symptom directly. In my own analogy, if you attack poverty only with welfare, you fail to end poverty and you get more demands for welfare. You need to change the system that generates poverty.

Definition of system.

Meadows defines a system as having elements (called agents or actors by other authors), interconnections, and a coherent organization that achieves something. For example, your digestive system has many elements, from throat to intestines, with chemical and nervous signals to communicate the information. In contrast, rocks scattered on the road are not a system. They have no connections. It’s easier to learn about a system’s elements than its interconnections, its flows of information.

 System behavior.

Meadows’ book doesn’t clearly identify emergence as a system property. However, she presents the concept that system behavior is a feature of the entire system. It is not the behavior of an individual actor. That’s the difference between a hive and one bee. The strength of the book is its identification of the way systems fail, and the methods for altering system behavior.


Meadows says the basic operating unit of a system is the feedback loop. That’s where an action generates a signal that may be interpreted, modified, acted upon, and changed by a sequence of the actions of other actors. Some message gets back to the first actor, perhaps after a delay. If the feedback is positive, that is, if it rewards the initial action, then the initial action is repeated. It grows. If the feedback is negative, the action is discouraged. Systems in balance have positive and negative feedbacks, so the system behavior may vary, but remain in bounds. In my analogy, the market price of putty may vary up and down, but not go to such an extreme that the resources are exhausted or that putty production stops. Meadows underscores something known to every physicist and engineer: A system with an unchecked positive feedback will ultimately self-destruct. But Meadows is not talking about electric circuits or machines; she’s applying the logic to social systems. Indeed, the collapse of entire societies has since been documented by Jared Diamond (Collapse, Viking, 2005).

System failures and fixes.

Meadows identifies undesirable system behaviors and the ways attempted fixes fail.

The system may be resistant to changes in policy, as when enforcement agencies try to limit the supply of illegal drugs. In response, the suppliers increase their efforts to meet the unchanged demand of the addicts. The solution is to give up ineffective policies, as was done during prohibition of alcohol.

A situation like the tragedy of the commons occurs when there is no feedback or long-delayed feedback from the public to a business that consumes a non-renewable resource. Examples are destruction of land by overgrazing, mining a mineral (e.g. oil), or dumping wastes in rivers and the atmosphere. A business can increase its profits by extracting the resource or by dumping waste on the public. That’s an immediate positive feedback, profit for dumping. It is the absence of timely negative feedback that causes the long-term public loss.

The goal for a system (that is, its desired state), may become degraded if we perceive the actual state as the desired state. When the actual state degrades, our expectation degrades, resulting in a continued degradation of the goal. This makes me think of money, for which we want a constant value. But, after each year’s inflation, we adjust the acceptable value to be the current value, and inflation proceeds—providing a loss to those who save cash. There’s a negative feedback, but politically we refuse to get the message.

A system may escalate its misbehavior when the goal of each actor is to get ahead of the other actors. Each actor sets a goal based on the perceived state of other actors. Escalation occurs in advertising, negative campaigning, or when hospitals compete in providing expensive facilities, costing health care customers more because facilities are not fully utilized.

A special case of escalation occurs when winners of a competition receive the ability to compete more effectively in the future. That’s a positive feedback. The winners get ahead, the losers lose more, and the system acts to eliminate competition. That happens in politics, in the marketplace, and in funding for schools. The poorest children receive the worst education.

Sometimes, a system is based on the wrong goal. In politics and governmental actions, the GNP (gross national product) is used to guide decisions. In effect, GNP is a measure of the total money flow, not a measure of satisfaction. If I clean your house and you clean mine, and we pay each other, the GNP increases but the real result is the same as when we cleaned our own houses.  Meadows points out that if the goal of society is GNP, the society will do its best to produce GNP, not necessarily producing welfare, equity, justice, or efficiency. Meadows suggests real wealth with low resource consumption, health, and a clean environment should be the goals. Meadows saw the painful results of systemic faults. In this writing, she appears as a loving analyst, neither an idealist nor an activist.

Beliefs and paradigms.

Meadows suggests that almost every culture has a belief in its superiority over other cultures. To me, that’s south versus north within the U.S., or west versus middle-east internationally. Beliefs are paradigms by which we set goals. Meadows suggests individual paradigms can be changed in an instant, offering leverage points for adjusting systems.

Because Meadows gives insightful methods for altering system behaviors without rhetoric or rancor. I wish our leaders would read her book. However, it’s up to the rest of us to recognize the behavior of systems, and to alter the goals, rules, and information flows within the systems that frustrate us.