Understanding Systems

 Understanding Systems

In the previous chapter, we asked the question ‘What is a system?’ and defined it as a collection of interacting parts, components, or actors in which the interactions result in system-level properties and behaviours not attributable to the sum of individual parts. With this definition, we also introduced several concepts that underpin systems thinking, including hard and soft systems; open and closed systems; complex vs. complicated systems; systems behaviour and emergence; systems problems (tame, messy, and wicked); worldview; and self-organization. In this chapter, we will build upon these concepts and definitions to provide you with a deeper and more complete understanding of systems, their structure and properties.

LEARNING OUTCOMES

• Understand
• the function and purpose of a system
• systems inputs and outputs
• system performance
• stocks, flows, and forces
• system structure and boundary
• interconnections and feedback loops
• processes in the context of systems
• entropy and homeostasis
• Able to conceptualize relatively complex systems visually and verbally
• Identify and discuss the factors that explain behaviours of simple systems

3.1 Function and purpose

Based on our discussions in the previous chapter we could surmise that systems have properties. The purpose and function of a system are two important properties that help us to understand systems. The purpose of a system is what it does, and the function is how it delivers this purpose. This is best explained through the following example. The purposes of an aeroplane and a horse and cart are all to transport people from A to B, but the ways in which they transport people to their purposes are the same. But the ways in which they transport people from A to B are quite different – this is the function of the system. Other examples include a digestive system which processes food (function) to produce energy for the body (purpose) or an irrigation system which directs water through canals (function) to irrigate land (purpose).

Although there is general agreement that all systems, however complex or simple, have a function, there is some debate as to whether all systems have a purpose. What is the purpose of the solar system? We know what its function is (to keep the planets within the system balanced), but for what purpose? From these simple examples we can observe that in some systems, such as the solar system, we are not sure of its purpose, i.e. why it exists, even though we can scientifically explain how it functions. In systems thinking, some argue that whilst all systems have a function and a purpose, it is a purpose. Others argue that all systems have a function and a purpose; it’s just that we may not know or understand the purpose of the system; the solar system may have a purpose but at this stage we do not know or understand what this purpose may be.

In business and management most of the systems we deal with usually have a function and a purpose, thus this tension between function and purpose may be a moot point but it is an important one to bear in mind as not everyone who is involved with the system may understand its function or its purpose. In fact, in many cases, in complex organizations a system may have multiple purposes depending on the viewpoints/worldviews of its stakeholders. For example, what is the purpose of a university? Students and their parents may say that it is to provide them with a good education that enables them to get good jobs. Industry may say that the purpose of a university is to create new knowledge and technologies through research and development that allows them to improve their business. The local government may say it is to fuel economic development by creating spinouts and supporting enterprises. The central government may say it is to create revenues from foreign students paying fees. Although these views appear radically different, there are some overlaps amongst them and in our experience they are valid views, which we have encountered in the past.

In this context it is worth thinking about the function and purpose of a system using the following structure. A system does something (function) for a particular purpose and even though its function may be known and understood, its purpose may not be so clear.

3.2 Inputs and outputs

Unless a system is completely closed it interacts with its environment through inputs and outputs. Even in completely closed systems, its subsystems will interact with each other through inputs and outputs. In short, it would be safe to say that all systems have inputs and outputs. Following from our earlier examples an information processing system takes one form of information as an input, it processes the input and outputs it in a different form. For instance, a digestive system takes food as an input and produces energy and waste as outputs. In a similar vein, the primary input to the solar system is energy from the big bang, which arguably is still continuing, and its output is the motion of the planets, moons, etc. that serve to keep that solar system in balance.

Examining this in further detail we will observe that there are different kinds of inputs and outputs to a system. We could categorize and define these as follows:

.Inputs are processed through the system and leave the system in a changed state. For example, a management information system receives raw data (input) and produces management reports (output).

.Resources (also a form of input) are consumed by the function of the system. They include energy, time of people, and other resources and assets; for example, when a machine is used for a time, it consumes part of its useful life.

.Controls (also a form of input) set the expectations, standards, or requirements a system should fulfill. In effect, these define the parameters within which the system needs to function.

.Outputs are produced by the system in line with its function and, where known, its purpose. Outputs may be useful when they are consistent with known, its purpose.

3.3 System performance

In simple terms performance has two dimensions: efficiency and effectiveness. Thus, performance of a system is broadly defined as the effectiveness and efficiency of a system. In this context, effectiveness is defined as the extent to which the output fulfills the specification, requirements or expectations defined by the controls. Whereas efficiency is defined as the resources consumed to produce the outputs.

Of course, if the other unintended outputs and/or waste are being produced along with the intended useful outputs, we would also need to consider the resources that are consumed to produce these outputs when we are thinking about efficiency. However, we must bear in mind that not all unintended outputs are waste; in some cases an unintended output from a system may be quite useful.

At this point it would also be useful to discuss the term efficacy, which is commonly used in systems thinking. Efficacy is similar to effectiveness and they are often used interchangeably. However, there is a difference between the two. For example, in understanding effectiveness of vaccines the term efficacy is used to measure the effectiveness of the vaccine in a laboratory setting, whereas the term effectiveness is used to measure its performance in the real world. Thus, a vaccine may have 90 per cent efficacy in the laboratory but only 60 per cent effective in the real world, such as the influenza vaccine. When we apply the same thinking to systems, efficacy is concerned with the potential performance of the system, and arguably when we ask about the efficacy of the system we are interested in, i.e. ‘is this the best performance?’.

3.4 Stocks, flows and forces

When discussing the inputs and outputs of a system we implied that things flow through a system. That is indeed the case. Let us look at different systems to understand what sort of things flows through the system.

In a manufacturing system, raw materials and components arrive as inputs and are processed through the system into products that leave the system as outputs. In this example, materials are flowing through the system.

In an information processing system, such as processing of mortgage applications, information arrives into the system as an input in the form of a mortgage application; the information is then processed, and the output is also information in the form of a decision. In this example, information is what flows through the system.

In a customer experience system, such as a cruise package, customers arrive as inputs to the system; they are processed throughout the cruise (entertainment, games, activities, dining, excursions, etc.), and then they become outputs of the system hopefully as healthier, more relaxed, and entertained customers.

In an electrical distribution system, energy, in the form of electricity, flows through the system.

Based on these rather simplistic examples we could surmise that a variety of things may flow through a system: materials, information, customers, people, energy, and so on. In reality, many things flow through a system to enable the system to function effectively and efficiently. For example, in the manufacturing system example, apart from materials, there are many other things that flow through, including:

.information to make the products;

.energy to run the machines and the factory;

.money to finance inputs (raw materials, components, energy) and the resources (people, equipment, consumables);

.customers who may interact with the manufacturing system at different points (such as specification, design approval, testing, acceptance, etc.).

Clearly, these flows need to come from somewhere; they do not just materialize from nothing. They come from stocks. Stocks and flows are concepts commonly used in systems thinking but they are also prevalent in the disciplines of engineering, business, accounting and economics. A stock is defined as a quantity existing at a given point in time. For example, today you might have £100 in your bank account. It may have accumulated in your account over a period of time; you might have been receiving £10 every week for the past 10 weeks. In this case the flow is £10 per week resulting in the stock of £100.

£100 that you have in your bank account today. Of course, if you start spending this money it creates a new flow, which may reduce the stock.

In short, stocks and flows are concepts used in systems thinking to understand the dynamics of a system over time. Stocks are measures of quantity at a given time (we have £100 today), whereas flows are measures of quantity over time (£10 per week over 10 weeks).

So far, the examples we have provided for stocks and flows have been for tangible things (money, materials, information, energy); however, in systems thinking intangible things (knowledge, goodwill, resentment, misunderstanding, etc.) can also be represented as stocks and flows. For example, several unpopular decisions taken by the management of an organization may result in the build-up of some resentment within the workforce that may in turn result in unexpected and/or unhelpful behaviors. In fact, often it is the tangible stocks and flows that help us to describe hard systems, and the intangible stocks and flows, sometimes referred to as forces, that help us understand soft systems (discussed later in this chapter).

Later in this book when we start discussing systems dynamics we will build upon the concept of stocks and flows. At this stage we will refrain from going into further detail and instead summarize what we have learned so far about the structure of a system.

3.5 System structure

In Figure 3.1 we have attempted to summarize the concepts introduced earlier in this chapter to define the structure of a typical system. According to this illustration the system has a purpose and a function. It has inputs and outputs. It has controls that govern the system and it uses resources to carry out its function and transform the inputs to outputs. While carrying out its function, things (materials, information, people, energy, etc.) flow through the system. These flows come from stocks which either exist within the system or come from outside the system. We can conceptualize and measure the performance of the system as effectiveness and efficiency of the system.

In the previous paragraph, when we say "the flows come from stocks which either exist within the system or come from outside the system," it implies that the system has a boundary and that some flows come from within the boundary and some from outside the boundary. This has further structural implications that are not captured in Figure 3.1. If we consider the definition of a system as ‘a collection of interacting parts, components or actors in which the interactions result in system-level properties and behaviours not attributable to the sum of individual parts’, introduced earlier, and the notion that ‘nothing in this world exists in isolation, everything is connected to something else and everything is part of something larger’, discussed in the previous chapter, it suggests that a system consists of interconnected parts, which are systems in their own right (i.e. subsystems) and that a system exists as part of a larger system as depicted in Figure 3.2. In this context the higher-level systems are systems of systems, and the lower-level systems are subsystems or systems within systems.

Examples of this hierarchical system structure are illustrated in Figure 3.3 showing two perspectives. Figure 3.3a illustrates how individual components are parts of a car, which in turn are parts of an enterprise system, which in turn is part of an industry system, which in turn is part of an economic system and so on. In contrast, Figure 3.3b illustrates the administrative systems starting from business processes and moving on to business units as higher-level systems that in turn exist as part of the overall business, which in turn is part of the automotive industry. These examples provide just two of many ways we can conceptualize and structure the system we wish to understand. Clearly, organization structure (i.e. individuals, teams, departments, divisions) is an alternative systems perspective of the organization.

However, if we take a horizontal slice through any system in the hierarchies shown in Figure 3.3, we will see that there is an additional level of complexity. For example, if we consider the economic system of a region, it is not concerned with just one single industry system. The economic system of a region typically contains several industries and other systems as illustrated by the diagram in Figure 3.4. The size of each bubble on this diagram depicts the significance of the subsystem on the wider economic system, while the thickness of the lines connecting different subsystems depicts the interdependency between different subsystems.

3.6 System boundary

The discussion above brings us to the point about the boundaries of a system. Indeed, all systems have boundaries and different components or subsystems are connected to one another within these boundaries. They can also be connected to components or subsystems of other systems through these boundaries. For instance, if we are an insurance company and interested in assessing the risk of a car, we set the system boundary around the car as a mechanical object. We assess its performance, track record, technical reliability, cost of repairs, age, etc. However, if we expand the boundary of the system and include the driver, then we are assessing a different system. We will pay equal if not more attention to the experience of the driver, track record of the driver and the driver’s age and gender. According to insurance companies, these parameters will play a much more dominant role in assessing the insurance risk of the driver.

a particular car, when the driver is considered. We can extend the boundaries of a system even further and look at where the car is parked and used. In rural areas the traffic density is lower, and the risk of traffic accidents is also lower. In this context, postcode can have an impact on the insurance risk as different areas within the same city will have different crime rates, frequency of accidents, etc.

But the big question is 'how do we know where the boundary of a system lies?' Often the answer to this question is 'it depends'. In reality, the boundary of a system is an artificial concept because the components of a system are connected together, and they act in concert. Thus, the system boundary is an artificial concept to aid the analyst or the manager in conceptualizing and understanding the system. For instance, in Figure 3.4 we see that the regional economy is a function of several interacting subsystems. If we wish to understand the impact of our innovation system on the regional economy, where do we draw our boundaries? Particularly if we consider that the higher education system, through its research and development activities, could be a significant contributor to the innovation output, where do we draw our boundaries? Do we make higher education a subsystem within the innovation system? Or do we keep higher education as part of the wider education system and show the connections between the higher education subsystem and the innovation system? The answer to this question often is it does not matter; the important point is that we understand the connections between the higher education subsystem and other subsystems within the innovation system. Indeed, it is not just the connections between the innovation and higher education systems that are in question here. If we need to understand innovation as a system within our economy, we need to understand how it is connected to other systems within the economy, In short, it is the systems analyst or the manager who needs to decide where the boundaries of the system should be as long as the pertinent connections between the system that is being studied and other systems that are important to its function are understood, and the connections are included in the analysis.

3.7 Interconnections and feedback loops

At the very start of this book, we said that in this world everything is connected to everything else, and that systems thinking enables us to see through this complexity and understand why complex systems behave the way they do. We also defined a system as a collection of interacting parts, components, or actors in which the interactions result in system-level properties and behaviours not attributable to the sum of individual parts. Earlier in this section, in describing the structure and properties of a system, we demonstrated that whilst within itself the parts of a system interact with each other in its stocks and flows, the system itself also interacts with other systems in its environment. Naturally, these interactions result in the system being influenced/changed by its environment as well as the system influencing/changing its environment. Indeed, these interactions take place at different levels.

Key to understanding these relationships are the connections between the system and the parts of the system as well as the connections between the system and other systems operating within the environment of the system. These connections are not just flows that describe the quantities of things flowing from one part of the system to another; they are also causal relations between the connected parts that may affect the behaviour of a part resulting in not just local but potentially system-wide behavioural changes. This idea of this causality may be demonstrated simply through the following example:

At a given interest rate, the greater the bank balance (A), the greater the interest earned (B).

This is a very simple example of causality where A causes B. But of course, the causality in this example does not stop there because:

At a given interest rate, the greater the interest earned (B) the greater the bank balance (A).

Here we can see that there is a similar causal relationship in the opposite direction. In systems thinking, this phenomenon is known as reinforcing loops or feedback loops as depicted in Figure 3.5.

In systems, particularly in complex ones, we can observe many such loops with different characteristics. However, we will refrain from going into further detail about connections as we will develop these ideas further as we learn about different methods later in this book. At this point it will suffice to say that the connections and causal relationships are critical in helping us to understand and predict the behaviour of complex systems.

3.8 Processes and systems

A common question we get at this point is ‘what is the difference between a process and a system? ... is a system not just a process?’ This is a fair question and it usually comes from people who have been trained in process management and improvement techniques such as Total Quality Management (TQM), Lean Management, and Six Sigma. In essence, a process is a system, but a system can be much bigger than a process. In TQM, Lean, and Six Sigma the process focuses on very specific systems such as the manufacturing process, the order fulfillment process, the planning process, and so on. We would argue that whilst all processes are systems, not all systems are processes. For example, the national economic system discussed before is not a process, but comprises many subsystems and many processes within it. In short, process thinking is consistent with systems thinking, but in systems thinking we are concerned with systems of all sizes and their interconnections with other relevant systems.

3.9 Entropy and homeostasis

Finally, before we conclude this chapter, it would be pertinent to introduce two further concepts that help us explain the behaviour of systems. Earlier we defined a system as a collection of interacting parts that result in system-level behaviours. The individual parts of the system, whether they are technical components, individual people or organizations, do exhibit autonomous behaviours. In most cases the parts of a system are divergent and as a result, if left to their own devices they move towards disorder or disorganization.

This phenomenon is best explained through an example like those shown below:

Example 1. Think about a garden that is well manicured and organized; it is quite nice to look at. But if you leave the garden unattended for a while, the plants will start competing with each other as they try to take over the soil and after a while you end up with a mess. In this case the organizing force is the gardener who regularly maintains the garden to keep it nice and organized. Without this organizing force the orderly garden will gradually decline into disorder.

Example 2. Thinking in similar lines about a car, which is also made of many parts, each having a specific function. Over time, each part wears out and their interactions with other parts change. Gradually the nice shiny new car deteriorates into an old car that does not drive or perform as well as it used to. However, as we all know, we can slow down and even prevent this gradual deterioration through regular servicing and maintenance. Without this intervention the orderly system will gradually decline to disorder.

Example 3. Thinking back to the flock of birds we discussed in the previous chapter, you may ask: if systems deteriorate to disorder over time, what keeps these birds dancing in the sky and creating these unique patterns? What keeps these birds together is herd or swarm behaviour (i.e. the behaviour of individuals or animals in a group acting collectively without centralized direction) which comes about because of very simple rules that are genetically coded into these birds. This rule may be something as simple as I will not be more than six inches apart from the other birds in my flock. In this case, unlike the previous two cases where there was an external intervention or control, there is an intrinsic system in place that keeps the flock together. In fact, it is this intrinsic control system that defines self-organizing systems we discussed earlier in this chapter.

Example 4. Let us think about a crossroads junction with no external controls such as a roundabout, traffic lights, or even a policeman directing the traffic the old-fashioned way. In this case only one rule applies and that is self-preservation; in other words do not hit anyone as it may injure me and/or it will cost money. Watching such a crossroads from a distance is quite unsettling; often it looks like chaos, but accidents are rare, and life goes on. If you wish to see one of these traffic junctions just search for crazy traffic junctions in YouTube and look for the ones with no traffic.

So far through these examples we have established that systems without an intrinsic or extrinsic control system tend to move from order to disorder. In systems thinking, this phenomenon is known as entropy. The term entropy is borrowed from physics where it is used to describe how organized heat energy is lost into the random background energy of the surrounding environment (the Second Law of Thermodynamics). As such, in systems thinking, entropy can be used as a metaphor for ageing, skill fade, obsolescence, or similar.

Entropy equally applies to organizations. In fact, from a systems thinking perspective, a key purpose of management is to prevent entropy. This is typically achieved through continuous improvement change and renewal.

Homeostasis is another term used within systems thinking to describe a system that maintains its ‘steady state’ or a system that is in a ‘dynamic equilibrium’. Examples include the human body’s ability to remain at a steady temperature and an organization’s ability to maintain its performance within its market. In the above examples, intrinsic and extrinsic controls or interventions prevent entropy and enable systems to achieve homeostasis.

In the next chapter we will explore common models and frameworks that underpin systems thinking, providing further insights into these intrinsic and extrinsic controls.

CASE STUDY

A systems thinking case study

As we have seen in this section, there are different ways to conceptualize systems. When it comes to thinking about organizations, one of the first things that comes to most people’s minds is the organization structure. Although this is a valid way of conceptualizing an organization in systems, there are other ways that are often more useful.

This case study is based around a whisky manufacturing company which distils whisky in its distilling plants and then stores the beverage for anything from 5 to 25 years. The older the whisky the more valuable it is. The company bottles the whisky in its bottling plants before the product is distributed to various customers and retailers around the world.

On the surface all the products appear the same; whisky is an output from a distillation process. It stays in stock for some time (5 to 25 years) and then in the

When we analyse how these products behave in the market and the customer buying behaviour, we see two distinctive groups emerging. One group of products, lower-value products, behave as commodities. They are sold throughout the year in reasonably high volumes. The customers are more interested in what is inside the bottle, i.e. the whisky, and they are less interested in the bottle or the packaging. The other group is high-value products that behave like fashion items or pieces of art. They are significantly more expensive, and are bought and used like trophies. Often they are given as presents or corporate gifts. They have a brand value and a story behind the product. They are sold in much lower quantities and the sales volumes are much more seasonal and variable. The customers are more interested in the brand value with time. In some cases the bottles become collectors’ items and increase in value with time. Some people treat these as investments; thus they do not open the bottle and drink its contents. The customers pay more attention to packaging, which must look elegant and be representative of high price.

If we examine how these products are made, the low-value products are made in fully automated high-speed production lines in high volumes. The quality and cost of the packaging (i.e. the bottle, label, cap/cork, etc.) are comparatively low. In contrast, the high-value product is made in smaller quantities in semi-automated and sometimes manual production lines using higher-quality/value components such as coated bottles, embossed labels, weighted bottle caps, metal cases, etc. The speed of production is much lower, and special care is taken to ensure that the bottles and labels are not scratched or damaged by rubbing against each other. The look and feel of the bottle must be right. In terms of finances, the margins from high-value products are significantly higher than those of the low-value products even though they cost more to produce.

When we analyse how the products behave from these perspectives we see two systems within this company: one system for high-value products and another one for low-value products, with each being produced differently and serving different markets. Although both of these systems may have the same purpose, i.e. to make money for the company, their functions are clearly different. The two systems would also have different performance measures, one emphasizing speed and efficiency, the other quality and image. If we were to improve the performance of this company we would need to understand how each of these systems behaves before designing an intervention. The consequence of looking at just one of these systems and designing a performance improvement intervention without differentiating. 

Reflective questions

In the above case study we have already identified how the two systems may differ from each other in terms of purpose, function and performance measures. Can you think about what may be the external forces that could influence the behaviour of each system? Think about how these forces may influence the behaviour of these systems.

Think about the organizational structure of this company. How would the structure we identified above impact the design of the organizational structure? Would you have two different management teams, one for each system, or would you have just one management team managing both systems?

3.10 Summary

We started this chapter with the aim of building upon the key systems thinking concepts and definitions introduced in the previous chapter by giving you, the reader, a deeper and more complete understanding of systems, their structure and properties. Based on the discussion so far, we can summarize that:

A system has a purpose (usually) and a function.

A variety of things (materials, people, information, knowledge, energy and so on) flow through the system as inputs and outputs.

Stocks and flows are useful ways of thinking about a system, and they are not limited to tangible things like materials, people and energy, but they can also be intangible things such as knowledge, experience and goodwill.

These stocks and flows create the forces that shape the emergent behaviour of systems.

The performance of a system may be conceptualized as effectiveness and efficiency of a system.

A system exists within a wider operating environment and is connected to other systems within this environment. Indeed, that wider system itself may be part of an even wider and bigger system.

.A system has a boundary, but these boundaries can be defined by the nature of the enquiry and/or the worldview of the manager/analyst studying the system.

.A system has a structure that comprises parts that are interconnected with one another within the system. These parts may also be connected to other systems in the wider environment through the system’s boundary. Indeed, these parts may be subsystems in their own right with their own internal structures.

.A system interacts with other systems in its environment. This interaction results in the system being influenced/changed by its environment as well as the system influencing/changing its environment.

.Connections between parts of the system are critical in helping us to understand the causalities between these parts and how these interact in concert to define the behaviour of the system.

Understanding the structure, purpose, function and performance measures of the systems within organizations from different perspectives is critical first step for understanding why organizations behave the way they do. When we are managing an organization, making decisions without this understanding is likely to produce unexpected and unintended outcomes.

REFLECTIVE EXERCISE

At this early stage, as a new manager or a student of systems thinking, it may be quite challenging to conceptualize your own organization (or part of it) as a system. It usually helps to start thinking about a system in simple terms and then move on to building a more complex picture. Thus, in order to help you ground your learning, we would encourage you to start thinking about your own organization or part of the organization by conceptualizing it as an image, illustrated in Figure 3.6. Here we have used just some of the systems concepts and properties we have introduced thus far. In your own exercise you do not have to use all of these or be limited to these, but it is a start.

You may start by drawing the boundary and then start putting in parts of the system that you think may belong to this system. If it is helpful, you may also put parts of other systems outside the boundary. Then start thinking about what flows between these parts but also about what happens during these flows and how they may influence each other. For example, sales forecasts may flow between two parts. You can also think about what happens when these forecasts are wrong or intentionally under-reported. Try to capture the feelings and perspectives of people or teams in the relevant parts of the system and try to explain the resultant behaviour.

In order to explain the picture, you will find it helpful to write a narrative of what is happening in the picture. Keep this short and to the point, but make an effort to capture the causalities behind various behaviours.

TEAM EXERCISE

As a facilitator, try to identify a system that is common to your group of managers or students. This may be their workplace or equally somewhere they all identify with, something like a local well-known fast-food restaurant, sports club, or even their place or programme of study. Ask each of them to do the above exercise individually. Then ask each member of the group to talk through their model. Try to draw attention to the similarities and differences between the models. Ask them to discuss why the differences may be there.