System Dynamics


Volume 16 – Number 2          Fall 2003


From the President
From the Executive Director
Member News
     Nelson Repenning Wins Forrester Award
     Dana Meadows Student Prize Awarded in New York
     In Memoriam – Lee Frost-Kumpf
     Nan Lux Retires from MIT
     New PhD’s
New York City Conference Report
     Defining and Refining Methodology
     Outstanding Applications
     Economics and System Dynamics
     Teaching System Dynamics
     Economics Roundtable
     Military Roundtable
     Group Model Building
     Letters from Conference Participants

Chapter News
     Latin America
     Education SIG
     Environmental Dynamics SIG
     Health Policy SIG
Policy Council Holds Summer Meeting in New York City
Announcements and Calls for Papers
    2004 International System Dynamics Conference, Oxford, England
System Dynamics Review Special Issue:  Dynamics of Supply Chains and Networks
Society Sponsors
Publication and Contact Information

Many thanks to all who contributed their writing, photographs, and information to this issue of the newsletter:  Henk Akkermans, Lina Al-Qatawneh, Yaman Barlas, John Barton, Carmine Bianchi, Bob Cavana, Dean Christensen, Henry Cole, Brian Dangerfield, Pål Davidsen, Vedat Diker, Isaac Dyner, David Exelby, Jay Forrester, Carol Frances, Paulo Gonçalves, Andreas Größler, Fazel Hayati, John Heinbokel, Jack Homer, Michael Kennedy, Hironori Kurono, Nan Lux, Rod MacDonald, Geoff McDonnell, Taehoon Moon, John Morecroft, Anastássios Perdicoúlis, Jack Pugh, Hal Rabbino, K. Raman, Rohita Singh, Krystyna Stave, Jim Thompson, Silvia Ulli-Beer, Qifan Wang, Elise Weaver, David Wheat, and Aldo Zagonel.

From the President

Adapted from the President’s Address given at the International System Dynamics Conference in New York, July 2003

Dear Members of the System Dynamics Society,

On behalf of the Society, let me first take this opportunity to officially congratulate the winner of this year’s Jay W. Forrester Award, Nelson Repenning, and the winner of the Dana Meadows Prize, David L. Cooke, along with honorable mentions to Jan Jaap Bezemer, Özge Pala, and Klaus Vogstad.

I want to take this opportunity to thank you all, members of the system dynamics community, for having inspired me to work in this fielda field of endless opportunities for establishing professional and personal relationships and for interdisciplinary research, education and practice in an exceptionally wide variety of domains.

In the last issue of our Newsletter, there is a picture of the participants at the International Conference on System Dynamics at Geilo, Norway, in 1976. It is a rare picture of a collection of pioneers in our fielda conference that resulted in the profoundly interesting book The System Dynamics Method, edited by Jørgen Randers. I would have liked to have been thereit would have saved me long detours in my research.

Eight years later, I met with several of you at the conference in Oslo. George Richardson and John Sterman subsequently introduced me to this community of scientists and teachers. To meet and work with the people behind the papers and presentations made a world of difference to me. I sensed a willingness in this community to address major, real, complex, dynamic issues. I soon understood that the dedication to the system dynamics method, so strongly articulated in this community, did not originate from a religious belief, but from a realization that the method was exceptionally well suited to address these kinds of issues, to help us understand the issues and, perhaps even more importantly, to help us understand why we often do not understand them. From this realization comes the profound respect for complexity, the critical thinking skills, and the truly exceptional willingness to share their insights that members of this community have in common.

This very personal experience of my first system dynamics conference I know I share with most people who meet us in such a context for the first time. It is such a precious moment that I think we who have been here “forever” should take a moment to reflect on how we welcome them. Let me take this opportunity to give a very warm welcome to our new members and first-time conference attendees.

In line with what I just said, I regret to have to report to you that for some people around the world it becomes increasingly difficult to attend our conference due to severe security restrictions imposed upon travellers across our globe. On behalf of our Society I can only regret that our members are facing such constraints and have to endure such discomfort in their sincere attempts to take part in the development of our field. And I can promise that I will find ways to meet for scientific discourse.

I believe that this community is a resource that can be more effectively put to use for the benefit of society at large. One of the most important tasks for the System Dynamics Society is to find ways to accomplish that.

In this address to you, I will talk about some of the challenges that the System Dynamics Society, our community, and society at large are facing and how I think we might be addressing those challenges.

Let me first tell you that I think it is important to the field and, therefore, extremely encouraging to find that there is a growing number of exceptionally competent women entering our field.

Then let me first turn to our own Society. I think it is important that you, members and potential members of our Society, are aware of the concerns of our Policy Council, our officers, and our home office. We want you involved as active participants in shaping the Society so that it can work effectively to the benefit of you, its members.

If you read the phenomenal report on the Society produced by our home office, you will see that our Society has grown exponentially in terms of members. Over the last couple of years, we have developed from a Society relying entirely on volunteer contributions to one that, for the most part, is professionally run by the executive office located at the New York State University at Albany.

So effective have been our Executive Director, Roberta Spencer, and her associates and volunteers at Albany that very many of the tasks previously handled on an ad hoc basis by volunteers are now taken care of.

We members have come to expect that the Home Office has infinite capacity. At the moment, however, there is not sufficient funding to support the huge amount of work currently undertaken by that office. Consequently, such a shift of the burden to the intervener has left the home office understaffed and underfunded. The reason for this is, in part, that the membership fee has not developed at the rate of the services offered by the Society.

So, while taking care that low- and regular-income members will be shielded, we consider a change in the fee structure of the Society. Moreover, we are actively looking for ways to generate additional income. For years, Beer Game sales has been an important source of income. I would like to challenge the membership to develop and donate a new game from which we may all benefit. In the meanwhile, other options are being considered, such as the licensing of courses. We would certainly be open to additional suggestions as to how the Society can, cost-effectively, serve you better.

A major portion of the work conducted at the Home Office is related to the annual System Dynamics Conference. For that, and a number of other good reasons, we are about to modify slightly the way we select conference sites, take a stronger hands-on approach in the organization of conferences, and improve the program structure so as to offer better exposure to quality work and increase participant interaction. Again, we welcome suggestions from our membership. I will return to the importance of such a restructuring of the conference to the way I believe our field will develop in the future.

The Society concerns itself with the growth of the field and its community of system dynamicists at large:  scientists, educators, students and practitioners in the private and public sectors. As we all realize, the dynamics of the field is governed by a non-linear feedback structure. I would like to challenge the students present at this conference who might want to develop and analyze a model of the field and help us come up with appropriate policies for a sustainable development.

The growth concept itself is not simple and has a variety of dimensions that we need to distinguish, in particular growth in volume and growth in quality.

The concern has been raised that rapid growth, in terms of number of practitioners, may compromise quality in the work accomplished. Moreover, there is the concern that such quality erosion may backfire to hurt the field in the form of a bad reputation.

In particular, this has been an issue in our debate on conferences. Indeed, if one takes our conference as an indicator, there is a spread of quality among the papers submitted. So this is a valid concern. Yet, such a distribution is a natural consequence of an exponentially growing field in which the seasoned masters constitutes a decreasing minority.

Some deviation from perfection is OK. Without it, there is no room for improvement, no potential for learning. More than any, our Society believes in corrective feedback and it is essential that we, in our conferences or any other fora where we meet, make room for imperfections and ensure that proper feedback is offered. The listserve entertained by Ventana and the PhD colloquia organized by the student chapter are among our more successful established feedback mechanisms. Yet, with improved capacity, we can do a lot better. And herein lies the key to success:  We need to build feedback capacity to strike a reasonable balance between quantitative and qualitative growth.

As a result of the current, intense press coverage of world events, there is undoubtedly an increased public awareness (recognition) of the dynamic complexity of that world. There is a growing sense that we need to understand that complexity and, based on such an understanding, develop policies that effectively deal with the challenges with which we are confronted. In theory, therefore, there should be a growing demand for system dynamics.

In practice, however, the specific call for a system dynamics approach to deal with existing problems is hardly audible. We understand the reasons for this:

For each and every domain of application, other theories, methods, techniques, and tools have already been developed and gained recognition, and are currently being employed and taught in schools, colleges, universities and courses by an establishment of scientists and practitioners. In emerging as well as more established fields, strong reinforcing mechanisms play out. They contribute to establish a field identity, to support the establishment of scientists and practitioners in the field, to build a portfolio of success stories, and also to inbreeding.

As indicated by Professor Forrester and demonstrated, in the case of economics, by Mike Radzicki, Glen Atkinson, Jed Schilling, and Jerry Barney, there is growing evidence that some of the ways currently employed to go about addressing real complex, dynamic issues have their shortcomings and have not always been successful.

There are two approaches we may take, and one does not exclude the other:

Professor Forrester suggested one. Based on the assumption that there is no hope for people who adhere to linear, open-loop, equilibrium thinking, we need to educate a new generation of scientists and practitioners starting in K-12. Jay Forrester, backed by his mentor, Professor Gordon Brown, sponsors such as the Waters Foundation, facilitators such as Nan Lux and Lees Stuntz, motivated students at MIT and innovative educators such as Diana Fisher and many, many others initiated an impressive program that has led to the introduction of system dynamics in North American schools.

Here is another way to go about it:  In the confidence cracks that result from a growing awareness of the shortcomings of alternative approaches, we may search for fertile soil and a golden opportunity for the introduction of system dynamics.

This way, we need not act with confrontation and emphasize system dynamics as an alternative approach, although this is actually the case, but may portray our approach as an enhancement, a reinforcement, of existing theories, methods, techniques, and tools. We may employ a friendly Trojan horse–not a fake one that wins by deceit, but a real one that can carry a load of success.

Here are some of the reasons why this might work in various disciplines and domains:

  1. It will allow us to tap into the lingo of the domain to effectively introduce more alien concepts such as structure and behavior, stock levels and flow rates, accumulation and delay, and, last but not least, feedback and non-linearity. These are concepts that, in each and every domain, take on a unique meaning (if any at all), and when used without a careful adaptation may seem meaningless at best. Yet more often they lead to confusion.
  1. Employing a friendly Trojan horse will help us address problems people believe they have, not problems we believe they have but that they may not recognize. In other words, we will tap into frustration and offer enlightenment and, potentially, solutions to real problems. This is what I mean by fertile soil. Marketing system dynamics otherwise is extremely difficult. We market system dynamics as a way to address and cope with dynamic complexity. Yet, in our marketing, we are often asked to demonstrate its effectiveness under severe time and resource constraints. These are constraints that do not allow for elaboration, and so it becomes paramount to address the specific, current needs of our clients.

What I have stated here, applies not only to various fields of applications, but also to disciplines themselves, such as OR, that may well benefit from the adaptation of system dynamics.

This way, I believe system dynamics may gradually gain exposure and demand may pick up. That takes good scholars working intimately with open-minded specialists in various domains and disciplines to assimilate that specialization. And it takes good textbooks, appropriately adapted to the variety of domains and disciplines that may benefit from system dynamics.

Our best system dynamicists are scholars of this kind and authors of such textbooks. They are among our best practitioners and educators. I think we all have a lot to learn from them on how to introduce system dynamics in ways that do not raise suspicion, aggravation, and resistance.

I think it is important to keep in mind that, because system dynamics, when applied appropriately, is extremely powerful, it may inadvertently be perceived as intimidating and confrontational. If people do not understand the dynamic complexity of their own field, it is not because they are stupid or incompetent, it is because they have not had the benefit of the theory, methods, techniques and tools that allow them to gain such an understanding. Again, through system dynamics we gain respect for complexity and that respect must extend to the people trying to cope with that complexity, whether they are school children, students, scientists or practitioners in the private or public sector.

So the need for system dynamics is there. For this need to manifest itself into actual demand, we need fora for exchange of ideas and insights. This is where I will come back to the structure of the system dynamics conferences. It has been proposed that future conference programs be organized around threads that exhibit the application of system dynamics in various disciplines and, across disciplines, in various domains. In addition, there will be threads addressing recent developments in theory, methods, techniques and tools. These threads will be designed and chaired by experienced system dynamicists with a keen interest in hosting such a thread. These chairs will serve as such for a number of consecutive conferences and will, together with the Program Chair, constitute a somewhat more permanent Conference Committee.

I expect that such an improvement of the program structure, in particular a more permanent responsibility for and identification with threads, will benefit the quality of the conference. We have seen an emerging example of that associated with the micro-thread on security. The presenters in that thread all met in a workshop earlier this year to prepare and critique their drafts and have published an affordable special collection of papers in that domain.

Moreover, and perhaps more important, such a program structure has the potential to greatly benefit the community of system dynamicists at large. I would expect that we will see a number of special interest groups (SIG’s) grow out of these threads. Each such group will be able to connect to experts in related fields, to invite them to our conferences, and to offer threads in their conferences. We need them as well as they need us. Such a mechanism will allow our field to effectively reach out to other disciplines and gain exposure in a variety of domains of application.

Moreover, it will serve to solve a dilemma:  To promote cross-fertilization, it has repeatedly been suggested that we co-organize our conference with other conferences, such as TIMS, AOS and others. Yet, we recognize that few such arrangements would satisfy even a majority, let alone all, of our members. With the establishment of threads and SIG’s emerging as a result, we may comfortably reach out to a variety of communities of scientists and practitioners.

Such interaction will contribute to strengthening our conferences, in terms of structure, variety, and quality.

Practice in various domains constitutes our growth engine. Good practice breeds demand for additional work. We fuel this engine with human capital through our education and with technology through software development. If scholars, educational material, including textbooks, effective educational programs, and technology are in short supply, inaccessible, or inadequate, then a rapid growth in the demand for system dynamics will cause:

  1. clients, in search of solutions, inevitably to turn to other competencies offering alternative approaches; and
  2. people to employ system dynamics and associated software less competently so that we, in the long term, will experience an increase in the amount of low-quality work offered, and the field will develop a bad reputation.

The growth of our field is effectively constrained by the supply and quality of human capital and technology.

It is, consequently, of utmost importance that we continue to develop our educational capacity, in terms of volume as well as quality. Among the set of goals that we should strive towards, are the following:

I said that I would return to basic system dynamics research. By this I mean research on the theory, methods, techniques and tools that constitute the foundation for our work. With a few, brilliant exceptions, little has been happened in the area of theory, methods, and techniques since Jay Forrester and his PhD students took their first steps up the path that subsequently led to what we now call system dynamics. As for tools, some major innovations have produced software that has improved the way we visualize and analyze models. In spite of this development, the lack of basic research is apparent and of serious concern. Such research should constitute the foundation for further software development in three major directions:

One is to make system dynamics more accessible to the public, i.e. to model developers, model consumers and model communicators.

Research is required into the bridging of spreadsheet with system dynamics technologies so as to bring on board a vast group of, in part, deeply concerned modelers struggling with an inadequate tool, just not really realizing what their problem is. This technology is exemplified by the work being done by Mohammad Mojtahedzadeh.

Research is also required to improve visualization. At the moment there is a major divide between the representation of system structure and system behavior. Except for a very important increase in the use of phase diagrams, minor progress has been made to illustrate the core issue in system dynamics–namely, the relationship between structure and behavior in complex, dynamic systems. A very encouraging exception to this fact is work behind the SyntheSim facility offered by Vensim. Along the same line, there has been very moderate research into the use of new media technologies to enable system dynamics to meet the public by distilling and “telling the story.”

And research is required to ensure ways to effectively generalize, store, retrieve and reuse a repository of generic models, through the process of encapsulation, inheritance, instantiation, parameterization, and initialization. We should more aggressively adapt an object-oriented approach to modeling–well-recognized as a key to boost productivity in software engineering.

The next area of basic research constitutes the foundation for experimental system dynamics aimed at revealing the mental models and processes governing our actions in complex, dynamic environments–i.e. to prepare system dynamics to work effectively with the community of cognitive, behavioral and social psychologists. We need to elaborate our ways of designing, conducting and analyzing the results of such experiments, and we need software support to set up interactive labs for this purpose, for effective interface design, and for data collection and subsequent analysis.

Finally, what I think is perhaps the most important challenge in basic research is to understand the relationship between structure and behavior in non-linear, dynamic systems. This is a challenge faced by model investigators and policy designers. This research has four components.

In our analysis, we need to find out what part of the structure, at any point in time, governs current model behavior. As this will change over time, we need to develop a method for model (system) surveillance.

Moreover, we need to develop measures for leverage point elasticityto obtain indications of how effective the utilization of various leverage points will be in improving model behavior, by way of modifying the relative significance of various structural components in the model.

We then need to develop ways to control the relative significance of the components of the underlying model structure in the form of policies, so as to reinforce favorable at the expense of unfavorable model behavior.

Ultimately, we need to find ways to implement our research findings in the form of software that visualizes the relationship between structure and behavior so effectively that the conclusions from our analysis and policy design become self-evident. Such an approach would increase the likelihood that our clients adopt the outcome of our analysis and implement our policy recommendations.

As exotic as this may sound, there is currently ongoing work in each of the basic areas of system dynamics that I have outlined. Due to the significance of this work, I believe that the results will find their way into software tailored to meet our needs. To speed up this process that will benefit us all, we should encourage, sponsor, and engage ourselves in this kind of research and establish tighter links between the community of researchers working in these areas, practitioners in need of the technology, and software developers.

Beyond what I have said about the need for and the opportunities to work with members of other disciplines and domains, I will not take time to comment on the variety of applications of system dynamics. It is sufficient to state that the scope and the quality of work being conducted are simply impressive.

But I would like to follow up what Jay Forrester, Mike Radzicki and others have emphasized throughout this conference–the fact that we need to address the major issues. Now, what those issues are depends on the perspective one takes.

Let me offer my perspective, that of people in the world in distress from poverty and armed conflicts that result from social injustice, political unrest, or plain malgovernance.

We are gathered in New York City–the USA’s melting pot and the capital of world diplomacy. The ultimate purpose of diplomacy is to prevent conflicts from arising and to resolve conflicts by peaceful means once they have arisen.

Earlier this year, diplomacy failed. Never before have so many people, all over the world, been able to observe cultural clashes, lies and deception, political pressure, and dirty deals, bringing the UN to a stalemate.

Regardless what side one chooses to take, I think one might agree that the associated political conflict resulted from poor political and diplomatic craftsmanship for which many nations must share responsibility.

The conflict over Iraq is the one that caught our attention this year. Remaining unresolved are several hundred local and regional conflicts that claim thousands of innocent victims every single day, most of them in the developing world.

Many of these conflicts remain unnoticed for decades. In several cases, they are not even considered conflicts because their more disadvantaged stakeholders have no power to speak up. These stakeholders are subject to bad, often corrupt, governance and are victims of social injustice. They are victims because they are not offered sustainable living conditions, including basic health care, are not given the opportunity to provide for themselves, say, through education, and remain at the mercy of social welfare or ruthless exploiters. It is a grave fact that most of these victims are women and their children who, if they ever had a chance, would constitute the future of our civilization. It is also a sad fact that many of these children are brought up to recognize crime as the only way by which they may escape what seems to be their destiny.

It does not take much analysis to recognize that, in these cases, we find ourselves in a fuzzy, uncertain, dynamic, non-linear feedback environment, governed by multiple stakeholders who hold a variety of misperceptions, unaligned goals, hidden agendas, and conflicting visions for the future. Needless to say, to a system dynamicist, this sounds like an opportunity not unlike what we have been used to find in corporations and one not to be missed.

Sincere compassion and professional inspiration compels me to ask the question:  How may we employ system dynamics as a means toward national and international conflict resolution, conflict prevention and social rehabilitation?

Maybe I am overly optimistic, but I am convinced that our approach has the potential to help people in conflict distil mental models, come to consensus as to what constitutes the core of their conflict, and represent that core in the form of an explicit road-map model that may be applied for analysis and policy alignment to secure peace and to bring about political stability and social sustainability. Personally, I am very much looking forward to the opportunity to work next year with the Millennium Institute, Mike Radzicki, the Carter Center and the World Bank along these lines.

We are faced with the daunting task of bringing system dynamics to the developing world. Here is my comfort:  As indicated in reports, such as the one published by the Structural Adjustment Participatory Review International Network (SAPRIN), I think it is safe to say that nowhere else do we find more fertile soil for our seeds than in the developing world.

Nowhere else has the limited success or even failure of traditional approaches to conflict resolution and development planning been more apparent. Nowhere else have their failures in addressing social problems had graver consequences. Nowhere else has the motivation to find alternative approaches been greater. And nowhere else have the educational systems from kindergarten to university been more ready for adaptation.

I conclude my address by wishing that you may find great satisfaction in the use of system dynamics for what you think are the major issues in your world.

Pål I. Davidsen

Society President Pål Davidsen

[ Return to contents ] [ Next ]