Part 1
Read the scenario as well as the Introduction and Chapter 1 of the Meadows text, the Cathon article on the Learning Organization, the Zemke article on Systems Thinking, and the other material required, and then respond to the prompts that follow.
Scenario
As an expert in systems analysis, you’ve received an e-mail from the chief financial officer (CFO) of XYZ Manufacturers to discuss a potential consulting project. The CFO is vaguely familiar with the concept of systems thinking but isn’t sure it could be successfully applied to her fast-paced, global business.
- Based on your knowledge of systems thinking and the learning organization, reply to the CFO’s enquiry by explaining at least three challenges of managing complex organizations and how and why effective systems thinking can help improve their performance. Support your response and reasoning with explicit and appropriate references to the readings and with at least two other theoretical frameworks or academic references about systems thinking and practice. (2 -3 paragraphs)
- Having learned about systems thinking and the learning organization, and reviewed at least one other theoretical framework or academic reference about general systems thinking and practice, give at least two examples, from your experiences in organizations, in which the application of systems thinking could have helped the organization become more effective. Be clear in explaining which specific principles and concepts from systems thinking could have helped in the examples you choose, and explain how and why. (4 – 5 paragraphs)
Part 2
Read the “Opportunity Consultants, Inc., Case Study” and then respond to the following prompts:
- Using a systems approach, analyze the performance of Opportunity Consultants, Inc. and develop a case-specific “effect-cause-effect logic” tree diagram using the 5-Whys analytic tool. See the Rubric for details on what this diagram should contain.
- Write a summary description of your diagram with specific recommendations for improving the club’s performance that are linked to your diagram analysis. (1–2 pages)
Part 3
Read the “Baria Planning Solutions Case Study” and “Facilitating Systemic Thinking in Business Classes” documents, and then respond to the following prompts:
- Using a formal systems diagramming approach, analyze Baria’s performance and develop a robust “effect-cause-effect logic” tree diagram using the 5-Whys tool, as in Part One.
- Create an appropriate, simple causal loop diagram (CLD) that incorporates relevant and logical feedback loops to capture the fundamental system behaviors, outcomes, and causes in the “Baria Planning Solutions Case Study.” See the Rubric for details on what this diagram should contain.
- Write a summary description, including specific recommendations, that links directly to your 5-Whys and CLD analyses for improving Baria’s sales support operations and organization as a whole. (1–2 pages)
Part 4
Read the “Bayonne Packaging, Inc., Case Study” and “The System Archetypes” documents, and then respond to the following prompts:
- Using a formal systems diagramming approach, analyze Bayonne’s organizational performance and develop a robust “effect-cause-effect logic” tree diagram using the 5-Whys tool, as in Part One.
- Create a robust causal loop diagram (CLD) that incorporates appropriate causal loop logic in the analysis and that also identifies common system archetype patterns within the diagram. This diagram should describe fundamental system behaviors and outcomes.
- Write a summary description, including specific recommendations, that links directly to your CLD analysis (which includes embedded archetype relationships) for improving the packaging company’s operations and the organization as a whole. (1–2 pages)
Decision Sciences Journal of Innovative Education Volume 4 Number 2 July 2006 Printed in the U.S.A.
Facilitating Systemic Thinking in Business Classes
J. Brian Atwater† Business Administration Department, College of Business, Utah State University, Logan, UT 84322-3510, e-mail: batwater@b202.usu.edu
Paul H. Pittman School of Business, Indiana University Southeast, New Albany, IN 47150, e-mail: ppittman@ius.edu
ABSTRACT
This article identifies and describes three dimensions of systemic thinking—thinking holistically, thinking dynamically, and thinking in terms of feedback loops. We contend that within the field of business a significant amount of attention has been paid to holistic thinking, but relatively little to the other two dimensions. In addition, we present evidence that managers will need to more fully develop systemic thinking skills as business systems become even more interdependent in the 21st century. Tools are presented that can be used in any area of business to help students develop their systemic thinking skills and gain better insights in their respective fields of study. Examples of how to introduce and use these tools in the classroom are also provided.
Subject Areas: Causal Loop Diagrams, Management Education, System Archetypes, and Systems Thinking.
INTRODUCTION
Are we providing business students with the tools they need to successfully manage organizations in the 21st century? During the last decade of the 20th century, sev- eral well-respected management experts published books and articles emphasizing that businesses are complex social systems and that management practices must change to be effective in this environment (Ackoff, 1994; Deming, 1994; Forrester, 1994; Senge, 1990). At first glance, it appears that universities have heard the crit- icisms and implemented the experts’ recommendations. The concept of systems is omnipresent in business programs. Students are taught about production systems, accounting systems, information systems, financial systems, and so on. Basically, all the functional areas discuss various systems housed within them. The critical question, however, is whether learning about these systems is adequate to prepare managers to succeed in what some are calling the age of systems. Decision making in today’s business environment is very complex. Successful managers must be able to make decisions that cut across organizational functions. They should be
†Corresponding author.
273
274 Facilitating Systemic Thinking in Business Classes
able to anticipate probable sources of policy resistance and identify both the short- and long-term ramifications of their decisions. They must be able to make these decisions while juggling with the objectives of multiple stakeholders. To handle this level of complexity, students must learn to think systemically.
It is important to note that this is not an indictment of current teaching prac- tices. There is no doubt that students must learn about systems. Studies have shown that students’ acquisition of operational skills is heavily dependent on the concep- tual knowledge they gain at earlier stages of their education (Wickens, 1992). Consequently, students must learn what a system is, how businesses fit the systems paradigm, and what types of subsystems are embedded within them. They also need to learn about the various elements making up the different types of subsystems in a business, along with how they work and interact. While this core knowledge is crucial, it does not develop their systemic thinking skills.
Graduate business programs need to cultivate systemic thinking skills for sev- eral reasons. First, trends in business like supply chain management, virtual work teams, global sourcing, and multinational operations will only increase the number of complex interactions that managers must juggle. Second, very few people truly know how to think systemically. Education programs are dominated by analytical techniques. The analytical paradigm is so pervasive that most students consider the terms “thinking” and “analysis” as synonyms. Consequently, there is a need to make students aware of the systems concept. Third, developing systemic think- ing skills helps managers understand better how the systems within their function operate and interact with other systems both inside and outside the organization. Finally, systemic thinking equips managers to develop better their own intuition about complex system behavior. These last two factors are important elements in the lifelong learning concept commonly recognized as an essential part of every manager’s development. Obviously, educational programs cannot really teach stu- dents all the possible system interactions they will face in their careers. However, if they learn basic systems principles and are taught specific tools that help them apply those principles to comprehend system interactions, students will be better prepared to evolve their own understanding.
What is systemic thinking? What skills are required to think systemically? How can systemic thinking skills be cultivated? This article attempts to answer these questions and provides examples of how systemic thinking skills can be integrated into graduate-level business classes.
WHAT IS SYSTEMIC THINKING AND WHY IS IT IMPORTANT?
Systemic thinking is a difficult process to describe. Nevertheless, if educational programs are going to help students learn to think systemically, the cognitive pro- cesses involved must be made explicit so that methods can be devised to help students develop these skills. Therefore, a specific description of systemic thinking is provided in this article. The description is based on the integration of ideas from several noted experts in systems theory. The focus of the definition developed here is on the cognitive processes necessary to gain holistic insight into a situation and the implications of making changes to the status quo.
The first element of systemic thinking focuses on the concept of holis- tic/synthetic thinking described by Ackoff (1981). Dr. Ackoff explains that, for the
Atwater and Pittman 275
past 400 years, the analytical paradigm dominated our approach to understanding the world around us. In reality, analysis is only one method of perceiving the world around us. Ackoff goes on to introduce the concept of synthetic or holistic thinking and then differentiates it from the analytical approach. Analytic thinking attempts to understand a system by breaking it into its smaller parts and studying these parts in isolation. Once the parts are understood, the analyst tries to explain the behavior of the whole based on the behavior of the parts. In contrast, synthetic thinking starts by trying to understand the larger context that the system operates within. Once the role of a system within its larger context is understood, the synthetic thinker tries to explain the behavior of the system based on that role. Looking at the two types of thinking from a different perspective, analytical thinking helps people understand what the parts do and how they work, while synthetic thinking explains why the parts do what they do. Ackoff goes on to point out that a crucial factor for un- derstanding system behavior is observing how the parts interact. Consequently, he maintains that the very act of analysis (i.e., studying the parts in isolation) makes it impossible to truly understand a system, thus highlighting the need for developing synthetic thinking skills.
Understanding why systems behave the way they do is particularly impor- tant in today’s businesses, which have evolved into multi-minded, multipurpose, complex social systems. A quick review of the evolution of the systems view of a business helps clarify this point. The industrial revolution was fueled by the view that business organizations were essentially mechanistic systems. Viewing busi- nesses as mechanistic systems, Frederick Winslow Taylor used analytical thinking to develop the scientific management movement, which was credited with signifi- cantly improving manufacturing operations in the early 20th century. Through the use of scientific management, work was broken down into simplified parts. Em- ployees were assigned these simple tasks and, as a result, workers were viewed and treated like interchangeable machine parts.
Over time, the systems view of a business shifted from mechanical to biolog- ical. This shift took place as a result of business owners using equity financing to fund the growth of their organizations. As ownership of the business was diffused, managers, rather than owners, became responsible for running organizations. A hierarchy was developed for management and a divisional structure was created to enable the various functions to grow with the business. The analytical perspec- tive was still applied and it was widely believed that a business would perform optimally if each function tried to optimize its performance in isolation from the other functions. There were clear separations between manager and worker activ- ities, which included decision making and problem solving to be performed by managers (i.e., the brain of the organization), while the fundamental work was performed by hourly labor (i.e., the body of the organization). This hierarchical arrangement and divisional structure became widely used and is still the dominant approach employed within business organizations as well as schools of business and their curricula today.
As businesses continued to evolve and each functional area pursued excel- lence in isolation, conflict between the functions started to emerge. Shortages of shared resources and inconsistencies across functional performance measures were just a few of the many symptomatic issues that created these conflicts. Furthermore, a variety of social changes forced managers to become aware of other stakeholders
276 Facilitating Systemic Thinking in Business Classes
in their organization. In addition to shareholders and customers, managers needed to address the concerns of employees, suppliers, government agencies, special in- terest groups, and society at large. Because of these changes, the systems view of a business was altered once again from biological to multi-minded, multipurpose, complex social systems. In essence, organizations are now recognized as being part of a larger purposeful system (i.e., society) with many subsystems (functional areas and/or teams) and parts (employees), all seeking to fulfill their own individual purposes.
Ackoff and Gharajedaghi (1996) assert that many of the problems we cur- rently see in business and other social systems are due, at least in part, to managing social systems as if they were mechanical or biological. If the various purposes held by the business, its subsystems, and its parts are not recognized and properly man- aged, organizations will experience high employee turnover, functional infighting, and a host of other problems. To effectively manage a multi-minded, multipurpose, social system, managers must understand the motivation behind the behavior of the various elements of the system. Understanding why the parts of the system behave as they do enables managers to acknowledge the concerns of these vari- ous subsystems. By acknowledging these concerns, managers reduce constituent fears that their interests are being overlooked or dismissed. In addition, managers need to communicate why they are taking specific actions. The process of learn- ing, acknowledging, and explaining the whys behind behavior is essential in social systems to create an environment where managers have the latitude to shuffle the relative priorities of objectives over time. This understanding is also essential for creating the changes necessary to remain agile and competitive in today’s global economy. In short, managers must take a holistic approach to understanding situ- ations, communicating how and why they are prioritizing various objectives, and making decisions. Consequently, the first part of the definition of systemic thinking used in this article is synthetic or holistic thinking.
While holistic thinking is an essential part of systemic thinking, it does not sufficiently describe all the cognitive processes necessary to think systemically. Forrester (1971) identified several characteristics of complex systems, which make it difficult for people to understand and work with them. These factors include:
� Cause and effect are often separated both in terms of time and space.� Problem resolutions that improve a situation in the short term often create bigger problems in the longer term, and actions that make things worse in the short term often have long-term positive effects.� Because of the first two characteristics, people often do not learn from their mistakes. – Long time delays often result in one person creating a cause and another
experiencing its ultimate effect.
– Due to the differences in short- and long-term effects, what a person learns from the short-term result of a decision is completely different from the true long-term outcome.� The subsystems and parts of a system interact using multiple, non-
linear feedback loops. This complex flow of interactions often creates
Atwater and Pittman 277
counterintuitive behavior. Consequently, what appears to be the obvious decision is, in fact, often a bad choice.
Note that the factors identified by Forrester focus on time and complex interac- tions. Therefore, an operational view of systemic thinking must include capabilities related to understanding the impact of these phenomena on a system.
Senge (1990) distinguishes between two types of complexities that further clarifiy the phenomena described by Forrester. These are detail complexity and dynamic complexity. Detail complexity involves numerous variables, details, and data and is linear in nature. This type of complexity comes about when a decision maker must determine an optimal, or at least satisficing, solution to a problem with many variables and a large number of possible solutions. The process most typically used to solve problems centers around detail complexity, because we typically gather large amounts of information and use the power of computers and information systems to help identify the best solution. Dynamic complexity comes about from the interactions of the decision variables over time. This type of complexity can occur even in very simple situations where there are only a few variables involved. Nevertheless, dynamic complexity is very difficult to visualize, because it involves situations where cause and effect are subtle and where the long-term effects of interventions are not always obvious.
As previously stated, the parts in a social system all have objectives and are constantly interacting; consequently, they exhibit high levels of dynamic complex- ity. Because of the interdependency of the parts in these systems, changes cannot be made in isolation. There are always feedback loops that create unintended con- sequences. Furthermore, these feedback loops often do not follow a simple linear course and commonly include time delays. Most people do not account for dynamic complexity in their decision making. In fact, many people do not even realize it exists. This is because our understanding of cause and effect is tacit rather than explicit, because we generally learn these dynamic relationships at an early age through simple situations such as; “If I cry someone will come to help me,” “If I touch a hot stove I get burned,” “If I don’t watch where I am walking I’ll stumble over something,” and so on. In addition, the analytical focus of our formal edu- cation further reinforces our linear rather than nonlinear understanding of cause and effect. Without tools to formally help us develop our systemic understanding, we naturally develop an event-oriented view of the world (Senge, 1990; Sterman, 2000). With this view, people see the world as a series of simple cause-and-effect relationships, where an effect has a single cause that occurs just prior to the effect surfacing. This perception prompts us to treat problems as isolated events and view solving them as a discrete, linear process; a problem is recognized, alternatives identified, solutions selected and implemented, and the problem is resolved. While this belief holds true when working with simple systems, it breaks down when dealing with complex social systems. Senge (1990) articulated the common mis- perception of event-oriented thinking very succinctly in his classic book The Fifth Discipline when he wrote, “Today’s problems come from yesterday’s solutions.”
To complete this description of systemic thinking, two terms described by Richmond (2000) are used. These terms are “Dynamic Thinking” and “Closed- Loop Thinking.” Dynamic Thinking describes the ability to see a phenomenon as
278 Facilitating Systemic Thinking in Business Classes
the result of behavior over time rather than a reaction to an isolated event. Closed- Loop Thinking requires the decision maker to examine the role that the structure of the system (i.e., reward systems, information flows, logistical processes, etc.) plays in creating behavior. It also examines interactions with external forces. Once the structure and interactions are recognized, closed-loop thinking attempts to understand how these interactions feed back to shape the ultimate result of an intervention. Synthesizing the above discussion results in the following three-part description of systemic thinking:
Synthetic Thinking— studying the role and purpose of a system and its parts to understand why they behave as they do.
Dynamic Thinking— examining how the system and its parts behave over time.
Closed-Loop Thinking— investigating how the parts of a system react and interact with each other and external factors.
THE NEED AND IMPACT OF TEACHING SYSTEMIC THINKING
Forrester (1971) asserted that the human mind is incapable of truly understanding the behavior of complex social systems without the assistance of tools and tech- nology. Several studies have been conducted which confirm this assertion. Dörner (1996) used simulation games to test people’s ability to make effective decisions when dealing with problems in complex social systems. His study showed that people do not intuitively think systemically. Booth Sweeney and Sterman (2000) developed an inventory of exercises designed to examine people’s ability to rec- ognize and anticipate the results of systemic interactions. Participants performed poorly even on the simplest tasks. Furthermore, this lack of ability appears to transcend age, national origin, educational background, and other demographic variables. Ossimitz (2002) conducted a sequel study to that of Booth Sweeney and Sterman using Austrian Business Administration students with identical re- sults. These studies provide strong evidence that systemic thinking is not an innate skill for most humans. Therefore, in order to develop the ability to understand and work effectively with complex social systems, people must be trained in systemic thinking tools and concepts.
Subsequent studies investigated the impact of teaching system dynamics tools on the ability of decision makers to grasp better the systemic nature of a situation. Kainz and Ossimitz (2002) showed that even a crash course of only 90 minutes, which exposed students to systemic thinking tools, significantly improved their ability to think systemically. Pala and Vennix (2005) showed that interventions that focused on stock and flow diagramming and more complex system dynamic tools were met with mixed results. In some cases, these interventions actually seemed to confuse the students, resulting in poorer systemic thinking. Maani and Maharaj (2004) showed that improved decision making resulted from students’ ability to apply higher level systems thinking skills. Specifically, skills related to seeing the dynamic thinking, closed-loop thinking, and operational thinking (de- fined as thinking in terms of causality rather than correlation). The findings of these three studies are particularly relevant to this article. The Kainz and Ossimitz study
Atwater and Pittman 279
provides support that substantial improvement in systemic thinking abilities can be achieved through fairly short, well-focused exercises that could be incorporated within an existing course. Combining the Pala and Vennix study with the Maani and Maharaj study provides support for focusing on higher-level systemic thinking tools. The two studies show that students can more easily understand these con- cepts and, in many cases, learning them will improve decision making. In the subsequent sections, we will describe tools that can help students develop higher- level systemic thinking skills and describe methods that have been used to teach these tools.
TOOLS FOR DEVELOPING SYSTEMIC THINKING SKILLS
How do you help develop a person’s higher-level systemic thinking skills? No one really teaches another person how to think. However, people can be taught various tools and techniques that help the mind focus its attention in a specific way. Over the years, many tools have been developed to assist people in performing analytical thinking. For example, the scientific method is a classic process designed to help researchers isolate variables and study how they impact a phenomenon of interest. Over the years, virtually every field has evolved its own set of analytical thinking tools. Similarly, the field of systems thinking has developed specific tools to help people think systemically.
In the field of business, several tools have been developed to help people understand better how parts of a system interact. For example, Value Stream Map- ping was developed within the area of lean manufacturing to help managers trace workflows, to see how the various parts of the process interact and where value- adding activities are performed. Accounting has developed the Balanced Score Card to help managers assess the performance of a business across multiple di- mensions simultaneously. Policy Deployment was developed within the field of strategic management to help understand how various initiatives can be deployed horizontally and vertically across an organization. Quality Function Deployment was developed within the field of quality management to integrate the voice of the customer into the product or service development process. These are only a few of the several tools that have been developed in every area of business to help man- agers think holistically. While these tools are all useful for helping managers think holistically, they do not really address the other two aspects of systemic thinking; feedback loops and behavior over time. Failing to address these aspects of system behavior helps to explain the mixed results organizations have had when applying the previously mentioned tools. Fortunately, specific techniques have been devel- oped to help people with these aspects of systemic thinking. Causal Loop Diagrams (CLDs) and System Archetypes are two tools that help students improve their skills in these areas. The next two sections provide a brief description of each tool.
Basic Causal Loops
Everyone involved with a specific situation has a mental image of how they believe the key elements in that situation interact. This mental image is often referred to as a mental model (Senge, 1990). The CLD is a tool that enables individuals to communicate their mental models to others in a way that can be easily understood.
280 Facilitating Systemic Thinking in Business Classes
As people learn how to develop CLDs, they improve their ability to organize and articulate their own understanding about the behavior of the systems they operate within. As individuals share and discuss their diagrams with others, they are able to acquire multiple perspectives on the phenomena and refine their understanding of the real system’s behavior.
A brief description of how to draw a CLD is provided. Readers who are interested in a more complete description of the process are encouraged to consult Anderson and Johnson (1997), Maani and Cavana (2000), and Sterman (2000). CLDs are composed of two basic types of causal loops. One causal loop is referred to as a reinforcing loop and the other is a balancing loop. These basic loops are created when two or more variables are linked together using arrows, which result in a closed loop. In each two-variable link, the variable at the back of the arrow is said to cause a change in the behavior of the variable the arrow points to. The type of change is depicted using either a + or a − sign. A + sign means that the two variables change in the same direction and a − sign means that the two variables change in opposite directions. For example, if two variables are linked by an arrow with a + sign, it means that an increase in the cause variable results in an increase in the effect variable. Similarly, two variables linked by an arrow with a − sign is read as an increase in the cause variable, resulting in a corresponding decrease in the effect variable. Figure 1 shows a simple diagram of a closed loop that would be read as follows: An increase in Variable A causes Variable B to increase. The increase in Variable B causes Variable C to increase. Finally, to close the loop, we see that an increase in Variable C causes Variable A to increase more.
The loop depicted in Figure 1 is referred to as a Reinforcing or Amplifying Loop, because the variables involved interact to encourage continuous change in the same direction. A reinforcing loop occurs when all the causal links have a positive relationship or when there is an even number of causal links with a negative rela- tionship within the loop. The picture in the middle of the circle showing a snowball running downhill signifies that this is a reinforcing loop. A reinforcing loop creates either a desirable or virtuous cycle or an undesirable or vicious cycle. For example,
Figure 1: Basic causal loop diagram.
Variable C
Variable A
Variable B
+
+
+
Atwater and Pittman 281
Figure 2: Reinforcing loop of stock value.
Variable C: Value of a
Company X’s Stock
Variable A:
Confidence in
Company X
Variable B: Demand for Company X’s Stock
+
+
+
Figure 3: Balancing loop of stock value.
Variable C: Demand for Company X’s Stock
Variable A: Current Trading Price of Company X’s Stock
Variable B: Percieved Relative Value of Company
X’s Stock
+
–
+
True market Value of Company X’s Stock
+
Figure 2 illustrates a reinforcing loop related to a company’s stock value. If for some reason there is an increase in the confidence that people place in a company, then that would cause a demand for the company’s stock to also increase. The in- creased demand would correspondingly increase the company’s stock value. Often the increase in stock value causes others to increase their confidence in the com- pany and also demand the company’s stock, and so on. This is a classic virtuous cycle (at least for the company in question). Note, however, the cycle can just as easily go in the opposite direction if people lose confidence in a company. In this case, the vicious cycle starts as people sell off the stock, lowering the price and causing others to lose confidence and sell their stock, and so on.
The recent stock market “bubble” and subsequent stock market sell-off (cor- rection), clearly illustrate that there are more factors influencing the dynamics of stock prices than those illustrated in Figure 2. These additional factors often have a slowing effect on reinforcing loops. These result in the other basic type of loop called the Balancing Loop. Balancing loops occur when there are an odd number of negative links in the loop. Figure 3 shows a balancing loop related to stock value.
282 Facilitating Systemic Thinking in Business Classes
Note in Figure 3 that another variable has been added that is not within the loop. This variable acts as a goal or desired state that the loop aspires to attain. In this example, the goal is the true market value of a company’s stock. As the true market value of a company’s stock increases, that causes the perceived value of the company’s stock relative to its current trading price to increase. This in turn causes the demand for its stock to increase, which causes its price to increase. However, when the trading price increases, this causes the perceived value of the stock relative to its true market value to decrease. Reading through the diagram, it can be seen that an initial in- crease in any variable will loop through the diagram ultimately causing Variable B to decrease. In fact, each complete cycle through the loop shows that an initial change in any variable ultimately works to create the opposite impact of that initial change. In this example, there was one negative link in the loop, but the same phenomenon occurs whenever there is any odd number of negative links within a loop.
Basic reinforcing and balancing loops can be joined together to form more complex feedback interactions. These more complex combinations often result in the counterintuitive behavior described by Forrester (1971). In some cases, these combinations of reinforcing and balancing loops have been identified in several real-world situations. Consequently, their behavior, along with methods for in- fluencing them, is better understood. These common complex combinations have become known as System Archetypes.
System Archetypes
CLDs can be used to create System Archetypes. System Archetypes are generic structures of common patterns that occur repeatedly in both personal and profes- sional settings (Senge, 1990). These archetypes illustrate classic scenarios that commonly occur in systems. Because they represent the common system behavior, archetypes are often used as the foundation for customized CLDs developed by a group investigating a specific situation within an organization. Archetypes are an excellent tool to transition from an elementary understanding of causal loops to systemic thinking. While reinforcing and balancing loops can be used to effec- tively explain a specific phenomenon, archetypes can be used to provide a more complete description of the interactions occurring within a system. Once people become aware of these archetypes, they are often able to see them at work within their systems and use them to explain counterintuitive outcomes and identify the best actions for improving system performance. There are nine commonly used archetypes, which include:� Fixes that Fail� Shifting the Burden� Escalation� Drifting Goals� Limits to Success� Success to the Successful� Growth and Underinvestment� Accidental Adversaries� Tragedy of the Commons
Atwater and Pittman 283
Figure 4: Causal loop diagrams of the nine basic archetypes.
Figure 4 provides the generic CLD for all nine of the basic archetypes. A detailed description of all nine of these archetypes is beyond the scope and intent of this article; however, a brief explanation of three of these archetypes, along with an example of how it could play out in a business setting, is provided here. A more extensive coverage of these tools can be found in Senge (1990), Senge et al. (1994), and Maani and Cavana (2000).
284 Facilitating Systemic Thinking in Business Classes
Fixes that fail
The Fixes that Fail archetype is also sometimes referred to as Fixes that Backfire. This archetype can be used to illustrate and understand situations where there is a trade-off between the short- and long-term consequences of our actions. Specifi- cally, this archetype occurs when we use a quick fix to address a problem. In many cases, decision makers may even be aware of the negative consequences result- ing from this quick fix, yet still continue because of the difficulty and/or delayed benefits of applying a better corrective measure.
The generic CLD illustrating this archetype can be found in Figure 4a. The archetype is composed of one reinforcing and one balancing loop. The balancing loop in the CLD for this archetype illustrates how a solution is used to temporarily bring a problem situation back into a more desirable state. Unfortunately, there is an unintended consequence of this solution that creates a reinforcing loop which makes the problem worse or reinforces the original problem.
Successfully overcoming situations illustrated by the Fixes that Fail archetype requires recognizing the role that time delays play in whichever solution is applied. If the quick-fix solution is chosen it will eventually, and inevitably, create a worse situation. On the other hand, implementing the better solution will also have a delayed impact. In the short term, things may actually get worse, but eventually the benefits from pursuing the longer-term solution will kick in and significantly improve the situation.
There are many examples of this archetype playing out in business. One classic operations management example of this archetype occurs when a production system consistently has poor delivery performance. Often, short-term remedies include the use of overtime or stocking more inventory. While these solutions may alleviate the short-term problem, they inevitably increase production costs and do nothing to make the company more competitive in the long run. However, if a company implements an improvement program such as Lean or Six Sigma, there is an initial investment and a period in the beginning while problems are being surfaced that makes the return on investment look very bad. Nevertheless, there is ample evidence that if companies stay on course, the long-term benefits far outweigh the short-term pain.
Escalation
As its name implies, an Escalation archetype occurs whenever two or more parties in a system view each other as a competitive threat and react to one another in a way that creates a continuous increasing spiral. The Escalation archetype is illustrated in Figure 4c with two counterbalancing loops. Each loop is identical and represents one party’s behavior in the scenario. In each case, one party takes an action, which causes the other party to feel threatened so that they respond with a similar action. This causes the original party to react again. This action/reaction process continues to seesaw between the parties, creating a continuing escalation in behaviors. Effectively managing this archetype lies in uncovering and addressing the underlying assumptions perpetuating the perceived threat. The solution to this scenario lies in at least one party looking for a solution to the situation other than the repeated escalation of the same action.
Atwater and Pittman 285
Probably, the most famous example of this scenario was the escalation in the arms race between the United States and the Soviet Union during the Cold War. However, the phenomenon also occurs frequently in business. One example of this archetype occurring in business is when a company uses a price incen- tive or gift to induce customers to try their product rather than developing a sus- tainable competitive edge through product innovations or superior relationships. Initially, the inducement increases the number of customers buying the product, which makes the tactic appear to be working. However, eventually competitors catch on to the inducement idea and copy it in some way. The initial company responds with yet another inducement, and the cycle repeats itself. The end re- sult is a very fickle customer base that repeatedly switches suppliers while the companies in the industry reduce their product to a commodity and diminish their profitability.
Limits to success
The Limits to Success archetype is sometimes referred to as Limits to Growth. Figure 4e illustrates the Limits to Success archetype, which includes one reinforcing and one balancing loop. The reinforcing loop represents a virtuous cycle where an action taken by the system creates a desirable outcome. The company continues to pursue this action with increasing success until, for some unforeseen reason, the beneficial results diminish, plateau, and/or even decline. The unexpected decrease in the results is depicted in the balancing loop on the right side of the diagram. The balancing loop has an external variable that represents some limiting condition, which has a slowing effect on the successful action. The Limits to Success archetype is based on the fact that all systems have factors that limit their rate of improvement. Furthermore, if these factors are not recognized and addressed, they can even cause the organization to collapse under its own success.
The Limits to Success archetype dovetails well with a popular business im- provement method known as the Theory of Constraints (TOC). Dr. Eli Goldratt, the founder of TOC, emphatically states that all systems have constraints. Fur- thermore, both the developers of the system archetypes and Dr. Goldratt stress that when faced with this phenomenon, managers should focus on removing the constraint rather than pursuing the successful activity more vigorously. A com- mon business example of this archetype in action occurs when the results of a company’s improvement program start to falter. Often, this is the result of a shortage of a resource, or increased dependence on another part of the system. In any case, managers need to surface the limiting condition and develop an ap- propriate plan of action for dealing with it before pursuing further improvement efforts.
Both CLDs and System Archetypes are tools designed to assist people in thinking systemically. They are not specific to any functional area such as opera- tions, accounting, finance, and so on. These are the types of tools that will enable managers to understand better the systems they work within and communicate that understanding to others. Developing the skills to use these tools will also help managers facilitate the lifelong learning requirement of working in complex social systems. In the next section, we discuss how the tools can be introduced into a
286 Facilitating Systemic Thinking in Business Classes
graduate-level business course and subsequently used to enhance classroom dis- cussions and help student gain a more systemic perspective on an issue.
TEACHING CAUSAL LOOP DIAGRAMMING
CLDs are the natural starting point for teaching these aspects of systemic thinking for two reasons. First, learning the mechanics of drawing these diagrams is fairly simple and typically can be handled in less than one class period. (Of course, learning to use them and mastering their use are two separate issues.) Second, experience with a vast number of students has shown that they tend to be intrigued with the concept of CLDs and enjoy using them to describe their mental models.
Being able to quickly teach the tool is important, because systemic thinking is typically not the focus of the course, but simply a tool to enhance deeper under- standing of the topic under study. One method for carrying out this introduction is provided here. Generally, a brief lecture (20–30 minutes) is sufficient to introduce the fundamentals (i.e., positive and negative causal links and basic reinforcing and balancing loops). Next, a scenario is selected that provides the details of an in- dustry cycle, a counterintuitive outcome, or a failed change effort. In selecting the scenario, there are three characteristics that are generally desirable. First, the topic should align well with the course being taught to facilitate teaching the subject and further highlight that this is a tool to enhance their understanding of the topic, not a different topic. Second, the topic should be something the students can easily understand and draw from their own knowledge to help construct a CLD. This may mean that the scenario provides enough detail for the students to get a solid background on the events described. Third, if possible, it should be taken from current events. This will help the students appreciate the prevalence of feedback loops and counterintuitive results.
One scenario that follows these parameters is an exercise used in an operations management course when teaching inventory management and its relationship to production capacity, sales, and marketing. Using an article database, the students are asked to review the Wall Street Journal to search for articles relating to sales in the automotive industry for the past 4–5 years. In the class, the article headlines and their dates are recorded on the board and then arranged sequentially by date. Very quickly, the students notice the repetitive nature of the headlines as they cycle across slow sales, increased inventory, concern about the excess inventory, production cutbacks/downsizing, the introduction of sales incentives, and finally increased sales. The students are then placed into groups of 3–4 and some groups are asked to draw the CLD that focuses on the relationship between inventory and sales incentives. Other groups are asked to focus their CLD on inventory and production cutbacks. Typical CLDs reflecting the students’ mental model of these cycles are shown in Figure 5. Once these diagrams are drawn on the board, they can be used to stimulate class discussion about other factors that could be playing a role in these cycles. In addition, they are asked to discuss why the cycle continues to repeat itself every year and what could be done to break out of it. This discussion is an excellent lead-in to lean manufacturing techniques and supply chain management.
Atwater and Pittman 287
Figure 5: Simple causal loop diagrams of automotive sales and inventory dynamics.
Automotive Sales
Inventory
Concern About Inventory
Use of Sales Incentives
–
+
+
+ Automotive Sales
Inventory
Concern About Inventory
Production Cutbacks –
+
+
–
TEACHING SYSTEM ARCHETYPES
System Archetypes are a natural next step after learning how to draw and work with CLDs. Unfortunately, students typically struggle when moving from working with a single causal loop to multiple interacting loops. In addition, many have difficulty with the idea that there is no perfect representation of reality. Consequently, prior to working with the archetypes, students are assigned readings on the topic and there is another brief class lecture (approximately 30 minutes) reviewing them. Even with this preparation, the potential exists for students to become frustrated when they first start working with this tool. In order to help students work through this frustration, system archetypes should first be applied to simple, fairly well defined situations. It is generally a good idea to have the students work in groups as they learn these concepts. Often these classes can be quite lively and jovial as students work through example scenarios.
Here again, as a first step, we suggest presenting current examples that can only be fully understood through systemic analysis. Examples can come from the popular media including The Wall Street Journal, local papers, Time, Newsweek, and so on. Look for key words appearing in the articles, such as reinforcing or vicious cycle, unexpected outcome, complexity, long-term consequences/change, and so on. In addition to the specific examples previously illustrated in this article, more general systemic phenomena can also be used to begin the discussion such as the Israeli–Palestinian conflict, childhood obesity, rising healthcare costs, and so on.
The next step is to get students to identify examples where system archetypes can be applied. An excellent reading and source for a group exercise is “Laws of the Fifth Discipline” (Senge, 1990, pp. 57–67). Each student is asked to review these 11 laws and think of at least one example that illustrates each law in action. The students are placed into groups of 3–5 members and asked to share the examples they have identified. Finally, each group is asked to share their best examples for each law. This in-class activity should include 10–15 minutes of group discussion time and 20–30 minutes of class discussion time. As a subsequent homework assignment, students can be asked to identify the system archetype they feel best applies to the examples provided by the groups.
288 Facilitating Systemic Thinking in Business Classes
Once the students have an understanding of the basic terminology and con- cepts related to the archetypes, we suggest the following process for teaching them how to apply system archetypes to specific real-world situations:
1. Assign students to read an article or short case outside of class and ask them to identify the archetype which they feel best fits the sce- nario being studied. Have them develop a written description of the archetype as applied to the scenario.
(a) It is important to have the students carry out this exercise inde- pendently, because that will prompt further reflection of both the important systemic aspects of the scenario and provide richer subsequent group discussions.
(b) The “Archetype Family Tree” (Senge, Kleiner, Roberts, Ross, & Smith, 1994, p. 150) is an excellent tool to help students select the archetype that most closely matches the situation.
2. In class, have students break into groups of 3–5 and share the archetypes they developed independently, explaining both their logic in selecting the particular archetype and its application to the scenario.
3. Have each group select which archetype they feel best fits and work together to develop the archetype so that it incorporates the ideas of the participants. Creating this group archetype results in team learning and a far deeper systemic understanding of the case being studied.
4. Select 2 or 3 groups to present their archetypes to the rest of the class.
5. Require those not presenting to play an active role. Specifically, ask them to identify what the presenting group did well and offer sugges- tions on how the archetype can be improved.
The first step should be performed outside of the class to assure proper reflection and preparation. The remaining four steps typically require about 60–75 minutes of class time.
Opportunities to apply archetypes to real-world scenarios appear in the news on a daily basis. The Enron debacle is one such example, which we will use to illustrate the above process. One aspect of the Enron case that has been used to help students learn to apply system archetypes relates to Enron’s extensive use of limited liability partnerships (LLPs) to absorb debt. Two articles that work well for reference information in this exercise are “The Enron Debacle: Byzantine deals have shattered the energy outfit’s credibility” (Forest, Zellner, & Timmons, 2001) and “Why Enron Went Bust: Start with arrogance, Add greed, deceit, and financial chicanery. What do you get? A company that wasn’t what it was cracked up to be” (McLean, Varchaver, Helyar, Revell, & Sung, 2001).
Through reading these articles on Enron, students quickly realize that the use of the LLPs were a creative short-term solution, used to correct the problem of bad decisions resulting in debt. In fact, the investors in these LLPs were predominantly internal to Enron and profited handsomely. As the convenient use of these LLPs increased, the long-term recognition that significant bad business decisions were being made became less obvious. Over time, Enron’s dependency on these LLPs
Atwater and Pittman 289
Figure 6: The Enron debacle—shift the burden archetype.
Creation of limited liability parrtnerships to absorb
debt
Enron losses
Recognition of mistakes
& poor decisions
Increase in market value of stock
–
+
–
+
–
Delay 1
Delay 2
+
continued to grow and their ability to face the mounting consequences for bad decision making became nearly impossible. Those in key leadership positions aware of the dangerous potential consequence of the LLPs either tried harder to make them work or quit their jobs, recognizing the significant threat to the future of the organization. These fears were ultimately realized in the sudden collapse of Enron. As the students work with the material, they begin to realize that it is an excellent example of the Shifting the Burden archetype. An example of the type of archetype developed by students for this case is provided in Figure 6.
COMMON STUDENT ERRORS
Students make several errors as they learn how to use CLDs and system archetypes. The errors fall into two categories. The first category is composed of mistakes related to the mechanics of drawing CLDs and archetypes. Students often, initially, put a minimal number of causal elements in their archetypes, fail to designate the direction of the causal connections (i.e., + or −), and fail to identify expected delays between some causes and effects. These issues can be easily overcome through faculty or peer feedback, including specific recommendations for improving the archetype presented.
The second category is composed of errors that are conceptual in nature. Here, students often use their diagrams to prove a predetermined conclusion or blame individuals for problems that exist in a system. Students must be made to understand the importance of representing all sides of a situation as realistically as possible. It is only through the most honest and accurate depiction of reality that the real dynamics of the system can be understood. One suggestion is to have students identify as many motivations for behavior as possible and develop CLDs that use the different motivations. In addition, instructors can assign students the responsibility for advocating a position that they disagree with. In these assignments, reinforce that students must learn to see multiple sides of an issue, because an unfair advocacy
290 Facilitating Systemic Thinking in Business Classes
of one part of the issue will not help them recognize and prepare for potential unintended consequences. In addition, the focus should be on using the concepts to help students see that behavior is generally motivated by the structure of the system (i.e., focus on short-term performance, faulty reward systems, conflicts in resource needs, unrealistic expectations, etc.) rather than wanton sabotage. In other words, the focus should be on the cause of the behavior, not the individuals themselves.
CONCLUSION
There were several factors that motivated the writing of this article. First, we wanted to differentiate between teaching systemic thinking and teaching students about systems. While the latter is an important aspect of students’ learning experience, it does little to prepare them for handling decision making in complex adaptive systems (e.g., business). Furthermore, we wanted to highlight the three dimensions of systemic thinking (i.e., synthesis, behavior over time, and feedback loops). While most business functions do highlight the importance of understanding the interaction of parts and have even developed tools to facilitate managers carrying out this activity, there is relatively little discussion about the other two aspects of systemic thinking.
Second, we wanted to make the case for the importance of developing sys- temic thinking skills in managers. We feel this fact cannot be stressed strongly enough for two reasons. First, while business has always been a system, recent business trends, such as lean thinking, supply chain management, globalization, outsourcing, and so on, have increased the interconnectedness of businesses and their subsystems. There is every reason to believe that these trends will continue and even increase in the future. Consequently, students must learn how to work more effectively within these systems. Second, several studies have shown that most people do not naturally think systemically, particularly with regard to time delays and feedback loops. Therefore, students must be trained to use tools and techniques that will help them.
Finally, we wanted to describe two tools that are very useful in stimulating systemic thinking and provide examples of how they can be introduced and used. Decisions made in any business area will always have ripple impact across and even outside the organization. Consequently, the need to understand and apply systemic thinking skills transcends a single course or business area. Indeed, the idea that systemic thinking should be relegated to a single course is very non- systemic. With this in mind, the tools presented in this article were meant to be non-function specific, so that they could be used in a variety of business classes. In addition, the exercises are simply illustrative of ways that the tools can be introduced and we are confident similar exercises can be developed for any business function.
The evidence that students need to develop systemic thinking skills is quite strong. This article presents a method that provides a way for professors to cover their content and still expose students to systemic thinking, hopefully in a way that is synergistic. Undoubtedly, business education will evolve to accommodate the demands of the system age. Hopefully, this article will serve as a stimulus for further discussion and research on this important topic.
Atwater and Pittman 291
REFERENCES
Ackoff, R. L. (1981). Creating the corporate future. New York: Wiley.
Ackoff, R. L. (1994). The democratic corporation: A radical prescription for recre- ating corporate America and rediscovering success. New York: Oxford Uni- versity Press.
Ackoff, R. L., & Gharajedaghi, J. (1996). Reflections on systems and their models. Systems Research, 13(1), 13–23.
Anderson, V., & Johnson, L. (1997). Systems thinking basics: From concepts to causal loops. Waltham, MA: Pegasus Communications Inc.
Booth Sweeney, L., & Sterman, J. D. (2000). Bathtub dynamics: Initial results of a systems thinking inventory. System Dynamics Review, 16(4), 249– 286.
Deming, W. E. (1994). The new economics for industry, government, education. Cambridge, MA: MIT Center for Advanced Educational Services.
Dörner, D. (1996). The logic of failure: Recognizing and avoiding error in complex situations. New York: Basic Books.
Forest, S. A., Zellner, W., & Timmons, H. (2001). The Enron debacle: Byzantine deals have shattered the energy outfit’s credibility. BusinessWeek, November 12.
Forrester, J. W. (1971). The counterintuitive behavior of social systems. Technology Review, 73(3), 52–68.
Forrester, J. W. (1994). Learning through systems dynamics as preparation for the 21st century. Keynote Address for Systems Thinking and Dynamic Modeling Conference for K-12 Education. Concord Academy, Concord, MA.
Kainz, D., & Ossimitz, G. (2002). Can students learn stock-flow-thinking? An empirical investigation. Proceedings of the 20th International Conference of the System Dynamics Society. Polermo, Italy.
Maani, K. E., & Maharaj, V. (2004). Links between systems thinking and complex decision making. System Dynamics Review, 20(1), 21–48.
Maani, K. E., & Cavana, R. Y. (2000). Systems thinking and modelling: Understand- ing change and complexity. New Zealand: Prentice Hall, Pearson Education.
McLean, B., Varchaver, N., Helyar, J., Revell, J., & Sung, J. (2001). Why Enron went bust: Start with arrogance. Add greed, deceit, and financial chicanery. What do you get? A company that wasn’t what it was cracked up to be. Fortune, 144(12), 58–68.
Ossimitz, G. (2002). Stock-flow-thinking and reading stock-flow-related graphs: An empirical investigation in dynamic thinking abilities. Proceedings of the 20th International Conference of the System Dynamics Society. Polermo, Italy.
Pala, Ö., & Vennix, A. M. (2005). Effects of system dynamics education on systems thinking inventory task performance. System Dynamics Review, 21(2), 147– 172.
292 Facilitating Systemic Thinking in Business Classes
Richmond, B. (2000). The thinking in systems thinking: Seven essential skills. Waltham, MA: Pegasus Communications Inc.
Senge, P. M. (1990). The fifth discipline: The art, & practice of the learning orga- nization. New York: Doubleday/Currency.
Senge, P. M., Kleiner, A., Roberts, C., Ross, R., & Smith, B. (1994). The fifth discipline fieldbook: Strategies and tools for building a learning organization. New York: Doubleday/Currency.
Sterman, J. D. (2000). Business dynamics: Systems thinking and modeling for a complex world. Boston: Irwin-McGraw-Hill.
Wickens, C. D. (1992). Engineering psychology and human performance (2nd ed.). New York: HarperCollins.
J. Brian Atwater is an associate professor of production/operations manage- ment at Utah State University. He earned his PhD in operations management at the University of Georgia. He currently teaches graduate courses in continuous improvement techniques and system dynamics. He is certified in Production & Inventory Management by the American Production & Inventory Control Society (APICS). He is a certified academic associate (JONAH) of the Goldratt Institute. He has also worked as an examiner for the Shingo Prize for Excellence in Manufac- turing. His current interests center around the teaching of systemic thinking and the integration of those concepts with other problem-solving approaches such as cre- ative problem solving, lean thinking, six sigma, and TOC. He has published several articles in a variety of journals, including the Production & Operations Manage- ment Journal, International Journal of Production Research, International Jour- nal of Operations & Production Management, Productions Inventory Management Journal, Cost Management Journal, and Industrial Management Journal.
Paul Pittman is Professor of Operations Management at Indiana University South- east. He earned his PhD in Production and Operations Management at the Uni- versity of Georgia. He is certified in Production & Inventory Management at the Fellow level and is certified in the Theory of Constraints. His publications include articles in International Journal of Operations and Production Management, The Journal of Learning in Higher Education, Journal of Educational and Psycholog- ical Consultation, and the Journal of Systems Improvement, along with being the coauthor of numerous APICS certification books and educational materials. He has earned numerous teaching honors including his university’s Distinguished Teach- ing Award. He is Chair of the Certified in Production and Inventory Management program for APICS, The Association for Operations Management, and a principal partner of The LAMP Group which assists organizations in learning and applying the concepts of systems thinking, decision making, quality improvement, project management, and organizational learning
