Book Notes: Thinking In Systems

Thinking In Systems

Thinking In Systems analyzes the structures and behaviours of systems.

Donella Meadows was an environmental scientist, teacher, and writer. She primarly focused on systems thinking and how that applied to the environment. This type of thinking can also be applied to many other domains.

The beefy part of the book lays out a number of systems traps and analyzes the ways to get out of them.


Page 1 · Location 189

The answer clearly lies within the Slinky itself. The hands that manipulate it suppress or release some behavior that is latent within the structure of the spring. That is a central insight of systems theory. Once we see the relationship between structure and behavior, we can begin to understand how systems work, what makes them produce poor results, and how to shift them into better behavior patterns.

Page 5 · Location 262

I start with the basics: the definition of a system and a dissection of its parts ( in a reductionist, unholistic way ). Then I put the parts back together to show how they interconnect to make the basic operating unit of a system: the feedback loop.

Page 5 · Location 264

Next I will introduce you to a systems zoo — a collection of some common and interesting types of systems. You’ll see how a few of these creatures behave and why and where they can be found. You’ll recognize them ; they’re all around you and even within you.

Page 5 · Location 266

With a few of the zoo “ animals ” — a set of specific examples — as a foundation, I’ll step back and talk about how and why systems work so beautifully and the reasons why they so often surprise and confound us.

Page 11 · Location 327

system * is an interconnected set of elements that is coherently organized in a way that achieves something. If you look at that definition closely for a minute, you can see that a system must consist of three kinds of things: elements, interconnections, and a function or purpose.

Page 12 · Location 346

A system is more than the sum of its parts. It may exhibit adaptive, dynamic, goal - seeking, self - preserving, and sometimes evolutionary behavior.

Page 14 · Location 384

Many of the interconnections in systems operate through the flow of information. Information holds systems together and plays a great role in determining how they operate.

Page 14 · Location 399

Purposes are deduced from behavior, not from rhetoric or stated goals.

Page 16 · Location 422

Keeping sub - purposes and overall system purposes in harmony is an essential function of successful systems.

Page 17 · Location 443

A change in purpose changes a system profoundly, even if every element and interconnection remains the same.

Page 18 · Location 465

A stock is the memory of the history of changing flows within the system.

Page 18 · Location 466

Stocks change over time through the actions of a flow. Flows are filling and draining, births and deaths, purchases and sales, growth and decay, deposits and withdrawals, successes and failures. A stock, then, is the present memory of the history of changing flows within the system.

Page 22 · Location 531

A stock can be increased by decreasing its outflow rate as well as by increasing its inflow rate. There’s more than one way to fill a bathtub!

Page 23 · Location 542

Stocks generally change slowly, even when the flows into or out of them change suddenly. Therefore, stocks act as delays or buffers or shock absorbers in systems.

Page 24 · Location 562

Stocks allow inflows and outflows to be decoupled and to be independent and temporarily out of balance with each other.

Page 27 · Location 613

A feedback loop is a closed chain of causal connections from a stock, through a set of decisions or rules or physical laws or actions that are dependent on the level of the stock, and back again through a flow to change the stock.

Page 28 · Location 639

This kind of stabilizing, goal - seeking, regulating loop is called a balancing feedback loop, so I put a B inside the loop in the diagram. Balancing feedback loops are goal - seeking or stability - seeking. Each tries to keep a stock at a given value or within a range of values. A balancing feedback loop opposes whatever direction of change is imposed on the system. If you push a stock too far up, a balancing loop will try to pull it back down. If you shove it too far down, a balancing loop will try to bring it back up.

Page 31 · Location 677

It is called a reinforcing feedback loop, and will be noted with an R in the diagrams. It generates more input to a stock the more that is already there ( and less input the less that is already there ). A reinforcing feedback loop enhances whatever direction of change is imposed on it.

Page 32 · Location 703

Reinforcing feedback loops are self - enhancing, leading to exponential growth or to runaway collapses over time. They are found whenever a stock has the capacity to reinforce or reproduce itself.

Page 39 · Location 805

The information delivered by a feedback loop — even nonphysical feedback — can only affect future behavior; it can’t deliver a signal fast enough to correct behavior that drove the current feedback. Even nonphysical information takes time to feedback into the system.

Page 40 · Location 820

A stock - maintaining balancing feedback loop must have its goal set appropriately to compensate for draining or inflowing processes that affect that stock. Otherwise, the feedback process will fall short of or exceed the target for the stock.

Page 48 · Location 924


  1. Are the driving factors likely to unfold this way?
  2. If they did, would the system react this way?
  3. What is driving the driving factors?

Page 50 · Location 977

The central question of economic development is how to keep the reinforcing loop of capital accumulation from growing more slowly than the reinforcing loop of population growth — so that people are getting richer instead of poorer.

Page 55 · Location 1040

A delay in a balancing feedback loop makes a system likely to oscillate.

Page 57 · Location 1069

Delays are pervasive in systems, and they are strong determinants of behavior. Changing the length of a delay may ( or may not, depending on the type of delay and the relative lengths of other delays ) make a large change in the behavior of a system.

Page 57 · Location 1076

We can’t begin to understand the dynamic behavior of systems unless we know where and how long the delays are. And we are aware that some delays can be powerful policy levers. Lengthening or shortening them can produce major changes in the behavior of systems.

Page 59 · Location 1113

In physical, exponentially growing systems, there must be at least one reinforcing loop driving the growth and at least one balancing loop constraining the growth, because no physical system can grow forever in a finite environment.

Page 71 · Location 1267

Nonrenewable resources are stock - limited. The entire stock is available at once, and can be extracted at any rate ( limited mainly by extraction capital ). But since the stock is not renewed, the faster the extraction rate, the shorter the lifetime of the resource.

Page 71 · Location 1270

Renewable resources are flowlimited. They can support extraction or harvest indefinitely, but only at a finite flow rate equal to their regeneration rate. If they are extracted faster than they regenerate, they may eventually be driven below a critical threshold and become, for all practical purposes, nonrenewable.

Page 75 · Location 1313

Consider the properties of highly functional systems — machines or human communities or ecosystems — which are familiar to you. Chances are good that you may have observed one of three characteristics: resilience, self - organization, or hierarchy.

Page 76 · Location 1323

Resilience is a measure of a system’s ability to survive and persist within a variable environment. The opposite of resilience is brittleness or rigidity.

Page 76 · Location 1328

Even higher meta - meta - resilience comes from feedback loops that can learn, create, design, and evolve ever more complex restorative structures. Systems that can do this are self - organizing,

Page 81 · Location 1430

Systems often have the property of self - organization — the ability to structure themselves, to create new structure, to learn, diversify, and complexify. Even complex forms of self - organization may arise from relatively simple organizing rules — or may not.

Page 85 · Location 1500

When a subsystem’s goals dominate at the expense of the total system’s goals, the resulting behavior is called suboptimization.

Page 85 · Location 1512

Hierarchical systems evolve from the bottom up. The purpose of the upper layers of the hierarchy is to serve the purposes of the lower layers.

Page 86 · Location 1525

  1. Everything we think we know about the world is a model.
  2. Our models usually have a strong congruence with the world.
  3. However, and conversely, our models fall far short of representing the world fully. That is why we make mistakes and why we are regularly surprised.

Page 87 · Location 1538

Everything we think we know about the world is a model. Our models do have a strong congruence with the world. Our models fall far short of representing the real world fully.

Page 89 · Location 1583

System structure is the source of system behavior. System behavior reveals itself as a series of events over time.

Page 90 · Location 1596

These behavior - based models are more useful than event - based ones, but they still have fundamental problems. First, they typically overemphasize system flows and underemphasize stocks.

Page 90 · Location 1601

Second, and more seriously, in trying to find statistical links that relate flows to each other, econometricians are searching for something that does not exist. There’s no reason to expect any flow to bear a stable relationship to any other flow. Flows go up and down, on and off, in all sorts of combinations, in response to stocks, not to other flows.

Page 91 · Location 1625

A linear relationship between two elements in a system can be drawn on a graph with a straight line. It’s a relationship with constant proportions.

Page 91 · Location 1628

A nonlinear relationship is one in which the cause does not produce a proportional effect. The relationship between cause and effect can only be drawn with curves or wiggles, not with a straight line.

Page 106 · Location 1920

Bounded rationality means that people make quite reasonable decisions based on the information they have. But they don’t have perfect information, especially about more distant parts of the system. Fishermen don’t know how many fish there are, much less how many fish will be caught by other fishermen that same day.

Page 107 · Location 1940

Economic theory as derived from Adam Smith assumes first that homo economicus acts with perfect optimality on complete information, and second that when many of the species homo economicus do that, their actions add up to the best possible outcome for everybody.

Page 111 · Location 2014

But some systems are more than surprising. They are perverse. These are the systems that are structured in ways that produce truly problematic behavior; they cause us great trouble. There are many forms of systems trouble, some of them unique, but many strikingly common. We call the system structures that produce such common patterns of problematic behavior archetypes. Some of the behaviors these archetypes manifest are addiction, drift to low performance, and escalation.

Page 116 · Location 2101


When various actors try to pull a system stock toward various goals, the result can be policy resistance. Any new policy, especially if it’s effective, just pulls the stock farther from the goals of other actors and produces additional resistance, with a result that no one likes, but that everyone expends considerable effort in maintaining.


Let go. Bring in all the actors and use the energy formerly expended on resistance to seek out mutually satisfactory ways for all goals to be realized — or redefinitions of larger and more important goals that everyone can pull toward together.

Page 119 · Location 2161

There are three ways to avoid the tragedy of the commons.

Page 121 · Location 2202


When there is a commonly shared resource, every user benefits directly from its use, but shares the costs of its abuse with everyone else. Therefore, there is very weak feedback from the condition of the resource to the decisions of the resource users. The consequence is overuse of the resource, eroding it until it becomes unavailable to anyone.


Educate and exhort the users, so they understand the consequences of abusing the resource. And also restore or strengthen the missing feedback link, either by privatizing the resource so each user feels the direct consequences of its abuse or ( since many resources cannot be privatized ) by regulating the access of all users to the resource.

Page 123 · Location 2249


Allowing performance standards to be influenced by past performance, especially if there is a negative bias in perceiving past performance, sets up a reinforcing feedback loop of eroding goals that sets a system drifting toward low performance.


Keep performance standards absolute. Even better, let standards be enhanced by the best actual performances instead of being discouraged by the worst. Use the same structure to set up a drift toward high performance!

Page 126 · Location 2302


When the state of one stock is determined by trying to surpass the state of another stock — and vice versa — then there is a reinforcing feedback loop carrying the system into an arms race, a wealth race, a smear campaign, escalating loudness, escalating violence. The escalation is exponential and can lead to extremes surprisingly quickly. If nothing is done, the spiral will be stopped by someone’s collapse — because exponential growth cannot go on forever.


The best way out of this trap is to avoid getting in it. If caught in an escalating system, one can refuse to compete ( unilaterally disarm ), thereby interrupting the reinforcing loop. Or one can negotiate a new system with balancing loops to control the escalation.

Page 130 · Location 2382


If the winners of a competition are systematically rewarded with the means to win again, a reinforcing feedback loop is created by which, if it is allowed to proceed uninhibited, the winners eventually take all, while the losers are eliminated.


Diversification, which allows those who are losing the competition to get out of that game and start another one; strict limitation on the fraction of the pie any one winner may win (antitrust laws); policies that level the playing field, removing some of the advantage of the strongest players or increasing the advantage of the weakest; policies that devise rewards for success that do not bias the next round of competition.

Page 135 · Location 2475


Shifting the burden, dependence, and addiction arise when a solution to a systemic problem reduces (or disguises) the symptoms, but does nothing to solve the underlying problem. Whether it is a substance that dulls one’s perception or a policy that hides the underlying trouble, the drug of choice interferes with the actions that could solve the real problem. If the intervention designed to correct the problem causes the self - maintaining capacity of the original system to atrophy or erode, then a destructive reinforcing feedback loop is set in motion. The system deteriorates; more and more of the solution is then required. The system will become more and more dependent on the intervention and less and less able to maintain its own desired state.


Again, the best way out of this trap is to avoid getting in. Beware of symptom - relieving or signal - denying policies or practices that don’t really address the problem. Take the focus off short - term relief and put it on long term restructuring.

Page 137 · Location 2524


Rules to govern a system can lead to rule beating — perverse behavior that gives the appearance of obeying the rules or achieving the goals, but that actually distorts the system.


Design, or redesign, rules to release creativity not in the direction of beating the rules, but in the direction of achieving the purpose of the rules.

Page 139 · Location 2566

GNP is a measure of throughput — flows of stuff made and purchased in a year — rather than capital stocks, the houses and cars and computers and stereos that are the source of real wealth and real pleasure.

Page 142 · Location 2584


System behavior is particularly sensitive to the goals of feedback loops. If the goals — the indicators of satisfaction of the rules — are defined inaccurately or incompletely, the system may obediently work to produce a result that is not really intended or wanted.


Specify indicators and goals that reflect the real welfare of the system. Be especially careful not to confuse effort with result or you will end up with a system that is producing effort, not result.

Page 164 · Location 2981

Systems modelers say that we change paradigms by building a model of the system, which takes us outside the system and forces us to see it whole.

Page 164 · Location 2983

There is yet one leverage point that is even higher than changing a paradigm. That is to keep oneself unattached in the arena of paradigms, to stay flexible, to realize that no paradigm is “ true, ” that every one, including the one that sweetly shapes your own worldview, is a tremendously limited understanding of an immense and amazing universe that is far beyond human comprehension. It is to “ get ” at a gut level the paradigm that there are paradigms, and to see that that itself is a paradigm, and to regard that whole realization as devastatingly funny.