Archive for the ‘Systems Thinking’ Category

The following discussion is based on, and inspired by, the work of Jack Harich and associates, which can be accessed here.

A systemic problem is one that originates in the structure of a system rather in the behavior of individuals participating in the system. This does not mean that individuals play no role in causing the problem; rather, it means that replacing the individuals or attempts at changing their behavior, without changing the system’s structure, will not solve the problem. This is because in any social system there is a dynamic and dialectical relationship between the structure of the system (including all the interconnections, feedback loops, explicit or implicit rules, goals, etc.) and the individuals who participate in the system. The human factor is important—it is, after all, the people whose aggregate behavior is largely responsible for bringing the structure into existence and whose continuing participation is what maintains that structure over time. The structure, however, tends to acquire a reality of its own—becoming stronger than its individual participants in many ways—that both influences and limits the behavior choices of the people operating within the system. The structure not only encourages and rewards certain behavioral tendencies but also makes alternative choices harder to imagine, let alone implement.


A social problem that persists—and often gets worse—over time, despite the application of various intuitive or commonsense solutions, is likely to be a systemic problem whose root causes lie in the structure of the system. Such a problem cannot be solved unless its root causes are accurately identified and the appropriate solution elements devised to push at high leverage points in order to address those root causes. A root cause is defined as the deepest element in a causal change that is susceptible to resolution.

When a social system operates in a way that produces desirable outcomes, we can say that it is functioning in the right mode; when it operates in way that produces undesirable outcomes, we can describe it as functioning in the wrong mode. Solving a social problem is therefore a matter of shifting the mode of the relevant social system from wrong to right. But the persistence of a social problem over a long period of time, despite huge efforts to solve it, indicates not only that the relevant social system is operating in the wrong mode, but that it has somehow become locked into that undesirable state. In other words, structural mechanisms such as feedback loops have developed that prevent the system from changing in the desired direction. As soon as any effort to shift the system’s mode starts to succeed, these mechanisms spring into action and immediately reverse those gains. When effort after effort fails to change the mode of the system, activists ought to realize that trying harder is not the solution. They need to go back to the drawing board and examine their own assumptions about the causes of the problem. In most cases of stubborn social problems, the lack of success is not due to a deficiency of effort on the part of the activists but due to their incorrect diagnosis.

To arrive at the correct causal analysis of a persistent large-scale problem, social diagnosticians must consider three types of causal forces: (1) root cause forces, (2) superficial solution forces, and (3) fundamental solution forces. These forces (as well as new root cause forces) appear in blue text in Jack Harich’s “Standard Social Force Diagram” depicted below.


When a social system is locked into the wrong mode, that’s because root cause forces arising from root causes are operating to keep the system in that mode and to oppose and defeat any and all efforts to bring about a mode change. A system operating in the wrong mode produces undesirable outcomes or symptoms, such as deforestation, high morbidity, ineffective government, too many industrial accidents, lack of sufficient housing, and so on. Concerned citizens, who find these outcomes disturbing and unacceptable, reason their way backwards from the symptoms to their possible causes. Most of the time, however, they end their analysis prematurely—as soon as they have identified what appears to them as a plausible explanation for the symptoms but is, in reality, only a set of intermediate causes. Thinking that they have found what they were looking for, they focus their efforts at unleashing superficial solution forces, which leads them to devise superficial solutions that push on low leverage points in order to address the intermediate causes. As expected, their strategy fails to solve the problem since superficial solution forces are, by definition, weaker than the root cause forces. The error, of course, lies in the incorrect diagnosis.

A successful strategy to bring about a systemic change must begin with the correct diagnosis. Instead of ending their analysis as soon as they’ve found the first plausible explanation, activists must keep digging until they have identified the root causes of the problem that lie hidden in the system’s structure, sometimes deep underneath the intermediate causes. Once the activists have identified the root causes, they will be able to focus their efforts at unleashing fundamental solution forces, which will lead them to devise fundamental solutions that push on high leverage points in order to address the root causes. Since fundamental solution forces are, by definition, stronger than root cause forces, this is the only strategy that can bring about an actual mode change in the system. As the fundamental solution forces introduce into the system new root causes, the latter give rise to new intermediate causes, which, in turn, produce new symptoms (or desirable outcomes). The new root cause forces will also give rise to new structural mechanisms including feedback loops that will keep the system locked into the right mode.The persistence of the desirable outcomes over time, and the ability of the system to maintain itself in the right mode despite any opposing efforts, will indicate that the system has, in fact, permanently shifted from being in the wrong mode to being in the right mode. This is the definition of a systemic change. Anything short of that does not deserve to be called a “success.”

Read Full Post »

The_fifth_discipline_coverThe following post consists of quotations from Peter Senge’s book, The Fifth Discipline: The Art and Practice of the Learning Organization, along with my own attempts at paraphrasing them. My purpose is to put in one place the most important lessons that I think I should learn from reading The Fifth Discipline, and so the following material is best described as consisting of my personal notes more than anything else. Since this isn’t a summary of the book, I will select passages that appeal to me for one reason or another, and not necessarily because they’re central to the author’s argument.

There are three basic principles of systems thinking, as described in chapter 3.

Structure Influences Behavior: Different people in the same structure tend to produce qualitatively similar results. When there are problems, or performance fails to live up to what is intended, it is easy to find someone or something to blame. But, more often than we realize, systems cause their own crises, not external forces or individuals’ mistakes.

Structure in Human Systems is Subtle: We tend to think of “structure” as external constraints on the individual. But structure in complex living systems … means the basic interrelationships that control behavior. In human systems, structure includes how people make decisions—the “operating policies” whereby we translate perceptions, goals, rules, and norms into action.

Leverage Often Comes from New Ways of Thinking: In human systems, people often have potential leverage that they do not exercise because they focus only on their own decisions and ignore how their decisions affect others.

The structure of a system is found in the pattern of the interrelationships among the system’s key variables. We can easily identify the key variables of a system, but it takes time and effort to understand how they influence each other and how the patterns of their mutual influence changes in response to a change elsewhere in the system. Even though structures are not obvious, we can discern their power simply by noticing the feeling that compels us to act in particular ways and do what is expected of us. When you find yourself saying or thinking “I have no choice but to …,” it is very likely that you’re facing a structural constraint or imperative.

pogoIndeed, the most important factor that shapes people’s actions is the structure of the system within which they are operating. This is why very different individuals often end up behaving in very similar ways when they are placed within the same position in a system—for instance, an otherwise peaceful person can turn overnight into an arrogant bully  when assigned the position of a prison guard. Yet, it would be incorrect to assume that people are merely cogs in a machine with no agency of their own. In the case of a human system—such as a school, a company, or a prison—individuals do not exist apart from the system’s structure; rather, they are very much an integral part of it. This means that blaming someone or something else is not helpful since it doesn’t take us off the hook; nor is blaming the “system” a legitimate excuse. It also means that we are not at the mercy of external forces that are lie somewhere else, above and beyond our control; rather, we have considerable leverage precisely because we are intimately connected to the web of influences that defines a system’s structure. Just as the system’s structure influences and shapes our behavior, we too have the power to change some of the structure within which we function. The flow of influence is active in both directions.

Towards the end of chapter 3, Senge discusses the three levels at which any complex situation can be analyzed and explained. The resulting explanations may all be valid, but they do not have the same usefulness. Event explanations focus on gathering such information as “who did what to whom” and tend to trap us in a reactive mode. Pattern of behavior explanations focus on identifying long-term trends and figuring out their implications; this approach begins to liberate us from the reactive mode and allows us to deliberately respond to the situation. Structural explanations are the least common and the most powerful. They focus on understanding the system’s structure in order to find the deeper causes that generate the observed patterns of behavior; such explanations can help us identify the root causes of the situation, thereby empowering us to take the appropriate corrective action. According to Senge:

The reason that structural explanations are so important is that only they address the underlying causes of behavior at a level that patterns of behavior can be changed. Structure produces behavior, and changing underlying structures can produce different patterns of behavior. In this sense, structural explanations are inherently generative. Moreover, since structure in human systems includes the “operating policies” of the decision makers in the system, redesigning our own decision making redesigns the system structure.

It seems that one progresses through these levels of explanation by repeatedly asking the “why” question. Suppose a particular event happened—a relatively new car broke down, a marriage disintegrated, a child contracted an illness that is supposed to have been eradicated, a family lost their home to foreclosure. When we ask “why” for the first time, the answer will provide an event explanation, consisting of the specific and immediate chain of causation that led to the particular event. Thus, the house was foreclosed because Mr. Smith was laid off by his employer, which prevented him for continuing to make his mortgage payments.

Asking “why” a second time will broaden our view and help us see that the specific behavior of particular individuals that caused the event in question was not a rare or isolated occurrence; rather, similar behavior could be observed in many other cases as well. In other words, we find that we are dealing with a relatively common phenomenon resulting from the same overall pattern of behavior. There are larger processes and long-term trends at play. It turns out, for example, that it wasn’t just Mr. Smith who lost his home; tens of thousands of similar cases were happening across the nation, all at the same time, mainly because of certain reckless practices by creditors.

Understanding the pattern of behavior explanation is empowering, for it allows us to see broader trends, which helps us anticipate future events and plan accordingly—but we have not yet arrived at the root causes. The second level of explanation calls for asking the “why” question a third time. Why is this particular pattern of behavior occurring? What elements of the system are allowing and/or incentivizing people to act in this way? To answer this question, we’d have to go into the structure of the system where the underlying  causes are to be found.

The structural explanation will help us see, for example, that the reckless practices by the creditors were made possible by the popularity of certain financial instruments on Wall Street, which in turn was due to financial deregulation many years earlier as well as decline in government oversight, and that these two trends were themselves caused by other processes or mechanisms. We would have to keep digging until we reach the root causes, i.e., the set of final elements in the causal chain that are susceptible to resolution. This is the level at which we must apply the corrective action if we are to have any real impact on the situation.


To summarize, focusing on a particular event and asking “why” produces an event explanation, revealing only the immediate causes; problem-solvers are trapped in a reactive mode. Asking “why” a second time yields a pattern of behavior explanation, revealing the intermediate causes; problem-solvers can now respond to shifting trends instead of reacting to particular events. Asking “why” a third time leads to a structural explanation, revealing the root causes; such an explanation is generative, since it empowers the problem-solvers to exercise maximum leverage.

Read Full Post »

The following discussion is based on, and inspired by, the work of Jack Harich and associates, found at their website.

Problems come in all shapes and sizes, and they vary in terms of causes, scope, duration, etc. Here I am concerned with problems that seem to result directly or indirectly from the decisions made by individuals. Some of these problems are rarely encountered, since they occur due to mistakes or accidents, while other problems tend to occur over and over again.The problems that show up repeatedly may simply be the result of individuals making wrong decisions on a regular basis. Such problems are relatively easy to solve, since it usually doesn’t take a lot of effort to trace the symptoms of the problem to the particular decisions of particular individuals and then educating or training those individuals so they start making the right decisions.

Not all problems are so easy to solve. The most important problems are those that are so widespread that we might file them under the category of “social problems.” Typically, these problems persist over multiple generations and tend to resist commonsense or intuitive solutions. Examples include drug addiction, prostitution, obesity, and domestic violence. In any given case, we may be tempted to identify the victim’s own past decisions as the cause of his or her current undesirable state; for example, obesity can be traced to the consumption of large amounts of junk food and domestic violence can be traced to poor judgment in selecting a spouse. Upon deeper investigation, however, we discover that those past decisions were made under constrains over which the individual had little or no control, and that relative to those constraints the decisions seemed pretty rational at the time they were made. This insight becomes even more compelling when we approach these problems not as individual events or isolated occurrences but in terms of large-scale patterns that repeat themselves over and over again. If only a handful of people were to become addicted to alcohol or another drug in the course of a year, we may attribute the problem to the individuals’ personal situation and bad decisions; but when drug addiction starts affecting millions of people at any given time, we have to abandon our individualistic approach and start thinking in terms of social systems. For this to happen, we must zoom out in order to look at the behavior of many more people than just the ones who are directly involved, at which point we would realize that the problems that occur repeatedly and on a large, society-wide scale cannot be traced to particular individuals making wrong decisions. Rather, these problems occur precisely because all who are directly or indirectly involved are trying to do the best they can with the resources they have under the objective conditions they’re facing.

Such problems may be called systemic problems because they originate not so much from the behavior or decisions of individuals, but rather from the structure of the system itself.

Imagine a situation where everyone is making the right decisions—they are doing more or less what they’re supposed to—and it is precisely these decisions that are giving rise to certain undesirable symptoms on a large scale. You might ask: how can right decisions lead to problems? The answer is that these decisions are “right” only within the constrains of the particular system that is shaping the behavior of the individuals in question; and so, when we look at the situation from a viewpoint that is not constrained by the system, it becomes obvious that the decisions being made are actually wrong. This does not mean that individual behavior plays no role in the genesis of these problems. Individuals, after all, do not stand totally aloof or apart from the system; they participate in the system and thereby make it happen through their decisions. The point, rather, is that the decisions of individuals are themselves influenced by the rules and goals of the system in which they participate. We cannot, therefore, pin the blame for a systemic problem on anyone in particular. The individuals making the problematic decisions are part of the system, yet they are not the real culprits; their decisions only represent the intermediate causes of the problem. The root causes of the problem lies elsewhere, i.e., in the way the system itself is structured.

Consider obesity as a case in point. The responsibility for the problem of obesity in the United States cannot be placed simply on the poor eating habits of large sections of population. We must ask: why are so many people eating unhealthy food? It turns out that certain government policies are providing strong incentives for doing so, including subsidies for corn farming that makes high-fructose corn syrup available at low prices, which results in junk food being considerably cheaper than healthy options like fruits and vegetables. And if we ask the next logical question—why do these policies exist in the first place?—we would only uncover additional layers of systemic causation.

The lesson here should be obvious: We cannot solve a systemic problem simply by replacing some individuals or by educating or training them differently; the fresh arrivals will behave in the same way as their predecessors, and education or training will not have a major effect—unless we change the rules and/or the goals of the system. The only way to solve a systemic problem is to find and implement a systemic solution. The individuals through their decisions are merely responding to the expectations and constraints that the system imposes on them; if those expectations and constraints are changed, so would their decisions.

A systemic solution is one that change those elements of the system’s structure that represent the root causes of the problem. In contrast, a superficial solution is one that addresses the intermediate causes while ignoring the root causes. By definition, a superficial solution cannot solve the problem, or solves it only at an insufficiently small scale and/or temporarily. Thus, Michelle Obama’s effort to inspire and encourage healthy eating habits among kids is a superficial solution to the problem of obesity, just as her husband’s effort kill suspected terrorists through drone strikes is a superficial solution to the problem of terrorism. Once it becomes obvious that a particular solution is not working, it’s time to take a step back and consider the possibility that the solution being implement is only addressing intermediate causes. What is needed at that point is a deeper analysis of the systemic structure (i.e., the rules and goals of the system) in order to find the root causes of the problem so that systemic solutions can be devised and implemented. The insistence on using the failed superficial solutions not only prevents the problem from being solved while wasting time, energy, and resources; it also makes the problem bigger and more complicated and therefore harder to solve. In due course, superficial solutions themselves become part of the problem.

A solution consists of several solution elements. The place in the structure of a system where a particular solution element is to be applied is known as a leverage point. Results vary depending on whether the leverage point being pushed will affect the intermediate causes or the root causes. When a relatively large amount of force is applied at a particular leverage point in the system for a relatively long period of time but the systems shows only a small amount of change, then this provides good evidence that we’ve been pushing at a low leverage point, and that our effort only affected the intermediate causes. In other words, we were implementing a superficial solution, not a systemic solution. On the other hand, when the solution elements are applied at high leverage points in the system, they affect the root causes of the problem; in this case, the application of a relatively small amount of force can produce a large amount of change in the system.

Superficial solutions are based on an incomplete understanding of the system’s structure, which is why they provide solution elements that are applied at low leverage points and only address the intermediate causes of the problem. Systemic solutions, in contrast, are based on a sufficiently complete understanding of the system’s structure, which is why they provide solution elements that are applied at high leverage points and affect the root causes of the problem.

Read Full Post »

%d bloggers like this: