JAY FORRESTER: I think we should start by realizing that system dynamics is a profession like learning engineering or medicine. The idea that it is quick and easy to acquire is fallacious. We’ve had the experience of running a basic training program in system dynamics here at MIT the last three years by e-mail for professionals around the world; about 20 took it each year. It was a very intensive program in which participants received an assignment each week that took them about 15 hours to complete. That is a big load on top of their normal activities. The program ran for 30 weeks. Thirty weeks at 15 hours a week is 450 hours of work on their part. At the end, many said, “I’m now beginning to see enough of this field to know that I need to go further.”

We have seen people going to three-day conferences and thinking they’re experts in system dynamics. They set themselves up as consultants or to bring the ideas into a corporation when, in fact, they don’t have enough insight to know how to approach the subject. You can draw an analogy to medicine. I think a one-day first-aid course is useful. It will help you with simple things in medicine, but it does not prepare you to do heart transplants. System dynamics covers fully that wide a range.

The activity called “systems thinking,” which is talking about systems, recognizing there are systems, and agreeing that systems are important, is really at the level of the one-day first-aid course. It is not sufficient for understanding the dynamics of an organization. I have no doubt that a brief introduction can be useful; it just isn’t sufficient. The introduction from systems thinking is not strong enough and not persuasive enough to reverse detrimental policies that are strongly held, because there’s no solid basis for the argument to change. A systems thinker cannot, I believe, achieve the kind of position that one can have working from a good system dynamics simulation model.

With a solid, thoroughly studied system dynamics model, you know the assumptions that are in the model, you know the behavior those assumptions lead to, and you know how the behavior will change from a wide variety of different policies. If everything you say at the level of policies, at the level of structure, and at the level of behavior is correct in the eyes of participants who know various parts of the real system, it becomes persuasive. And that’s what the expert system dynamics practitioner should aspire to.

I would say that any attempt to introduce a deep understanding of systems on a broad sweep through the organization by simply talking about it probably will not be effective. The right way is to go deeply into some specific issues, do serious computer simulation modeling, and show how the troubles of the organization are being generated, how problems evolved out of past policies, and how alternative policies would improve behavior. To effectively create and use a model requires skill. People will spend tens of millions of dollars over a period of five years or more to develop a new product, but are reluctant to spend anything like that amount in preparing themselves to make the corporation more successful.

I think the way to introduce system dynamics is to find colleagues, build a grassroots understanding, develop skills, and get help from people who understand system dynamics very well. Engaging an expert consultant in system dynamics can accelerate the launching of a program. System dynamics is for solving problems. An effort should start by selecting an important problem. Decide what difficulty to work on and model its causes, rather than starting to model the entire system.

Most people believe their problems are created from outside. In system dynamics modeling, we usually find that the problems are being created on the inside. A corporation that is having problems and is being overtaken by other organizations is operating in the same outside world as those other corporations. Therefore, it must be something they themselves are doing that is causing them to be different and less effective.

Beyond the “Quick Fix” Mindset

DC: In some of my meetings with management teams, something seems to be missing in terms of the way that problems or issues are approached. The thinking somehow doesn’t encompass what you just talked about.

JWF: People expect a quick fix in a year. A company’s problems take years to develop and the fixes take years to repair the damage. If you look carefully at the difficulties of many major corporations, you find the cause of troubles began 10, 15, or 20 years before symptoms are recognized by management or the public.

It’s a mindset of the whole society that’s standing in the way — the mindset on short-term results. In complex systems, we usually see that policies that are good in the short run produce troubles in the long run and vice versa. Therefore, to do something good in the long run probably imposes some pain in the short run. With hired managers who are in their positions only one to five years, their personal interests tend to be in the short run. They are not committed to the long-run good of their organizations. The financial markets also tend to impose that same detrimental short-run view.

People will spend tens of millions of dollars to develop a new product, but are reluctant to spend anything like that amount to make the corporation more successful.

The attitude of founder owner managers is substantially different. Those who found a company, who are significant owners, and who are managing without expectation of going to some other corporation can have a 20 year view or more. I think many of today’s corporations that are being run by short view point managers will disappear in favor of a new wave of founder-owner-manager companies.

DC: Compare the current state of the field of system dynamics its development and acceptance with your own hopes and expectations.

JWF: It’s probably developing faster than I would have expected. The growth rate in the field number of people interested in it is probably doubling every four years or so, which is a very rapid growth rate. We are arriving now at the point where it really can’t be ignored. The consulting companies are looking more and more for people with system dynamics backgrounds. And, of course, my own work is helping to in filtrate system dynamics into the kindergarten through 12th-grade educational levels. I think it’s going very, very well.

A lot of people in the field express disappointment. They say, “Why isn’t it developing faster?” Well, it can’t because one of the great dangers is running ahead of the number of people who can practice it effectively and correctly.

I think one can argue that great frontiers don’t stay as frontiers; they become a part of everyday life. The most recent frontier has been exploring science and technology. I think the next great frontier is to truly understand and be able to improve the behavior of our social, economic, and managerial systems. The understanding of those systems has not improved markedly since the time of the ancient Greeks. My wife and I were taking a tour through the Alhambra, the great Moorish castle above Granada, Spain. Our guide stopped at one point to show us the room where the Moors around the year 1300 met to discuss their problems of inflation and balance of trade problems that are still with us.

It’s clear why the understanding has not advanced until recently there has not been any effective means to understand dynamic complexity of social systems. Mathematics is a weak science when it comes to dealing with dynamics. It can deal with extremely simple systems, but in realistic social systems, there is no possibility of getting mathematical solutions. Therefore, simulation is the only known approach. System dynamics modeling is like doing experiments in the laboratory instead of trying new policies on a corporation. The simulation experiment is much clearer than trying policy changes in real life because you know the circumstances under which you did the experiment, you know the policy changes that were made, and you know nothing else has changed. If you carry on those same experiments in real life, results are very ambiguous, because you’re never quite sure what other things affected the results.

Are We Doomed?

DC: In my work, I keep running into deep frustration that employees lower within an organization experience, and the high levels of burn out, cynicism, and turnover that result. I believe that this is linked to not using system dynamics more effectively in running organizations.

JWF: Well, you find a lot of reasons for frustration. Demands are put on people that they cannot possibly achieve, because they aren’t given the resources and authority to produce what is demanded. Further more, people haven’t been educated in what they need to know to succeed. Those lower-level people are squeezed by demands for impossible performance. That’s very frustrating.

DC: My sense is that people tend to blame upper management for the problem.

JWF: I suppose it’s fair enough from the viewpoint of the middle level to blame upper management. But upper managers are in the same situation; they also do not know what to do and are themselves under pressures from the outside world. Everybody is operating under the pressure of growth. Why should they be? Why not just run a successful business? It is not possible for everybody to grow beyond the capability of the system. We have great pleading to allow people to immigrate into this country because we think we need more labor. Why should we want to do more than we are able to do? The attitudes that society has drifted into create pressures not only at the middle levels, but also at the top of corporations. Many top managers are coping with problems created by predecessors who focused on the short run.

DC: Are we doomed?

JWF: A lot of large corporations are doomed, yes, and properly they should be. It is good for society to have decaying organizations eliminated. Great depressions like in the 1930s have traditionally helped solve such decay and inefficiency. Severe economic downturns wipe out a lot of dead wood incorporations and open the door for new, vital growth.

DC: What do you see in the future as potential breakthrough areas in the field of system dynamics?

JWF: We need a different approach from the one followed in the early days of system dynamics. The result of a system dynamics study will almost always show that the serious problem at hand arises from policies people know they are following but believe are the solution to their problems. You often have to say to people, “Your problems are because of what you’re doing that you’re proud of, and you must reverse your course.” It takes three or four years for them to accept that point.

We’ve often worked many years to introduce system dynamics into corporations from the top down. It takes three or four years with even a receptive top management before they fully understand. Then the people you’re dealing with retire or die, and you have to start over with another group. This is one of the main reasons why I shifted over to system dynamics in kindergarten through 12th grade, to bring up a society that has a better understanding of the nature of the systems within which we live.

It is much easier to teach system dynamics to fifth graders than it is to CEOs or parents. As children ask questions that people can’t answer, they learn it’s politically incorrect to ask questions that embarrass people. So they finally stop addressing the big issues. However, at age 10, children have much less to unlearn and they have more open minds. They are more inquisitive; they have not yet had stamped out of them the desire to understand what is difficult.

The Linking of People and Modeling Skills

DC: I have found a lot of fear in people trusting that there might be a different way to do things.

JWF: That’s right. There are two big hurdles in system dynamics: having enough people with enough competence, and finding out what to do about people being afraid to take the steps necessary to improve their situations. There are great challenges in implementing policies that are opposite to what people have been doing and what they believe is successful, even though those policies are getting them into worse and worse difficulty.

DC: Does it mean that system dynamics professionals also need to be fairly skillful in dealing with people? They can’t just be good modelers?

JWF: To be successful, yes. Sometimes we see a team where several people have different roles, with one doing dynamic modeling work and other spaying attention to how to get people to understand it, why are they balking, and why they find it so difficult.

DC: What career advice would you give young system dynamics practitioners?

JWF: I suppose the main thing would be to keep building skills. Very few have read all of the available good system dynamics literature, and probably very few are trying to establish an apprenticeship with an expert. Just as in medicine, one needs to go through an internship. One does not just go to medical school and then do major operations or deal with the most serious illnesses most learning comes from experience. You don’t learn system dynamics by just going to conferences; you must have working experience. System dynamics is not a spectator sport. Like learning to ride a bicycle, listening to lectures is not sufficient.

I would say that a young system dynamicist who goes by himself into a company and is doing simulations will have great difficulty in building acceptance. The best choice would be to go into a place where there is already some level of acceptance, but also to be careful not to be subverted by bad system dynamics that may already be going on there. There’s a lot of so called system dynamics work that is very bad practice. People who are not yet competent are trying to do things well beyond their ability because system dynamics seems deceptively simple. The major books in the field can be read by almost anyone. Urban Dynamics was, I think, on the list of books for discussion by the League of Women Voters and PTAs. People can read those books and understand them, and the process looks very straight forward. Then, when the person closes the book and says, “I’ll do some of that for myself,” there isn’t the slightest idea of what to do next. Also, the accessibility of system dynamics software allows people to build things that look like system dynamics models but may not be useful. There is a great need for processes for developing high levels of skill.

DC: If you were going to design a corporation, how would you introduce system dynamics?

JWF: It’s like introducing system dynamics into K–12 education. InK–12, very few places have a way to learn system dynamics. System dynamics needs to become a part of everything else that’s going on. It has to be widespread to be most effective. In a corporation, suppose we have a top management that has some serious problems, and the long-term dynamic solution requires reversing cherished policies. Assume top management accepts the reversed policies, they believe in the new policies, they are willing to act, and they issue instructions to do so. Below the top will be several levels of managers who see it as their duty to protect the organization from the idiosyncrasies of the top. The understanding of policy design must extend down through many levels.

DC: So even if you had a group of executives who decide to change policies, they may still be in trouble.

JWF: As an example, take one of the early studies in system dynamics done by a couple of graduate students. It dealt with a two terminal trucking company between Boston and Philadelphia. Their big problem was that trucks tended to be at the wrong end of the route. When they needed to ship things from Boston to Philadelphia, they had too many trucks in Philadelphia and vice versa.

The modeling showed how the terminal with extra trucks could provide prompt service, and business there would increase until trucks were concentrated at the opposite end, resulting in poor service and decline of business at the first terminal. Then service would improve and business would pick up at the second terminal. Business would swing back and forth as the stock of trucks shifted.

“The high-school teachers who know what’s going on here are terrified. They see the day coming when the elementary schools and the middle schools will be delivering to them little monsters who can think!”

The solution then became relatively obvious: They needed to keep their trucks balanced, even if they sometimes had to send empty trucks from one end to the other. Management understood and issued orders but nothing happened, because the people in the dispatching office and on the shipping dock knew the company didn’t make money driving empty trucks around. Nothing happened until the model was explained on the shipping room platform and in the dispatchers’ offices. I was told that one could hear truck loaders on the dock discussing the model. At that point, management could get the policy implemented, because now it made sense and everyone understood why. Previously, the lower level employees, who were making day-by-day decisions, saw the policies as totally irrational and believed they should resist in the best interests of doing a good job.

“Delivering Little Monsters Who Can Think!”

DC: Is your K–12 initiative developing as you had anticipated? What have you been learning from it?

JWF: Yes, I think it’s developing as rapidly as is reasonable. Moving to system dynamics is a difficult transition. A particular school may start with one enthusiast but, even in a receptive environment, it may take 10 years to get to a self-sustaining, school wide activity.

There are now conferences for teachers interested in system dynamics in K–12 education. My wife and I went to one at the end of June 2000. There were nearly 200 teachers present. Never before in any field have I gone to professional meetings where the excitement is so high and the enthusiasm about the future so great. Teachers come up to me and say, “I had no idea these students could do so much.” One teacher reported that in the junior high, the students in detention hall for misbehavior dropped from an average of 50 to an average of 5. It’s the first time they’ve seen students who want to come in before school starts and stay after it ends. My favorite sound bite was from the teacher who said, “The high school teachers who know what’s going on here are terrified. They see the day coming when the elementary schools and the middle schools will be delivering to them little monsters who can think!”.

DC: Jay, what are they doing in high schools around system dynamics?

JWF: Most of what’s going on so far is very elementary. Furthermore, there is not yet any continuous program that builds on itself from kindergarten through 12th grade, to say nothing about going on through undergraduate college and graduate schools. There’s no organized developmental path yet. Eighth, ninth, or tenth graders should be able to move well head of what is now being taught in the graduate schools. There’s the challenge of developing new material for high school, and for four years of undergraduate and three years of graduate school. That would be some 12 years of material not yet on the books. The system dynamics challenge for the future lies in developing such additional depth and breadth.

Worcester Polytechnic Institute is the first school to organize a four year undergraduate program leading to a bachelor’s degree in system dynamics. It will take a long time to get system dynamics introduced widely into teachers’ colleges. The teachers that are coming into system dynamics now are doing so through knowing others who have become enthusiastic and by reading the limited amount of available material. They become intrigued, they introduce a little into a class, and then it begins to evolve. You hear stories like a teacher begins to introduce system dynamics in biology, and other teachers find that their students are taking notes in system dynamics stock and flow diagrams. Then the other teachers go to find out what is happening.

System dynamics runs across all disciplines. The application to physics is fairly obvious because of the dynamics involved, but it’s probably not one of the most active areas. In social studies, students can use system dynamics to explore the economic and social forces causing various things to unfold in history. There are English teachers doing computer simulation modeling of the psychological dynamics in various pieces of literature. System dynamics provides a foundation that underlies most of the subjects. A student discovers a new mobility between subjects. If one understands a particular dynamic structure in one setting, the behavior is the same in all settings.

DC: So it seems as though for children who experience feedback dynamics over a number of years, there’s a deep difference in the way they perceive things.

JWF: Entirely different. One father of a junior high boy told me his son had gone on a tour of Europe. He came back with descriptions of the way things are interrelated far beyond what you see in the newspapers, because he looked at things differently. Another boy, when asked what all this has meant to him, said, “I’m much better able to deal with my mother.” It gets down to things that matter. After all, that’s a very complicated interacting feedback system.

The system dynamics mentor to the schools in Glynn County, Georgia, has written that some of her most interesting experiences come when she is talking with teachers and students about modeling discipline problems. As she develops a diagram of the processes and interactions going on, the students suddenly see why what they’re doing gets the teachers so frustrated. The teachers also see that the discipline system they’ve set up is preordained to create trouble.

On the Cliff’s Edge

DC: When you wrote Urban Dynamics, you bumped up against people’s cherished beliefs by putting out something very different. What allows you to continue to work when you stir up a lot of controversy or receive a lot of criticism?

JWF: Pioneers always find that. The very definition of pioneering is that you’re doing something that people don’t already know and don’t already believe. I have been a pioneer in several different fields. In the early days of digital computers, we were building a computer using the binary system of calculation. Many people said it would be useless unless we did decimal calculations. Of course, all modern digital computers are binary.

Before that, I’d been in the pioneering of feedback control systems for the military in World War II. I graduated in electrical engineering, electronics, and we found ourselves working on systems to control Army gun mounts. The Army wouldn’t trust anything made out of electronics except their radios. So my first professional job was to design high performance controls using hydraulic oil pressure, with an emphasis on reliability. We were doing one of these for the Navy, and the question naturally arose, what will happen with the equipment in an ocean environment? So I thought I’d better know. I went down to the beach and brought back a gallon of genuine Atlantic Ocean water, mixed it half and half with the oil, and ran the equipment in it all winter. Everything still worked.

DC: So part of the success of moving the field of system dynamics forward has to do with your own comfort in being a pioneer?

JWF: Yes.

DC: And knowing what that entails as you’re moving things forward.

JWF: Knowing the opposition, knowing a little bit about how you bridge the gap. But it takes time. And, of course, being very much in touch with the real world. I grew up on a cattle ranch in Nebraska. And there, if things didn’t work, you found out fast. In my senior year in high school, I built a wind driven electric plant that provided the first electricity we had on our ranch. And it worked. So I think you develop a feeling for where the edge of the cliff is. If you step out too far, you’re a crackpot and you fall off. If you stay back too far, you’re just part of the crowd.

The post A Pioneer on the Next Frontier: An Interview with Jay Forrester appeared first on The Systems Thinker.

RIGOR VS. SUPPORT

The Facilitative Modeling approach for making important decisions combines high levels of analytical rigor with high levels of stakeholder support.

Common Decision-Making Approaches

Because of this level of complexity in all aspects of organizational life, organizations usually rely on what I refer to as the “shoot from-the-hip” approach for making important decisions. You’ve seen this technique if you’ve ever been in a team meeting in which a decision must be made today. Some members of the group toss out their ideas; most participants stay silent. Eventually, the team leader contributes his or her opinion, and everyone agrees. Decision made! Most meeting participants later bemoan the “poor” decision, claiming they won’t support it. The result? The new policy dies on the vine prior to implementation, leaving the organization the same as it was before.

In analyzing “shoot-from-the-hip” decisions, we observe that they lack strength in at least two major areas: analytical rigor and stakeholder support (see “Rigor vs. Support”). This isn’t a novel observation: Organizations have struggled with these two shortcomings for years and have devised various ways to overcome them.

1. The Technological Approach Before making a major decision, in order to increase the level of analytical rigor (or understanding of the issues), managers often rely on analysts and their toolkit — what I call the Technological Approach. Organizations adopting the Technological Approach generally do so because they’ve fallen victim to the mindset that they must find the perfect answer. The idea is that if you throw enough analysis at an issue, you can completely understand everything and uncover an ideal solution. These organizations think the answer must be found in the numbers.

To process the data they generate, organizations subscribing to the Technological Approach employ spreadsheets and statistical techniques. Some even build large simulation models to test nearly infinite possible scenarios. However, these tools can obscure the assumptions underlying the analysis. And because decision-makers aren’t privy to these hidden assumptions, they cannot compare them to their own mental models — so they do not trust the resulting recommendations. This lack of trust in the analysis is a major factor in why, although usually carefully applied, the Technological Approach rarely generates the support needed to lead to effective policy-making.

2. The Stakeholder Approach In contrast, proponents of a Stakeholder Approach often put technology aside and instead try to build knowledge and support through stakeholder involvement. Well-known techniques that follow this approach include Future Search, Open Space Technology, the World Café, various forms of dialogue — even some facilitated mapping sessions using causal loop diagrams and systems archetypes. These methodologies share an underlying mindset — by getting representation from different players in “the system,” everyone will gain a broader view of the problem at hand. Further, by allowing participants to express divergent perspectives in an unconstrained fashion, the Stakeholder Approach lets them formulate creative, systemic recommendations.

Whether trying to define the problem or to generate solutions, people applying these processes (if only implicitly) tend to follow a model of interaction described by Interaction Associates as the Open-Narrow-Close model. During the Open phase, participants get all of the data on the table while defining the problem; if they’re generating solutions, this is the stage in which creative solutions spring forth from the group’s collective wisdom. During the Narrow phase, contributors take an overwhelming list of choices (problems or solutions) and narrow them down to a few to consider further. During the Close phase, they actually choose which problems to tackle or solutions to implement and how to do so. Managers then often assign groups to each of the major action items identified during this stage and give them their blessing to “go forth and implement.”

The Stakeholder Approach includes processes that build broad support — unlike what often occurs in the Technological Approach. Plus, it helps those involved to see the system from a broad spatial and sometimes temporal perspective. These results are necessary and important for creating effective changes in any system.

A major weakness of the Stakeholder Approach, however, is that the processes used to narrow and choose

APPROACHES FOR IMPLEMENTING SYSTEMS THINKING

Facilitative Modeling serves as a middle ground between the Technological Approach and the Stakeholder Approach.

among the resulting divergent issues/strategies lack rigor and usually

rely on the assumption that, simply by having enough stakeholder representation, the group will make excellent decisions. But as Irving Janis learned by studying extremely poor decisions (such as the Bay of Pigs fiasco and the escalation of the Vietnam War, which he described in his book Groupthink), groups with very high average IQs can function well below expectations.

Barry Richmond of High Performance Systems, Inc. created a simple example called the Rookie-Pro exercise that also illustrates this point. Despite working with a much simpler human. resource system than that found in most organizations, only 10 to 15 percent of individuals can guess the system’s future behavior — even after lengthy discussion! So the assumption that the collective wisdom of the group will surface in a way that leads to optimal decision-making is tenuous at best.

In addition, the framework employed to guide team members in narrowing and choosing among different options doesn’t help to determine if elements of the proposed solutions need to be implemented at different times and in varying degrees. The result is that the organization often chooses to put the same amount of resources and effort into each action item. Nor does the Stakeholder Approach determine if the issues are interconnected — different groups may be separately implementing policies that should be done together or, even worse, are mutually exclusive.

Facilitative Modeling

The good news is that there is a way to both rigorously understand (or

even reduce) complexity and improve stakeholder support! Practitioners are often drawn to the field of systems thinking because of its promise to build collective understanding — to get everyone on the same page. Even so, these managers can be pulled between the Technological Approach (big simulation models created by experts) or the Stakeholder Approach (facilitated sessions using causal loop diagrams or systems archetypes). But there’s a middle ground — a large range of activities that I refer to as “Facilitative Modeling” — where tremendous power resides (see “Approaches for Implementing Systems Thinking”).

Facilitative Modeling is a Technological Approach, because it uses computer simulation and the scientific method to build understanding. It is also a Stakeholder Approach, because it requires the input of the important stakeholder groups, uses a common language so everyone can get on the same page, and creates small, simple, and easy-to-understand models. The models don’t generate the answer; rather they facilitate rigorous discussion. Facilitative Modeling usually culminates in a facilitated multi-stake-holder session in which the participants generate common understanding and make well-informed decisions.

Overview of the Process

In the Facilitative Modeling process, a group of stakeholders identifies and addresses an issue critical to their collective success. The issue is often one that has been resistant to organizational efforts to “fix” it. After choosing the area for exploration, the group sets the agenda for a facilitate

session. In preparation for that meeting, several individuals in the group serve as a modeling team and develop (alone or working with a modeler) a series of simple systems thinking simulation models that clearly articulate important components of the issue. These components may include the historical trend for that issue, the future implications if the trend continues, possible interventions, and the unintended consequences of some of these solutions. The models are deliberately kept small so that stakeholders will understand them and the development process remains manageable.

However, it’s not enough just to make models! In fact, building useful models is probably less than half of what makes a Facilitative Modeling initiative successful. The process requires the modeling team and perhaps others to create additional materials for the facilitated session, such as workbooks for tracking experiments and writing reflections, as well as CDs of the models for after the session. A facilitator and/or design team needs to carefully plan various aspects of the session, such as appropriate questions, suggested experiments to run on the model, and a mix of small and large group discussion.

The facilitated session represents the culmination of the process. During the gathering, teams of two to four people explore the models on computers. The session includes large group interludes and debriefs between exercises. And at the end of the session, participants discuss and agree on

next steps based on the insights that emerged during the event (see “The Facilitative Modeling Process”).

Facilitative Modeling in Action

Using the Facilitative Modeling Process outlined above, a nonprofit organization recently explored potential issues associated with implementing new funding policies. This organization was responsible for improving the health and welfare of the poor population in a community by giving funds to other local nonprofits to provide services. Originally, the organization had determined which organizations to fund and how much funding to supply by analyzing the services that the target organization would provide; in recent years, it had settled into just increasing the amount of funding incrementally over the previous year’s figure. To create more accountability among the local organizations and improve outcomes in the community, the nonprofit had decided to apply a performance driven approach to funding (that is, base funding on projected improvements to performance indicators and then renew the funding if the community experienced noticeable improvement in those areas).

Some members of the organization, as well as members of an important partner group, were concerned about the potential barriers to implementing this updated approach and were eager to understand possible unintended consequences that might result from the change. They agreed that a Facilitative Modeling approach would be an excellent way to surface and discuss these issues in a way that would give all stakeholders shared insight. In little more than five days of working with a facilitator and a few representatives from the organization and its partner, the team developed three small “conversational” models for a one-day facilitated session.

At the beginning of the session, the group adopted a set of ground rules to guide their interactions. Once participants agreed to the guidelines, they began by experimenting with the first model. The purpose of this initial simulation was to surface and discuss the potential dynamics associated with implementing the new funding approach. Allowing “sub groups” to work with the models at their own speed often increases their level of understanding. However, even those with some skill at reading stock and flow diagrams similar to the one shown here can be quickly overwhelmed by maps. The simulation included a function that let the sub groups slowly unfurl pieces of the map so that they more easily followed its logic (see “The First Map” on p. 5).

The map shown here represents one way to look at the different organizations affected by the nonprofit’s funding decisions. The language of stocks and flows is ideally suited for looking at this issue. The three stocks at the top of the diagram (the rectangles labeled “Resistant,” “Not Committed,” and “Committed”) represent groups of organizations. Currently, because the new approach has yet to be implemented, all organizations would belong in the “Not Committed” stock. Eventually, as the new funding approach is made into policy, organizations would begin to move into the “Committed” or “Resistant” stocks. Obviously, if possible, the funding group wanted to avoid any organizations becoming “Resistant.”

At the session, the individual groups discussed the meaning of each of the stocks. What does it mean to be “Committed”? “Resistant”? They mulled over the question, What number of “Resistant” organizations would pose a problem for the program as a whole? Can “Committed” organizations become “Resistant”? Is it realistic to assume (as the model does) that “Resistant” organizations never become “Committed”?

Talking about the diagram helped he sub groups, and eventually the entire group, reach consensus about how organizations might become committed or resistant to the changed funding policies. For many of the participants, it was the first time they had discussed the potential that some of their client organizations might resist the changes! By working with the model, the group was able to surface an unpleasant concept in a way that allowed them to grapple with its implications for their changed strategy.

They then entered different values into the model to experiment with how the funding organization might allocate its resources in the coming months. How much effort should they put into developing the performance-driven funding program? How much into explaining the program to the funded organizations? And how much of each should they do prior to officially announcing the program? After announcing it? In short, the group wrestled with the systemic or “chestration” (a concept developed by Barry Richmond) of resources the magnitude and timing of efforts required to successfully implement the strategy.

The group concluded that, in the first phase of development, they should apply most of their efforts to designing the new policy. Doing so builds the “Clarity of the Program,” which is useful in preventing “Doubts About the New Approach” down the road. They realized that they would need to allocate at least some resources in the first phase to working with the client groups and addressing their doubts about the change. This process would also help them to refine the approach (see “Implementation Timetable” on page 6). The next phase would require additional work with the other stakeholder groups to explain the program prior to release. The third and fourth phases would involve implementation; this is when the nonprofit’s staff members would spend most of their time addressing the doubts of the affected organizations.
The group realized that the exact numbers of organizations in each category wouldn’t be the same in real life as in the simulation, but that the stories described by the model were consistent with what they now expected might happen when overhauling their approach to funding. In keeping with the need for systemic orchestration the group concluded that their allocation of strategic resources must shift over time, depending on which phase they were in (for example, in the second phase, they would need to apply some resources to program development and even more to working with stakeholders).

Working with Subsequent Models

In Facilitative Modeling, each model tends to add to the understanding generated by previous ones. Because the performance-based funding approach would require implementing a new IT system, the second model helped participants explore how a funded organization would need to allocate resources in order to develop a new IT system and build its staff ’s capacity to use it. The third model served as the capstone exercise, because it required participants to explore how client organizations might allocate their resources across the following needs: providing services, building and maintaining the IT system, investing in staff skill development, and collaborating with partner organizations.

THE FIRST MAP

The three stocks at the top of the diagram (the rectangles labeled “Resistant,” “Not Committed,” and “Committed”) represent groups of organizations. As the new funding approach is made into policy, organizations would begin to move from the “Not Committed” stock into the “Committed” or “Resistant” stocks.

During the large-group debrief of the third model, the nonprofit’s senior director said that he didn’t like one dynamic that he experienced with the model. In all cases, after the funding change, the youth population’s sense of disconnection from the community initially worsened, even when the simulated strategies encouraged a majority of client agencies to be committed to the shift and to effectively implement performance-based approaches to providing services. When he experimented with the model, the director kept trying to avoid this “worse-before-better” dynamic. Through probing questions, the group learned that it wasn’t that he didn’t expect this behavior to happen, he just wished it wouldn’t!

IMPLEMENTATION TIMETABLE

By using the model to explore the magnitude and timing of efforts required to successfully implement the strategy, the group concluded that, in the first phase of development, they should focus on designing the new policy.

This revelation led to an interesting discussion of what is often an undiscussable in the public sector: that policies designed to improve social systems often take time before they lead to noticeable improvements and that there is often conspicuous degradation of performance in the interim. The director expressed that it was political suicide to admit that things might actually get worse before improving. Ultimately, through the facilitated discussion, he came to understand that regardless of whether he wanted to admit that such a dynamic might occur, it was inevitable, given the long delays before activities such as IT development and skill-building would have a positive effect on services. Through this admission, he and his staff were then able to explore options for mitigating the effects of this unavoidable dynamic.

Ultimately, the nonprofit’s staff left the session with useful insight in several areas. First, they all understood that some of their client organizations might resist the new approach. Second, they realized that it would be helpful for them to include those organizations in developing the program. Third, the group agreed that building staff skills was likely to be a more challenging impediment to successful implementation of the changed approach than developing the IT infrastructure. Finally, they accepted that systemwide implementation would require orchestrating a series of activities that, even in the best of circumstances, would cause a “worse-before-better” dynamic. All of these insights were just the beginnings of an ongoing dialogue, and all were facilitated by using small models to focus the conversation.

The Value of Facilitative Modeling

As shown in the example above, there is a powerful place for small models in a facilitated environment. The process used for developing good systems thinking models increases the rigor of the analysis and captures the benefits of a Technological Approach. At the same time, by keeping models small, Facilitative Modeling improves on the benefits of a Stakeholder Approach and increases the likelihood that all participants end up in alignment. Moreover, the Facilitative Modeling approach uses a language — stocks and flows — that is more representative of reality than other visual mapping languages. For this reason, the participants are able to discuss and come to a novel understanding of the assumptions built into the model. Running the simulation provides an essential test of the group’s understanding and facilitates further conversations about the likelihood of different results. The computer-generated “microworld” creates a safe environment for experimentation.

The post Facilitative Modeling: Using Small Models to Generate Big Insights appeared first on The Systems Thinker.

Executives today are constantly seeking to predict how their organizations and the marketplace will behave. But because many leaders continue to use traditional reductionist methods to understand organizational behavior ones that focus more on symptoms than on causes of a company’s success they fail to gain real insight into how to build and sustain that success. The result is often reactive, crisis driven management with unanticipated side effects and unforeseen outcomes.

Contrary to this rigid perception of organizations as predictable machines, some management thinkers have come to view them as complex and evolving organisms. Accordingly, the tendency in the business world to define companies in terms of simple formulas and numerical results is slowly being replaced by the recognition that, to be effective in leading organizations, we must think of them in terms of the underlying structures and dynamic patterns of behavior that produce those results. In other words, we must begin to complement or replace linear thinking about how our businesses work with nonlinear approaches by applying the principles and tools of system dynamics.

System Dynamics Theory

Why do so many brilliant management strategies lead firms directly into decline or not produce the anticipated results?

In his classic 1961 book Industrial Dynamics, Massachusetts Institute of Technology professor Jay Forrester originated the ideas and methodology of system dynamics. He pointed out that the traditional approaches of the management sciences could not satisfactorily explain the causes of corporate growth or decline because they focused on simply explaining behavior. He believed that a system’s behavior is actually a product of its structure and that leaders should seek to identify where changes in structure might lead to significant, enduring improvements. They could then design organizational policies and processes that would lead to even greater success.

In order for managers to undertake this design process, Forrester advocated that they must analyze their organizations using dynamic models. For this purpose, he developed tools such as causal loop and stock and flow diagrams. These tools serve to illustrate the interconnected feedback loops that form a complex system. By identifying these feedback loops, management can figure out a system’s basic patterns of behavior, which include growth(caused by positive feedback), balance(caused by negative feedback), oscillations(caused by negative feedback combined with a time delay), and further complex interconnections.

Applied to organizations, this way of thinking challenges the notion of measuring success only through financial results. Because people can see financial results, they think they have control over them. But these results are actually produced by the organization’s underlying structures. These structures consist of:

• Organizational Architecture: the basic organizational design (such as the functions or divisions that the company includes) and the governance system (such as the planning and control system)
• Organizational Routines: standard operating procedures, decision making processes, behavioral archetypes
• Tangible and Intangible Resources: financial capital, human resources, buildings, machinery, land, brands
• Organizational Knowledge and Value Base: patents, core competencies, cultural beliefs, attitudes

When we focus on systemic structures and behavioral patterns, we gain the knowledge to design our organizations to produce desirable day-to-day results in areas such as profits, employee motivation, customer satisfaction, and so on (see, “Structure, Behavior, and Results”). The basic idea of the dynamic approach is that, although people shape their organizations, their behavior is ultimately influenced, and therefore limited, by the organizational framework in which they operate. Consequently, leadership means much more than optimizing businesses for short-term outcomes; it involves creating and cultivating structures and enabling organizational behaviors that guarantee the viability of the whole firm. Therefore, in order to manage their organizations successfully, leaders must realize that the best way to achieve sustainable results is not by relying only on what they see or measure but by:

• Describing and assessing the observable behavior of the system;
• Understanding the interdependencies between a system’s behavior and its underlying structure;
• Making assumptions about and modeling these interdependencies using system dynamics tools; and
• Finding and implementing policies to redesign the structure of the system in order to improve its performance.

Building on this system dynamics foundation, we propose to take leadership one step further, to what we call evolutionary leadership. The natural process of evolution offers a compelling model of how leaders might intentionally design and guide growth and balancing processes to create a viable organization. Evolutionary leadership involves the deliberate interplay of two management functions: strategic management (designing structures and processes that stimulate growth) and management control (guiding the external and internal factors that regulate growth). But before we explore the synergy between these two functions, we need to talk about how evolution works in nature and in organizations.

Evolutionary Theory in Organizations

Evolutionary theory has been the predominant paradigm in natural sciences for more than a century. Recently, theorists and practitioners in the social and management sciences have begun to adopt the ideas of evolutionary theory as a framework for describing and analyzing organizational development. The basic concept these pioneers have set forth is that processes of variation, selection, and retention as well as the struggle for scarce resources trigger the evolution of an organization.

Sociocultural evolution differs from biological evolution in that it allows for the intentional variation and selection of ideas. In this context, an organization’s fitness its “viability,” or ability to survive and thrive depends on how its decisions and strategies affect its position in product and resource markets and on its legitimacy from the point of view of important stakeholders. Chilean neurobiologists Humberto Maturana and Francisco Varela have deeply influenced thinking about viability with their theory that living systems are complex systems that can self-generate. A system dies when it loses its ability to renew itself. In the business world, a company that fails to renew itself by changing its strategic orientation and/or internal structure in response to shifting conditions will die. In contrast, a viable organization is one that can continually create its own future and there by assure its fitness in an evolutionary sense.

But how does a viable organism develop this capacity to self generate? According to Maturana and Varela, it happens when the organism

• Preserves its identify by repeatedly drawing system boundaries (i.e., defining what is “internal” and, “external”); and
• Maintains its ability to adapt to a changing environment.

Within ever-changing environments, external forces constantly threaten the existence of a species by altering its living space. To survive, a species must adapt to the changing conditions successfully without losing its identity. For example, in nature, many kinds of birds have adapted from natural to urban environments, but not all have managed to do so. In the banking industry, banks have profoundly shifted their strategies in the past decade in response to technology changes and new competitors. Many brick and mortar institutions have gone “virtual.” In doing so, they are able to maintain their existence by simultaneously preserving their identity while adapting their strategy and structure to a changing environment.

The key to an organization’s survival lies in mastering the trade-off between preserving its identity and adapting to a changing environment. Leaders do so through strategic thinking and acting, and by asking how they can maintain the fit of the organizational structure and its environment. There are two ways to achieve this goal:

• Maintain your identity and structure and avoid fundamental adaptations by changing the environment or searching for an appropriate new environment.
• Fundamentally change your structure and redefine your identity to reestablish a fit between the organization and its ever changing environment.

In reality, most organizations choose adaptation strategies that lie somewhere between these two extremes.

Organizations can only make alterations to the extent that their structures and resources make modifications possible. A firm has a good chance to successfully adapt to a changing environment when it has a strong learning capacity, that is, the ability to anticipate, influence, and quickly react to environmental changes, along with the ability to recognize, vary, and advance the underlying mechanisms of the learning process itself. For example, Shell Oil enhances its learning capacity by combining strategic planning and organizational learning through scenario planning. Scenario planning provides a mechanism for thinking in alternatives and making underlying assumptions explicit. This process reduces the company’s risk of encountering negative surprises and increases the speed with which it can implement changes. In short, organizational learning is a dynamic feedback process that can help organizations remain viable and therefore survive the external pressures of natural selection (see “The Evolutionary Cycle in Organizations”).

Growth and Balance

In addition to having the ability to adapt and learn, systems must be able to grow. Generally speaking, growing means incorporating more and more available resources like nutrients for a plant or natural or human resources for a company in order to become larger and larger. For a company, growth can mean an increase in market share or market value. But is growth in itself sufficient for survival? Clearly, the answer is no, because nothing grows forever. But where and what are the limits to growth?

In nature, reinforcing processes, such as population growth, are slowed by balancing processes, such as limited food supplies and the spread of diseases. If normal balancing processes aren’t blocked and assert themselves before a population reaches the limits of its habitat, that species can maintain a harmonious relationship with its environment. Such balancing processes ensure that the evolving system remains within a viable range of activities, in this case, healthy population density. Indeed, these balancing processes are more crucial than reinforcing processes, in that they keep the overall system alive. If, on the other hand, important balancing processes are missing, the species might become extinct by overtaxing the resources in its environment.

Are there similar natural boundaries to the development of social systems? The answer is yes. For example, a firm’s development can be limited by its production capacity, the size of its market, or the number of its competitors. The faster the company grows, the more rapidly it reaches these boundaries. From time to time, such limits to growth can change. For example, shifts in market conditions, such as those created by the Internet boom or the world oil crisis of the 1970s, can increase or decrease the time it takes an organization to reach a certain limit, unless people find ways to use their limited resources more efficiently.

We can say that an organization is evolving when its configuration, routines, tangible and intangible resources, knowledge, and value base develop in accordance with the changing external environment. Scientists now know that most healthy living systems follow a developmental path described as punctuated equilibrium periods of balanced growth that are interrupted by periods of exponential growth (see “The Stages of Organizational Evolution” on p. 4).

We regularly underestimate the tremendous power of exponential, or reinforcing, growth. We tend to assume that growth is linear and increases consistently over time. However, exponential growth happens much more precipitously. If we observe the two over a short period of time, exponential growth approximates linear growth. Over a longer period, however, the gap between the two becomes enormous.

Because human beings tend to perceive short term rather than long-term changes, we often reach the boundaries of exponential growth faster than we anticipated, often completely unexpectedly. We see this happen to companies when booming success is followed by equally dramatic failure. For example, cellular telephone companies experienced this phenomenon when they projected that their sales would continue to increase at a high level. But they eventually saturated the market and experienced declining sales. For this reason, unless we understand and anticipate the impact and boundaries of exponential growth, we will have a distorted perception of the evolutionary process, leading to unpleasant surprises and even to an existential crisis for the whole enterprise.

Organizations sustain themselves when they attain a balanced evolution off setting reinforcing growth action with timely balancing impulses. Sustaining this balance is the only way to ensure that companies remain in the realm of “sound growth” as they develop and that they don’t exceed the limits of their environment or resources. Balanced evolution plays an especially critical role during periods of exponential growth, when the organization is at a much higher risk of losing its viability than in periods of balanced growth, when the stakes aren’t as high.

For example, when a leap in growth occurs for a limited time(through external factors such as deregulation or new developments in technology, or through internal factors such as changes in top management or a merger and acquisition), leaders need to off set that growth by intentionally introducing balancing feedback loops. They can do so through control and coordination systems as well as productivity enhancement programs. These loops keep the organization’s growth from consuming the company.

Leadership in Organizational Evolution

But how can leaders help firms achieve the balanced growth they need to evolve? Through strategic management, leaders expand the business; through management control, they regulate the growth process, making sure that it remains within a sustainable range. Together, the two functions form a balanced leadership cycle for guiding and controlling the company’s evolution.

Strategic Management. Through strategic management, leaders cultivate the conditions for a company’s sustainable growth. Specifically, they perform the following three functions:

1. Set Direction. As mentioned earlier, leaders need to preserve or redefine the organization’s core identity and develop its structures in ways that lead to lasting success. They do so by communicating the company’s values and beliefs to employees and external stakeholders through shared vision and mission statements, and by strengthening internal rein forcing processes such as employee morale. They also formulate and implement strategy, not by detailing a map of action but rather by defining a corridor of learning opportunities.
2. Build Resources. Leaders need resources to support entrepreneurial activity. They can acquire them externally (such as machinery or capital) or develop them internally (such as people or policies). From a resource based perspective, only internally built resources can provide the basis for competitive advantages and above average returns, because they are specific to the company and therefore more difficult to imitate. On the other hand, resources that are available on the open market are available to all competitors.
3. Create Infrastructure. Leaders must not attempt to drive growth but rather to influence the factors that can block or support it. As such, they need to design an organizational context that eliminates barriers to company development (such as fear, distrust, centralized decision making, too tight control, and insufficient resources) and develop processes to promote learning (such as organizing flexible teams, supporting communities of practice, creating incentive systems for transferring knowledge, and creating learning spaces).

From a system dynamics perspective, these three functions combine to form a reinforcing process called the “Strategic Management Loop,” which strengthens the company’s growth(see “The Balanced Leadership Cycle”). But for the organization to remain viable, this reinforcing loop must be reined in by balancing processes, such as those that make up the “Management Control Loop.”

Management Control. Management control acts to bring equilibrium to the expanding system. To do so, leaders must perform three central functions:

1. Assure Internal Consistency of Infrastructure, Resources, and Direction. Leaders need to maintain the coherence of a system, particularly in large companies where management functions often get split among different organizational units or departments. To handle this specialization of functions, they must synchronize the development of strategy, resources, structure, and systems. They do so by working with others to develop a shared view of the system, which acts as a basis of companywide activity. However, this model is necessarily a subjective simplification of complex reality, so it can easily become selective and distorted.
2. Compensate for Selective Perception. Therefore, leaders and their teams must compensate for their selective perception by continually enriching their assumptions with relevant new information and challenging their mental models. For example, they might use management information and decision support systems, which provide comprehensive data and make blind spots of organizational perception visible. Management control thus leads to more informed decision making and better anticipation of the consequences of those decisions.
3. Appropriately Limit Developmental Dynamics. Designing appropriate limits on developmental dynamics involves two realms: content and time. Leaders must analyze whether the firm’s expansion exceeds the limits set by its internal conditions (for instance, the number of staff with expertise in certain areas) and the external forces of its environment (for example, the size of the market), thus endangering its boundaries. They also must regulate how fast the firm grows. They do so by pacing the speed of growth so it doesn’t over tax the current management capacity (resources and infrastructure) or environmental limits (size and growth of the market).

{page5 image1 title=”THE BALANCED LEADERSHIP CYCLE”}

Leaders put these functions into action using different diagnostic tools, such as the balanced score card and budgeting. The balanced scorecard helps them see the inter connections among the key measures of the business, for instance, between employee capacity and customer satisfaction, or between customer satisfaction and market share. Executives can then ensure that key measures stay in balance. Through the budgeting process, they translate strategic direction into financial objectives, setting the frame work for the allocation of resources and the utilization of infrastructures to assure internal consistency. By limiting and balancing developmental dynamics as well as by assuring internal consistency, these tools contribute to the fulfillment of the management control function in the balanced leadership cycle.

In order to avoid survival threatening oscillations between growth and decline, leaders need to take into account the time delays that occur before balancing impulses take effect. Working properly, the interplay of strategic management (growth actions) and management control (balancing impulses) assures a synergistic rhythm of a company’s evolution, a characteristic of particularly successful firms in dynamic environments.

The post Evolutionary Leadership: A Dynamic Approach to Managing Complexity appeared first on The Systems Thinker.

Anticipating System Behavior

School districts are made up of many components, including district staff, individual schools, teachers and administrators within those schools, parent councils, and students. The sheer number and variety of these actors make it difficult to see their interdependence and to notice how an action in one part of the system affects the others. Add to this complexity policies originating from agencies outside the district, such as state education departments and the U. S. Department of Education, and the task of assessing how best to direct resources to meet students’ needs becomes almost hopelessly confusing.

Systems thinking and system dynamics tools, including casual loop diagrams, stocks and flows, and computer simulation, can shed light on the interrelationships among components and, perhaps more important, illustrate how outcomes may result from feedback loops rather than from simple, linear chains of cause and effect. These tools also make explicit the delays that often occur between a change in one component of a system and its effect on others. The interplay of feedback and delays can produce unanticipated system behavior, as shown by the mandating of smaller class sizes in California. When the legislature passed the new law, schools had to increase the number of classes they offered at each grade level to accommodate the same number of students. To do so, they needed to hire more teachers. Because becoming a teacher through traditional means requires at least four years of pre-service training, the number of teachers available fell short of meeting the needs of all schools. Suburban districts with greater resources filled their spots by recruiting teachers from urban districts, leaving those schools woefully understaffed. Proponents of the new law had failed to anticipate this unfortunate outcome of the change in class size.

By showing the potential behavior over time of multiple scenarios based on specific inputs, computer modeling offers policymakers and administrators the ability to visualize the long-term effects of specific decisions before those decisions are implemented. We can also use models to identify unexpected interactions between system components; ask “what if questions about changes in system parameters; run no-cost experiments that compress time and space; and reflect on, expose, test, and improve the mental models upon which we rely to make decisions about difficult problems. Thus, computer modeling could allow school-system leaders to make more effective decisions by building their understanding of long-term consequences of resource decisions in a complex environment.

Evaluating Professional Development Programs

To illustrate how a district can use computer modeling to analyze its options, I have created a simulation that explores the impact of professional development programs for teachers. Many school districts have responded to the call for better educational performance by implementing a standards-based curriculum. They offer professional development workshops to increase teachers’ ability to communicate this new curriculum to their students. The workshops are often formatted as multi-week summer programs.

Research has shown that teachers can learn to communicate the new curriculum through professional development training, so the question for a district is not whether summer workshops can build capacity, but whether they can do so for a critical mass of teachers in a reasonable time period. What factors play a role in this issue? Which workshops are most effective? What are the costs associated with this form of professional development? These questions are amenable to modeling because we can determine quantitative values for most of the important variables — such as the number of teachers in training and the turnover rate of teachers — and reasonable estimates for the qualitative variables — such as the effectiveness of the workshops and the relationship between the length of the workshop and the willingness of teachers to enroll in it.

I followed these steps to build the model:

1. Define the teacher stocks. All the teachers in the district fall into three stocks: Those who are not familiar with the standards; those who are attending a workshop to learn about the standards; and those who are familiar with the standards.

2. Establish the flow between stocks. Teachers who aren’t familiar with the standards can take a workshop to gain familiarity; teachers in the workshop may become familiar with the standards and move into the “familiar” stock or may not gain much from the workshop and return to the “unfamiliar” stock; and both “familiar” and “unfamiliar” teachers may leave the system each year.

3. Identify and assign values to the important system parameters and variables.

4. Incorporate funding components.

The model is based on the following assumptions:

• The number of teachers in the system remains constant at 10,000, and at the starting point, 10 percent of the teachers are already familiar with the standards-based curriculum. Workshops vary in length from one day to five weeks.
• Ten percent of the teachers leave and are replaced each year (with 10 percent of new teachers entering in the “familiar” stage), and the rate at which teachers leave the system is higher for teachers in the “unfamiliar” pool than in the “familiar” pool.
• In the baseline simulation, 1,000 teachers participate in the three-week workshop; this number can vary up or down by a factor of three.
• Fewer teachers participate in longer workshops, more in shorter ones. However, longer workshops are more effective. The initial success rate for teachers reaching the “familiar-with-standards” stage in a three-week workshop is 30 percent. This base rate increases linearly over time as more and more teachers (those for whom training was not effective the first time) retake the workshop.
• There are 25 teachers in each workshop. The cost of the workshop includes a stipend of $300/week/ teacher for each of 25 participating teachers and an additional cost of $2,500/week for the instructor, supplies, and space.

“Modeling Professional Development” illustrates the model’s basic features.

Analyzing Results

The simulation yields several non-intuitive results, the most important being that these workshops alone cannot adequately deal with the problem of building the necessary capacity in the teacher workforce. Even after 10 years of providing three-week workshops, only 52 percent of the teachers are skilled in presenting a standards-based curriculum — and this number includes teachers who were capable before they enrolled in the workshops. The results clearly show that the workshops do not produce a critical mass of teachers with the desired capabilities in a reasonable amount of time.

Another unexpected result of this analysis is that the five-week workshops result in the largest number of trained teachers over a 10-year period, even though the smallest number of teachers enrolls in them. Holding all else constant, approximately 5,200 teachers achieve the desired level of ability after participating in a five-week workshop, while only about 2,800 teachers reach this stage through one week workshops. The longer workshop is also the most cost-effective per teacher trained: $2,300 per teacher for a five-week workshop; $2,635 for a three-week workshop; and $3,100 for a one-week workshop.

We can generalize this kind of model to other areas of professional development, because the results are independent of the workshop content. Administrators have access to the quantitative data for their district (such as number of teachers in the system, distribution by length of service, teacher leaving rate, funding available for workshops) and can reasonably estimate values for the qualitative variables (such as percent of teachers who require specific professional development, workshop effectiveness, relationship of workshop length to teacher resistance and workshop effectiveness) from prior experience. Plugging these numbers into a computer simulation would give them a general tool for predicting the impact of a summer workshop on professional development in any content area.

Similar models could let stakeholders examine other questions, such as the impact of rationing workshop participation depending on teachers’ average time of service in the system.

Should administrators concentrate on those who will remain in the system longest, that is, younger teachers? Or is there value in offering training opportunities to experienced teachers, who can serve as opinion leaders in changing the system’s culture? This analysis could also be incorporated into an expanded model to include the use of mentors and school and web-based professional development. By exploring these variables as well, districts might come upon a formula for producing a multi-component professional development system with the capacity to bring a critical mass of teachers up to speed on new curriculum requirements in an acceptable time period.

As I hope I’ve shown here, computer modeling offers a valuable planning and decision-support tool for school districts. This approach permits “no-risk” analysis of competing policy choices and resource allocations and, while it does not offer definitive answers, it can help school-system leaders understand the impact of their decisions and guide them toward making better-informed allocations of scarce resources.

Daniel D. Burke, Ph. D., has a broad understanding of K-graduate educational systems. As deputy director for education, the CNA Corporation (CNAC), he leads the research and analysis activities of CNAC’s public education group. Before joining CNAC, Dan was a researcher in molecular biology and produced an extensive record of curriculum innovations. He also played an important role in the National Science Foundation’s K-12 education reform programs.

The post Making Better School Policy Decisions Using Computer Modeling appeared first on The Systems Thinker.

We’ll use paper and pencil (the only items you’ll have to pay for, assuming you own a fairly up-to-date computer) to sketch out the initial model, a simulator called SimPack to code the model, a compiler called gcc to compile the model into an executable program, gnuplot for the graphics, and Dia for producing stock and flow diagrams and other documentation (see “Shoestring Resources”).

The Process, Briefly

Let’s start with model building. I still find paper and pencil the best way to get started — it’s easy to use, there is no syntax checking to constrain creativity, and editing is quick and satisfying — erase small mistakes or, for bigger changes, ball up the paper and throw it in the recycling bin.

Once you have sketched a model, it’s time to convert it into something the computer can understand. While it’s not too hard to write a basic system dynamics simulator from scratch, SimPack, a collection of simulation programs produced by Paul Fishwick at the University of Florida, makes life easier.

SimPack provides two system dynamics simulator programs that you’ll find in the constraint/differential/integrate subdirectory of SimPack: conte.c for Euler integration and contrk.c for Runge-Kutta integration. You’ll need to enter your model equations as statements in the C programming language.

Once you’ve coded the model in C, you’ll need to compile it to turn the C program into an executable program for your computer. If you’re running on Linux or Mac OS X, you probably have ready access to the compiler gcc. If you’re running Windows, Cygwin offers a free environment that includes gcc and other tools you might use.

After you compile and run the program, you’ll get a file full of numbers, representing the value of each variable of interest at each point of simulated time. Gnuplot can graph that data.

What’s left? Perhaps you want to communicate your thinking about the relationship between stocks and flows in the system to others. Use Dia to create traditional or creative stock and flow diagrams. Both Dia and gnuplot can produce results suitable for casual viewing on the screen, incorporation into a Web site, or publication in a report or journal.

Taking the Next Step

If you haven’t programmed much, you may be a bit overwhelmed. Move forward in small steps. Start by installing and exploring gnuplot and Dia; you’ll likely find many uses for them, including plotting data and drawing diagrams.

Then install Cygwin, if you’re on Windows. While it can be a massive download, all you really need is the basic installation plus gcc. You’ll probably want the man (manual) pages, too.

Even if you decide you prefer to use a commercial simulator, you might find that some of these tools can augment your normal processes. For example, I’ve used gnuplot to produce publication-quality graphics from data generated using a commercial simulator.

Who knows? You might enjoy systems thinking on a shoestring!

Bill Harris (bill_harris@facilitatedsystems.com) is principal and founder of Facilitated Systems, a company dedicated to helping organizations address complex problems, work more productively in meetings and groups, and learn more effectively from experience.

this is a continuation…

Making It Concrete

The Model. To illustrate, I’ll carry a simple model about the spread of infection through a population through the entire process. If you need more information, refer to “Shoestring Resources” in the main article.

The SI model is a simple model of disease. It divides a population into two groups: Susceptible (S) and Infected (I) people (thus the SI moniker). There’s only one flow, from S to I. The number of susceptibles becoming infected per day, ipd, is the product of the number of susceptible people S, the number of contacts a susceptible person has per day c, the probability of any one contact being with an infectious person, and the probability of getting infected from contact with an infected person p:

ipd = S * c * (I / (S + I)) * p

For more information, see chapter 9 of John Sterman’s Business Dynamics (McGraw-Hill/Irwin, 2000).

To create the simulation program, open the SimPack file conte.c in a text editor (e.g., Notepad; I use GNU Emacs) and save it as si.c in a convenient directory. Then set the number of initial susceptibles S to 9999, the number of initial infecteds I to 1, the probability of getting infected from one contact p to 10% (0.1), and the number of contacts per day c to 4 by editing the initialization function init_conditions in si.c:

init_conditions() { out[1] = 9999.0; /* Susceptibles */ out[2] = 1.0; /* Infecteds */ p = 0.1; c = 4.0 time = 0.0; delta_time = 0.125; } Note that we don’t refer to S and I by those names; rather, we use elements of the array in and out to represent values of stocks. out[1] holds S, the number of susceptible people just calculated by the model, while in[1] holds the value of S to be used at the start of the next iteration. The function init_conditions also sets the initial time to 0 and the simulation time increment (often referred to as DT in system dynamics models) to 0.125.

We need one more SimPack function to calculate the results:

state() { /* Calculate flows */ ipd = out[1] * c * (out[2] / (out[1] + out[2])) * p; /* Update stocks */ /* Susceptibles */ in[1] = 0.0 – ipd; /* Infecteds */ in[2] = ipd – 0.0; }

We have a bit of housekeeping to attend to, as well; we’ll need this declaration at the start of the program:

double d, c, ipd;

In the function main, we’ll set the program to run when time (in days) is less than 50.0, and we’ll add a statement to print out all the results. To speed experimentation, I print all the stocks and flows and let gnuplot select the values to plot:

printf( “%f %f %f %f n”,time, out[1], out[2], ipd);

If you’re not a C programmer, some of this may look like gobbledygook; you can just try it, or you can read some of the “Shoestring Resources.”

Compile this program by typing two simple commands in a Cygwin command shell window:

gcc –o si.exe si.c ./si.exe > si.dat

The first line compiles the program si.c, creating si.exe, and the second runs si.exe, putting its results in the file si.dat. Si.dat has four columns: the time, S, I, and ipd. You can look at si.dat with your favorite text editor to see the results. If you’ve made it this far, pat yourself on the back; you’ve created a system dynamics model!

Plotting the Output

Few of us are satisfied looking at such a long list of numbers. Start gnuplot, and enter the command

cd “/Documents and Settings/My Documents/My Name/”

if that’s where you put your program. Then the command

plot “si.dat” using 1:2

will plot the number of susceptible people over time.

plot “si.dat” using 1:3

will plot the number of infectious people over time.

plot “si.dat” using 2:4

will create a phase plot showing the number of people getting sick per day for different values of the susceptible population. In each case, the numbers in the plot statement refer to columns in the data file you wrote when you ran si.exe.

Gnuplot can plot multiple graphs on the same sheet and can format the output for the screen or publication.

Making Life Simpler and More Powerful

As you advance, you’ll likely need a table function to represent nonlinearities. With basic C programming skills, it isn’t hard to create. You might also want to be able to give your model new parameter values without recompiling the program; the getopt C library function can help. If you want to reduce the typing required for this work, check out gnuplot mode for the Emacs text editor in the “Shoestring Resources” section of the main article.

Stock and Flow Diagrams. Those of us accustomed to using commercial simulators expect to see computer-generated stock and flow diagrams. We’ll use Dia. While you can draw straightforward stock and flow diagrams easily with Dia, you can also exercise a bit of creativity, using a symbol for stocks that suggests the type of object being accumulated and a symbol for flows that matches the symbol for stocks.

Now that you’ve seen that you can do system dynamics on a shoestring, remember all the good practices you’ve learned elsewhere. Happy modeling!

Bill Harris (bill_harris@facilitatedsystems.com) is principal and founder of Facilitated Systems, a company dedicated to helping organizations address complex problems, work more productively in meetings and groups, and learn more effectively from experience.

The post System Dynamics on a Shoestring appeared first on The Systems Thinker.

Nevertheless, despite the progress illustrated by these and other cases, the forces of unsustainable growth and resource exploitation have continued to compound. So the release of Limits to Growth: The 30-Year Update (Chelsea Green, 2004) by Donella Meadows, Jorgen Randers, and Dennis Meadows comes at an important time. For newcomers to the systems approach, the 30-Year Update presents the logic of overshoot and collapse and emphasizes the urgent need for sustainability without dwelling too much on the mechanics of the methodology (see “Key Terms”). At the same time, those already inclined to see things from a systems perspective not only have their mental models reinforced and refined, but also have a series of cogent examples to draw upon when spreading the gospel of sustainable development.

Systems and Growth

Three themes emerge in the book: background on systems and the mechanics of growth; the introduction of a formal computer model, known as World3, and some of the scenarios that it produces; and implications and recommendations (see “The World3 Model” on p. 9). Throughout the volume, but particularly in the first three chapters, the authors explain the basic laws of system structure and behavior with a lucidity that comes from decades devoted to the dissemination of these concepts.

Though usually considered “best practice,” it is not common to come across computer modelers who clearly communicate the purpose of their model and its associated boundaries; that is, the question the model was intended to address and those for which it loses its ability to provide meaningful insight. So it is a treat (for modeling geeks, anyway) to have the authors devote several pages to just these concerns in the course of their introduction to the World3 model. The central question they mean to address is: Faced with the possibility of global collapse, what actions can we take that will make a difference and lead to a sustainable future? It is clear that this is a model whose primary purpose is to help us think, not to provide the answer. In the course of laying out their model’s purpose, the authors make one of the best cases for “modeling for learning” that I have come across.

The authors introduce a variety of potential actions into the World3 model, at first, one-by-one, then in logically consistent groups. Each run, or scenario, provides insight into how that potential action or group of actions might affect the course of future events. In this way, Meadows, Randers, and Meadows are able to prioritize potential actions in order to come up with the set that offers the greatest opportunity for avoiding the worst consequences of collapse.

Recommendations for Action

In the end, World3 does provide an answer. Of the various assumptions tested and given the boundary conditions of the model, we can still make a transition to a sustainable global society if people around the world immediately take the following actions:

1. Stabilize the population
2. Stabilize industrial output per person
3. Add technologies to:
   • Abate pollution
   • Conserve resources
   • Increase land yield
   • Protect agricultural land

The bad news is that we have already begun to experience symptoms of overshoot—water tables are dropping rapidly in some areas and incidents of coral bleaching have risen but two of the most urgent signals. The good news is that, as the authors’ account of the ozone story demonstrates, once the global community sees the clear need for change, change can come about quickly.

According to the authors, people respond to signals that a system has overshot its limit in one of three ways:

1. Deny, disguise, or confuse the signal that the system is sending
2. Relax the limits through technological or economic action
3. Change the system structure Certain elements of society are

stuck in response 1, regardless of the growing mountain of evidence calling for action. We see this mindset in the refusal by some politicians to acknowledge the science behind global warming. Others place their faith solely in the market and/or technology, even though the price would be extremely high if the market system and new technologies fail to save the day. The only truly effective response is to change the system structure, the sooner the better.

This was the core message of the original Limits to Growth. And while that message became a part of society’s broad environmental consciousness, the warning went largely unheeded. The result is that the party’s nearly over, and we need to figure out how to minimize the hangover.

Restructuring Society

Because structure determines behavior, the highest-leverage approach to these problems is to change the underlying structures that have created them, such as farming techniques, forest management policy, end-user attitudes toward consumption, recycling, and reuse, and legislation regulating pollution. So how do we go about restructuring the global system? The authors share the tools they have found to be useful: rational analysis, data gathering, systems thinking, computer modeling, and clear communication.

Notice that these tools really have more to do with making the case for change than they do with enacting change that has been agreed upon. That is, they are exceptionally useful for helping lawmakers understand the need for change and explaining to corporate decision makers the logic behind a shift. These tools can even guide the overall implementation of a change effort. But once the case has been made, the day-to-day activities can look somewhat like business as usual: rewriting laws, redesigning products and processes, reorganizing departments, and so forth. The difference is that the guidance offered by these tools means the change is less like rearranging deck chairs on the Titanic and more like fixing the hole in the ship.

The 30-Year Update is compelling: We have already overshot the planet’s carrying capacity on numerous vital resources. Whether humanity is successful in avoiding the most disastrous effects of collapse will be determined in part by the actions taken by people across our society and planet. Unfortunately, politicians and other leaders often seem to be linear and “black-and-white” thinkers. Navigating the turbulence ahead will require decision making that appreciates non-linearities and shades of grey. The 30-Year Update will bring some to the sustainability camp. But more important, it will inspire others—those with the necessary perspective—to take action. There’s no time to waste.

Gregory Hennessy is honored to have worked with Dennis Meadows on several occasions and to have met the late Dana Meadows once.

The post Scenarios of the Future: The Urgent Case for Sustainability appeared first on The Systems Thinker.

I strongly believe that stock and flow diagrams offer a deeper understanding of a system than do causal loop diagrams. Nevertheless, in the past, I found it difficult to get more than a small handful of clients to develop the facility to build them. Despite its obvious benefits, the rigor of the stock and flow language comes at a price—it is more difficult to learn. While companies that create simulation software have made enormous advances in their products, vastly simplifying the model-building process, we still have a long way to go in learning how to help people develop the facility to create even simple models.

High Performance Systems, the creators of the ithink® software, have stuck firmly to their belief that a true understanding of the dynamics of any system requires an appreciation of the underlying stock and flow structure. For that reason, their software does not provide the facility to build CLDs. The best they offer are “Loop Pads,” which they describe as “. . . simple pictures that identify the cause and effect processes that work to generate dynamic behavior patterns.” To display these pictures, you have to build the stock and flow model first.

The challenge is to find more effective ways of helping clients develop an understanding of the structural dynamics of the system they are studying, while acknowledging that they usually find CLDs an easier place to start than stock and flow diagrams. To that end, I have developed an approach to model building that uses ithink in a slightly unorthodox way to start clients at a relatively easy place and move them quickly to a more sophisticated understanding of a given system using stocks and flows. Paradoxically, this technique capitalizes on the software’s unwillingness to let users draw CLDs.

From Feedback Loops . . .

To follow this process, you must use version 6.0 of the ithink software, which allows you to minimize the size of the converter icon. Start by changing the defaults to set the converters to small. Doing so lets you use the converters as you would the variables in a CLD. Then use the text box facility, which is one of the objects on the menu bar, to create the polarity signs (, “+” or “-,” which correspond to “s” and “o”). For example:

So far so good. Up to this point, we are simply building a CLD in which “resources allocated to process improvement” are influenced by current “performance” and “desired performance.” However, when we try to create a causal link between “resources allocated to process improvement” and “process errors,” an error message appears indicating that such an action would create a circular connection.

The nature of the mathematics that underlies the stock and flow language means that the software is unable to calculate the value of any converter or flow when they loop back on themselves. As the help files state: “In drawing connector linkages, you may encounter an alert which tells you that circular connections are not allowed. Mechanically, this alert means that you have attempted to create a chain of converters or flows, such that one converter or flow ultimately depends upon itself. The software cannot resolve the resultant simultaneous equations.”

. . . to Stocks and Flows

To get past this barrier, we must create at least one stock somewhere in the loop. This process forces us to look more closely at the structure of the loop we are creating and identify one or more stocks. Every feedback loop has an accumulation—it is this accumulation that generates the feedback dynamics. In this example, the issue for the team was how actual performance levels drove “resource allocation to process improvement.” With this in mind, we can now make a simple modification to the CLD by converting the variable “performance” into a stock.

We are now a step closer to gaining a deeper understanding of the feedback processes involved in this structure. We have done so, however, by beginning with a process that clients are familiar and comfortable with and then moving to a structural understanding through one simple step. How we develop the model from this point forward depends on your goals. We could stay with this loop and simply develop the stock and flow structure for each variable.

Going into Greater Detail

We also might want to explore a certain part of the structure in greater detail. For example, we might be interested in the dynamics involved in a process-improvement program. In this case, the team realized that the resource allocation decisions were not only determined by actual performance but by the gap between actual and desired performance:

On the other hand, we may want to develop a loop to explore the impact of process quality. One possibility could be:

Once again, when we try to close the loop by connecting “investing in process quality” with “process quality,” we will receive an error message that circular connections are not allowed. How we respond to this message depends on what we are trying to understand with the model. If we want to examine the financial implications in more detail, we could begin to unravel the structure underlying the variable called “profits.” For example:

The key point is that we always anchor the model development process in something the client is familiar and comfortable with—the development of CLDs. We then force the software to highlight a logical error to provide a stepping stone to unfolding the stock and flow structure. I have found that, using this technique, more clients are able to develop an ability to create their own stock and flow models than before. Prior to using this approach, I found that clients viewed CLDs and stock and flow diagrams as separate and distinct languages. Since I’ve implemented this process, I have noticed that they have begun to see the similarities, rather than the differences, between the two. As a result people are less mystified when working with stocks and flows, seeing them embedded in the feedback loops of CLDs.

David Rees is the director of High Performance Learning Systems, a consultancy firm specializing in applying systems thinking principles and tools in public and private sector organizations. He is also a research fellow at the Centre for the Design of Innovative Systems at UNITEC in Auckland, New Zealand.

The post Anchoring Model Development in Causal Loop Diagrams appeared first on The Systems Thinker.

For example, in August, White House Science Adviser John Marburger briefed a Senate panel on climate change, saying, “We know we have to make very large changes if this turns out to be a problem. The consequences of human-induced global warming could be quite severe.” Yet at the same briefing, the administration stood behind its “wait and see” policy, articulated by President Bush in February: we should only “slow the growth of greenhouse gas emissions,

and—as the science justifies—stop, and then reverse that growth.” Climate change could be severe, and yet we should wait before acting. How can U. S. leadership reconcile these two seemingly contradictory statements?

Climate As a Delayed System

MIT professor John Sterman and Harvard’s Linda Booth Sweeney explain that this “wait and see” approach makes sense if you believe the world’s climate to be a nondelayed, responsive system in which a change in human activity has an immediate effect. Their recent experiments confirm that many highly competent people instinctively see climate as behaving this way. Most of their business-school student subjects thought that if humans reduced emissions of GHGs, the storehouse of those gases in the atmosphere would promptly decline and global temperature would follow.

CO2 IN THE ATMOSPHERE

Carbon dioxide (CO2) in the atmosphere, the primary heat-trapping gas, can be thought of as a stock or accumulation. The stock is now at its highest level in almost half a million years. To reduce the ecological and economic changes from producing global warming, we need to lower the level of the stock by reducing the inflow (CO2 emission rate) to less than the outflow (Net CO2 removal rate).

The inflow is currently about double the outflow. If we were to reduce the inflow by, say, 20 percent, it would still be greater than the outflow and the level of CO2 would continue to rise, albeit at a slower rate. No wonder the students predicted incorrectly—it is counterintuitive to think that the CO2 emission rate can go down while the level of CO2 in the atmosphere continues to go up! Nevertheless, it’s true. If the removal rate were constant, we would need to cut the inflow rate by more than 50 percent to finally begin to lower the CO2 level. The bottom line is this: If we, as the Bush administration says, “slow the growth of greenhouse gas emissions, and—as the science justifies—stop, and then reverse that growth,” it could still take many decades for levels of CO2 in the atmosphere to decline.

A Robust Plan

If we believe Mr. Marburger that the effects of climate change could be “quite severe,” we need a robust plan.

The Bush administration’s plan would work well if the climate had short delays. But the plan is not robust when managing a slow-responding system like our climate; the possibility of negative, irreversible effects from waiting are too high.

We see two important steps. 1. Teach ourselves the basic mechanics of our climate.

Prudence leads us to act now to educate ourselves about the dynamics of the climate system and to address the source of the problem with practical measures. If Sterman and Booth Sweeney are right, our generally poor intuition about the climate enables many of us to accept a “wait and see” approach. For our society to engage in an effective public discourse about global warming, we need to ground ourselves in the basics of the climate inflows, levels, and outflows. Then we can evaluate the impact of national-level proposals and really understand the challenge that we face in stabilizing the climate.

2. Act now to reduce GHG emissions. The best way to deal with a slow-moving system in which we know we will eventually need to make a change is to begin the change as early as possible. We need not initially focus on retooling our entire industrial base; we can begin with the significant reduction in emissions available through hybrid cars, better designed industrial motors, fuel cells, and renewable energy production. Such improvements could come at relatively low cost, improve our short-term economic vitality, and reduce energy dependence.

How can we get the process started? We suggest designing incentives and rewards that would unleash people’s tremendous capacity for innovation. A similar outpouring of new ideas came as a result of the ban on CFCs to prevent additional damage to the ozone layer. Let’s introduce similar mechanisms into our market system to encourage technological and behavioral changes for reducing GHG emissions.

Prudence leads us to act now to educate ourselves about the dynamics of the climate system and to address the source of the problem with practical measures. These actions will not be easy—technologically, culturally, or politically. But they are certainly easier than navigating a barge while pretending it will handle like a Ferrari.

The post We Can’t Afford to “Wait and See” on Climate Changes appeared first on The Systems Thinker.

This long and broad view is in direct opposition to the “quick-fix” mentality that increasingly dominates our complex world. Perhaps the reliance on “band aids” results from our economic system, in which managers focus on short-term results to keep stock prices and option values high, and shareholders care more about quarterly returns than long-term corporate health (accentuated by technology that provides instant access to massive amounts of data). Perhaps it comes from our political system, in which politicians invest in symptomatic rather than fundamental solutions—which take longer to show results than the person’s term in office—in order to ensure reelection. Or perhaps it is an outcome of our educational system, which fails to expose people to the basic ways in which feedback processes work in the world.

Whatever the reason, despite the promise of systems thinking, its impact has been surprisingly limited. But I fear that, unless a critical mass of people and organizations adopt a systems view, our organizations will continue to fall short of their potential. Even worse, the dire consequences of non-systemic approaches to issues such as global terrorism, the environment, and poverty will threaten the world for us, our children, and future generations. By offering the perspectives that follow, I hope to widen the circle of systems thinkers by attracting newcomers and convincing experts to stay the course.

The Systems Thinking Difference

Let me start with a personal story. As a student, I attended a lecture by Norbert Weiner, the famed mathematician. He discussed a key project in which scientists of the day were working—unsuccessfully—to get computers to translate text from one language to another.

Weiner identified a possible breakthrough in the project: The goal should be to create a system for excellent translation by including a computer component to perform routine elements and a human component to handle the non-routine tasks. Together, they could elegantly and affordably achieve the overall goal. The fundamental idea of a system as an entity that was different from its components—and not merely the sum of the components—was, to me, original, new, and powerful.

Over the years, I have heard many people say that the simple act of thinking systems rather than components, the whole rather than the pieces, enables them to better understand why things behave as they do and take more effective actions. I have seen, for example, executives who are dealing with a critical product issue come to the realization that the answer is not in making marketing or manufacturing work better, but in improving the quality of interaction and influence between the two functions. The notion of recognizing the interactions among component parts as critical to the system’s performance leads people to accept the system as the major determinant of the behaviors and events that occur.

Once we can see the whole (the system) as something different from its parts (the components), it isn’t too far a leap to accept Deming’s observation that optimizing a system requires sub-optimizing its components. This idea is profoundly paradoxical. It says that functional excellence will not guarantee overall success and that working “across the stovepipes” provides the greatest possibility for superior performance. Bridging the gap between functions requires compromises from each department for the benefit of the firm as a whole. My observation is that once people “get” the concept of systems, they become sensitive to the harm the stovepipe mentality can bring, and they open themselves to seeing linkages among the pieces that may be important, even in areas beyond their control. Because talking across stovepipes is not easy, the mastery of dialogue, skillful conversation, and concepts such as the ladder of inference—all part of today’s organizational learning focus are essential to fundamental and sustainable performance improvement. A dozen years ago, I scoffed at such things as too soft and fuzzy. Now I am convinced that these tools play a critical role in improving systems. (Of course, Peter Senge already understood this point in 1990 when he popularized systems thinking and integrated it with team learning and other skills in his surprise best-selling management book, The Fifth Discipline: The Art and Practice of the Learning Organization!)

The ability to visually represent the interrelationships among the components of a system through different kinds of diagrams represents another benefit of systems thinking. These “maps” reveal the cause and effect linkages thought to underlie behaviors by depicting the “system behind the story.” Causal loop diagrams are especially effective in displaying the feedback processes at play. By recognizing the behaviors associated with each of the two kinds of loops (balancing and reinforcing) and through the process of collaborating on creating the diagrams, people are able to reach important and sometimes profound insights. Stock and flow diagrams are especially effective in displaying the dynamics among accumulations or stocks (such as backlog, inventory, or morale) and the flows (such as orders, shipments, and successes) that increase or decrease them. By identifying stocks and flows, we gain knowledge about a system’s behavior and take a step toward building simulation models. We know that the “map is not the terrain,” but maps of structure predictably add insight to our ability to better know the real terrain by giving us a shared view of its complexity.

System archetypes also provide a strong basis for learning about systems. Archetypes are a set of relatively simple structures that have been observed to occur again and again in social systems. These structures typically consist of two or three causal loops and have names like “Fixes That Fail” (the story of unintended consequences), “Shifting the Burden” (the story of addiction), “Limits to Growth” (the story of resource depletion), and “Escalation” (the story of violence and war).

It has been interesting to see the rapidity with which relative newcomers can relate to an archetype and apply it to their own experiences. In workshops, the energy and insights that emerge from archetype examples are often startling. More than once, I have heard someone say that the understanding of a single archetype changed his or her life!

Breadth and Depth

The most rigorous end of this spectrum is computer simulation, which stems from the breakthrough work of Jay W. Forrester at MIT in the 1950s. He brought the first application of engineering control theory to social systems, taking advantage of advances in computer technology for simulating non-linear systems. His 1963 book, Industrial Dynamics, provided the initial codification of the ideas, tools, and learnings of the nascent field and remains a classic today. (It was, by the way, my privilege to be a research assistant in Forrester’s group during the early phase of the field. That was how I was hooked!) Simulation enables users to view the system’s behavior in action and to experiment with various scenarios. These are very powerful capabilities.

All approaches that stem from a fundamental understanding of systems—whether broad or deep—can add value by offering insights far beyond traditional linear thinking.

Efforts involving simulation models around specific organizational issues have had a positive impact on corporate decisions and strategy assessment in a number of cases. However, building such simulations takes enormous time, money, and expertise. In addition, decision-makers who don’t fully understand the model may be uncomfortable changing policies based on its outcomes.

A broader or at least more visible source of impact, I think, has come from “models for learning” developed in academic and other non-corporate environments around major social issues and generic problem behaviors. Limits to Growth (by Donella Meadows et al.) and World Dynamics (by Jay For rester) were based on simulations that explore the extent to which our planet’s resources can support the rapid growth of human population and industrial activity. Forrester’s Urban Dynamics deals with the system structure underlying the growth and decay of cities. These bodies of work have created a widespread awareness—with significant controversy—of the critical environmental and social issues facing humankind by demonstrating the potentially catastrophic trends that can result from certain systemic structures.

As another example, my own work on the dynamics of corporate growth (a master’s thesis also published as a monograph) outlines how balance among functional decisions in a growing company can be more important than specific functional expertise. The study used computer simulations to show how a company, by its own actions and with inadequate understanding of its systemic structure, could easily fail even though its market was virtually infinite. It demonstrated how an enormous range of behaviors, from wildly successful growth to stagnation to collapse, depended solely on the firm’s internal decisions! As the cartoon character Pogo said, “We found the enemy and it is us.” I contend, though it is impossible to prove, that this work had a positive impact on the company that sponsored it (which was highly successful for more than 20 years afterward).

The very breadth of the systems arena has created some barriers that I believe have slowed acceptance of the field. From a systems perspective, the obstacles I have seen relate to our own stovepipes, represented by different approaches such as simulation, causal loop analysis, stock and flow diagrams and the like. When practitioners in particular areas imply that their approach is the only valid one, the credibility of the whole spectrum of activities suffers. Here, I have tried to convey that all approaches that stem from a fundamental understanding of systems—whether broad or deep—can add value by offering insights far beyond traditional linear thinking. As in most systems, the right balance among the components is the path to a stronger whole.

Looking Ahead

In closing, my objective in this article has been to present my observations of the compelling potential for creating a better world through applying systems concepts and tools to our own circumstances and issues. Thinking systemically can change lives, improve businesses, help economies, and maybe even save the planet. Equally important, the broad range of approaches for application provides great accessibility. Opportunities for demonstrating the impact of systems thinking should be embraced, wherever they happen, and the diversity of approaches should be used to full advantage! I hope I have provided some incentive for doing just that.

The post Taking a Systems View: A Reflection appeared first on The Systems Thinker.

Systems thinking is a perspective and a set of conceptual tools that enables us to understand the structure and predict the behavior of complex systems. While already commonplace in engineering and in business, the use of systems thinking in personal health is less widely adopted. Yet health is precisely the setting where dynamic complexity is most problematic and where the stakes are highest. Thinking in Circles About Obesity: Applying Systems Thinking to Weight Management (Springer, 2009), aims to fill this gap. The book applies systems thinking to personal health in a form that’s accessible to the general reader, with the hope that it will have a profound influence on how ordinary people think about and manage their health and well-being.

Systems Thinking . . . and Thinking About Systems

The great shock of 20th-century science has been that systems cannot be understood by analysis alone. While the performance of any system—whether it is an oil refinery, an economy, or the human body—obviously depends on the performance of its parts, it is never equal to the sum of the actions of its parts taken separately. Rather, it is a function of their interactions. Breaking a system into its component pieces and studying the pieces separately is, thus, an inadequate way to understand the whole.

Human weight and energy regulation provide a good case in point. They are parts of a complex psychobiological system that involves the behavioral act of eating, the processes of ingestion and assimilation of food, the storage and utilization of energy, as well as interactions with the external environment (cultural and physical). All these various factors are interconnected, pushing on each other and being pushed on in return. Appetite shapes body weight, and body weight influences appetite. Weight reflects activity levels (which are also shaped by the socioeconomic environment), and activity levels reflect weight. And on and on (see “Learning to Squint”).

Understandably, putting systems pieces back together and recognizing the interactions between them can appear slippery and elusive. So much will be going on, and some of the things that are going on will cause still other things to go on. Making sense of it all becomes a daunting task. It’s why one of the most important and potentially most empowering insights to come from the field of systems thinking is that certain patterns of structure recur again and again in many systems – whether physical, biological, or social—revealing an elegant simplicity underlying the complexity of systems (Peter Senge, The Fifth Discipline: The Art & Practice of the Learning Organization, Doubleday/Currency, 1990). And it’s why learning to recognize these recurring building blocks is a powerful conceptual leverage that allows us to see through complexity into the underlying structures that drive system behavior (or misbehavior).

Stock and Flow Basics

All dynamic systems—the human body being a perfect example—can be modeled as stocks and rates of flow threaded together by information feedback loops. Stocks and flows constitute the two fundamentally different processes— accumulation and flow—that characterize how reality works and how systems change. You’ll find these stock and flow structures in systems of all kinds. A familiar “plumbing” example is that of water in a bathtub. A bathtub is a (hydraulic) stock whose level changes as a function of the rates of water flowing in and draining out. And just like a bathtub, the level of energy stored in the human body constitutes a stock (primarily of fat), with food intake as its inflow rate and energy expenditure as its outflow rate.

Stock and flow structures are not limited to physical “stuff,” however. For example, experimental research is demonstrating that the human capacity for self-regulation—a critical faculty for successful weight regulation—is a limited resource. In a manner analogous to the storage and depletion of physical energy, the human capacity for self-regulation can be conceptualized as a reservoir—or stock—that is consumed and replenished with the exertion of self-control and rest (M. Muraven, D. M. Tice, and R. F. Baumeister, “Self-control as a limited resource: Regulatory depletion patterns,” Journal of Personality and Social Psychology, 74, 1998).

Behavior and Physiology Interactions

The strength of the systems approach lies in its capacity to integrate variables that otherwise would be isolated from each other. In the case of human weight and energy regulation, it allows us, for example, to examine (and better understand) the feedback interactions between the physiological and the behavioral.

The diagram “Dieting Regulation System” integrates the two sets of stocks and flows in the human psychobiological system for feeding regulation discussed above: (1) the stock of human self-control, with its replenishment and exertion rates; and (2) the body’s energy stock, with food intake as its inflow rate and energy expenditure as its outflow rate. As we shall see, these two sets of processes are not isolated phenomena. Indeed, it is the (mismanaged) interaction between these two stock and flow systems that gives rise to the weight-cycling dynamic—the “lose-gain” phenomenon widespread among and dreaded by dieters.

When the two stock and flow processes are combined into an integrated whole (see “Dieting Regulation System”), what we end up with is one of the classic archetypes for oscillatory behavior: that of two stocks (resources) interacting with one another such that the rise in one drains the other and vice versa.

DIETING REGULATION SYSTEM

This diagram integrates two sets of stocks and flows in the human psychobiological system for feeding regulation: (1) the stock of human self-control, with its replenishment and exertion rates; and (2) the body’s energy stock, with food intake as its inflow rate and energy expenditure as its outflow rate. The interaction between these two systems gives rise to the weight-cycling dynamic widespread among and dreaded by dieters.

Specifically, in this integrated psychobiological system for human feeding regulation, “Self-Control Strength” (which we can designate as stock 1) affects adherence to the diet and, hence, the regulation of the food intake rate into stock 2, “Weight.” This regulatory function is not a free lunch—constraining food intake to decrease and/or maintain the weight stock at a certain level requires effort which, in turn, consumes self-control strength. This means that the state of the body-weight stock (stock 2) regulates the exertion rate (the outflow rate) of the self-control stock. Stock 1 acts as a catalyst for the inflow rate to stock 2, and, likewise, stock 2 returns the favor and acts as a catalyst for the outflow from stock 1.

For any such stock and flow system, if and how fast total depletion of a stock occurs depends on the initial size of the stock and the magnitude of the imbalance between the inflow and outflow. In the case of self-regulation, we know from personal experience that most people are capable of exerting modest levels of self-control and sustaining the effort day in and day out. This suggests that the amount of self-control needed for our daily social functioning—for example, stopping at a stop sign, standing in line even when in a hurry, holding our tempers, and so forth—is low enough that normal periods of rest can compensate for the moderate depletion rate.

But what about when we have to (or choose to) exert more-than-modest levels of self-control? Resisting stronger impulses, such as not eating even when persistently hungry, obviously requires more self-control than resisting less appealing temptations or weaker impulses, such as speeding on the highway. Would normal rest be enough, then, to compensate for the faster depletion rate? Or is the human capacity for self-regulation a limited resource that intense exertion depletes relatively quickly—akin say to our bodies’ limited glycogen stores that fuel intense physical activity?

Over the last 20 years, a wide range of studies have been conducted to assess self-regulatory depletion in humans. (Many of these studies were conducted by Dr. Roy Baumeister and his group at Case Western Reserve University.) The results generally point toward the following conclusions: The capacity for self-regulation, just like muscular strength, is a limited resource that is subject to temporary depletion. Furthermore, the research results suggest that, for most people, this resource is rather scarce.

So, how effective are dieters at managing their limited capacity for self-regulation? The record indicates that successful long-term “losers” remain a minority, and that the vast majority of dieters are trapped in a recurring cycle of weight loss and regain—Yo-Yo dieting is the colloquial term for this process. In this all-toofamiliar pattern, dieters seeking lofty weight-loss goals are able to slash off large amounts of weight by eating very little or even starving themselves, but then run out of regulatory gas and end up, after a period of short-lived success, regaining the weight—often with “interest.”

But why?

Where More Is Less

When embarking on a diet, most overweight individuals tend to set weightloss goals that reflect their image of what their ideal body weight should be—based, perhaps, on personal notions of aesthetics, advertised “poster” success stories, or standard height/weight charts read in a book or magazine article. The greater the weight-loss goal, the greater the caloric deficit must be. The greater the caloric deficit, the more acute the person’s hunger and the greater the self-control needed to override the deprivation and sustain the diet—that is, the greater the drain rate on the dieter’s self-control capacity (stock). That’s obvious. But what is often less obvious is how much harder doing so becomes over time.

The capacity for self-regulation, just like muscular strength, is a limited resource that is subject to temporary depletion.

Dieters can seriously underestimate the escalation in hardship because, as psychologists have found, most people intuitively view causality in linear terms, expecting effect to be always proportional to cause. That is to say, we to tend to think that if A causes B to happen, then 2As must cause 2Bs to happen.

But the effort needed to accomplish a task often increases exponentially, not linearly, as the difficulty of the task increases. This principle is not unique to dieting, but applies to many tasks, both cognitive and physical. And it is, perhaps, easier to grasp in physical tasks such as, say, muscular exertion. Consider, for example, walking, which for most people is their major physical activity in a relatively sedentary lifestyle., “Escalating Energy Expenditure” portrays how energy expenditure escalates as walking speed increases, at speeds ranging from one to 10 km per hour (0.62 to 6.2 mph). It shows that as speed increases, energy expenditure rises, not in a linear fashion, but exponentially.

ESCALATING ENERGY EXPENDITURE

The effort needed to accomplish a task often increases exponentially, not linearly, as the difficulty of the task increases. For example, energy expenditure escalates as walking speed increases.

At low walking speeds—at the one- to two-mph pace of normal daily activities—the exertion of muscular energy (the stock’s outflow rate) is modest enough that the drain on energy reserves can be adequately compensated for by daily rest and food intake (the inflow rate). It is, in other words, a level of exertion that is sustainable, meaning that if we chose to, we could sustain this level of physical activity for extended periods of time without depleting our muscular energy stock. In fact, we can sustain it for very extended periods, as in the case of Deborah DeWilliams. On Friday, October 15, 2004, DeWilliams arrived back in her hometown of Melbourne after having set a world record as the first woman to walk around Australia— traveling in a clockwise direction along Australia’s National Highway 1. She completed the 9,715-mile walk in 343 days (which also earned her a second world record for the “longest walk in the shortest time”). Deborah De Williams had walked close to 30 miles per day, at a speed of two miles per hour. That translates into walking 15 hours a day, every day for almost a year—a sustained stock, if there ever was one.

As the speed versus energy expenditure plot in “Escalating Energy Expenditure” shows, walking faster can quickly increase the rate of energy expenditure. Once our rate of energy expenditure exceeds our ability to replace it, our energy reserves deplete over time. How fast? Consider what it takes to run a marathon. The human energy “stock” (even the best stocked) is barely large enough to sustain a 26- mile marathon run (quite a bit less than De Williams’ 9,715 miles.) And those resilient enough to endure that challenge will most certainly arrive with empty tanks.

Not unlike walking or running, the self-regulatory effort in weight loss escalates not linearly, but exponentially, with the difficulty of the goal. Our body’s weight set point seems to have a certain give to it, so that a person can stay a bit below it with relatively little effort. Larger weight losses, on the other hand, are difficult to tolerate. Fat-cell theory provides one possible mechanism for this physiological nonlinearity. As the enlarged fat cells of an overweight dieter (which had expanded in size during weight gain to accommodate excess energy storage) shrink back to their normal size (or slightly below it) subsequent to modest weight loss, the physiological signals to overeat and regain the weight are often easy to override. But if the weight-loss effort persists and the fat cells deplete to below-normal levels, the “volume” of the physiological message to the brain’s appetite control center increases, eventually becoming a scream:, “EAT, EAT, EAT.”

Understanding How Weight Cycling Happens

To understand how unrealistic goals can induce weight-cycling behavior, in the lower part of “The Lose-Gain Cycle,” we “walk through” one such cycle by following the numbered arrows down from top to bottom. At the start of a diet cycle, both stocks— “Self-Control Strength” and “Weight”—would typically be relatively full (such as at point 1). Voluntary restriction of one’s food intake when starting a diet causes “Weight”—stock 2—to gradually drop. Because the dieting process consumes self-control energy, the dieter drops to point 2 in the figure with both stocks partially depleted.

But this particular dieter doesn’t stop there. Her futile persistence to shed an unrealistic amount of weight causes her to keep going, depleting both stocks further. When that process ultimately depletes her self-control strength, she hits bottom—at point 3 in the cycle. While, from a weight-loss standpoint, reaching that juncture may be cause for celebration, unfortunately for her, she will not stay at that point. With a depleted stock 1, the dieter’s grip on the feeding inflow “spigot” loosens. And with adherence to the diet progressively weakening as a result, the weight stock invariably refills— propelling her back to the top of the cycle, at point 4.

THE LOSE-GAIN CYCLE

At the start of a diet cycle, both stocks—, “Self-Control Strength” and “Weight”—are relatively full (point 1). Voluntary restriction of food intake causes “Weight” to gradually drop. Because the dieting process consumes self-control energy, the dieter drops to point 2 with both stocks partially depleted. As she continues to lose weight, she depletes both stocks further, hitting bottom at point 3. With depleted self-control, the dieter’s grip on the feeding inflow “spigot” loosens, and the weight stock invariably refills—propelling her back to the top of the cycle (point 4).

This two-stock feedback structure, while admittedly far too simplified to capture the full complexity and idiosyncrasies of human weight regulation, does in fact capture the essential elements that underlie human weight-cycling behavior. Interestingly, this particular two-stock structure—two resources (stocks) interacting with one another such that the rise in one drains the other and vice versa—is fundamentally the same structure that underlies cyclic behavior in many other familiar systems, such as the pendulum clock and a child’s Slinky toy. And if we were to mathematically represent the variables in these systems and their interrelationships, the variables would assume different names— rather than body weight, feeding, and energy expenditure, we would have, for example, pendulum or spring mass, force, and momentum—but the differential equations that capture their dynamic interactions will have similar forms.

While weight cycling is surely a source of frustration to many dieters, the risks associated with repeated cycles of weight loss and regain far exceed mere disappointment. A substantial body of epidemiologic research clearly shows that repeated cycles of weight loss and regain increase the risks of chronic diseases (particularly coronary heart disease) and even premature death —independent of obesity itself.

Learning to “Manage Our Stocks”

This particular two-stock structure is fundamentally the same structure that underlies cyclic behavior in many other familiar systems.

Like any other limited (and exhaustible) resource, self-regulatory capacity needs to be managed and must not be squandered. But squandering it, not managing it, is what most dieters habitually do. The unrealistic goals that people set escalate self-regulatory exertion and over time induce regulatory depletion and ultimately relapse (not unlike a marathoner who sprints early, only to run out of gas later).

Unfortunately, setting more realistic goals rarely coincide with most dieters’ personal agendas. Nor are they encouraged to. The diet industry thrives for two reasons—big promises and repeat customers. The big promises attract the customers in the first place, and the magnitude of the promises virtually guarantees that they cannot be maintained. It makes for a very attractive business model. (J. Polivy and C. P. Herman, “If at first you don’t succeed: False hopes of self-change,” American Psychologist, 57(9), 2002).

Thankfully, however, things may be changing.

A growing understanding of the biological factors that regulate body weight and of the cognitive difficulty of maintaining large weight losses is prompting a redefinition of the “successful” goals of obesity treatment. Slowly but surely, moderation is becoming the overriding theme in weight-loss efforts. A major impetus for this shift has been the growing evidence that moderate weight losses of only 10–15 percent of initial weight, even among substantially overweight individuals, are associated with a significant improvement in nearly all parameters of health—including blood pressure, heart morphology and functioning, lipid profile, glucose tolerance (among diabetics), sleep disorders, and respiratory functioning. And these findings are now prompting a growing number of federal agencies and health organizations to call for setting more realistic weight goals rather than striving for an “ideal” weight.

To this system thinker, that’s music to the ear.

Tarek K. A. Hamid, PhD, is an MIT-trained system dynamicist with expertise in human metabolism and energy regulation. He is a professor of system dynamics at the Naval Postgraduate School in Monterey, California. He is the author of Thinking in Circles About Obesity: Applying Systems Thinking to Weight Management (Springer, 2009).

The post Thinking in Circles About Obesity appeared first on The Systems Thinker.