2.2. A Start on Causal Loop Diagrams

A causal loop diagram is a visual tool for the feedback systems thinker. As in the transportation example, such diagrams show cause and effect relationships and feedback processes. All causal loop diagrams are constructed from the same basic elements: words, phrases, links and loops – with special conventions for naming variables and for depicting the polarity of links and loops. Figure 2.6 is a very simple causal loop diagram, just a single loop, connecting hunger and amount eaten in a tiny model of appetite. Deliberately there is very little detail. Imagine the situation for yourself. You are hungry, so you eat. How would you describe the process that regulates food intake? Common sense and experience says there is a relationship between hunger and amount eaten and this is shown by two causal links. In the top link hunger influences amount eaten, while in the bottom link amount eaten has a reverse influence on hunger. Each link is assigned a polarity, either positive or negative. A positive '+' link means that if the cause increases then the effect increases too. So an increase in hunger causes an increase in the amount eaten. A negative '−' link means that if the cause increases then the effect decreases. So an increase in the amount eaten causes a decrease in hunger. In fact, the assignment of link polarity is just a bit more sophisticated. In general it is better to imagine the effect (whether an increase or decrease) relative to what it would otherwise have been, in the absence of an increase in the cause. This turns out to be a more robust test.^[] In any case, the two concepts, hunger and amount eaten, are mutually dependent, and this two-way dependence is shown as a closed feedback loop. The feedback loop represents, in outline, the control of food intake.

^[] This more sophisticated assignment of link polarity works equally well for normal straightforward causal links and for links that correspond to stock accumulation processes. The distinction will become clear in Chapter 3 where stock accumulation is introduced as a vital concept for modelling and simulating dynamical systems.

Figure 2.6. Simple causal loop diagram of food intake

There are a few more details to explain in the diagram. The bottom link contains a box labelled 'DELAY'. This symbol shows a time delay in a causal link where a given cause leads to an effect, but not immediately. There is a lag. So here the more you eat the less hungry you feel, but it takes a while for hunger pangs to diminish. Such time delays add dynamic complexity because cause and effect is less obvious. Where eating is concerned, a time delay of 20 minutes or so can make it much more difficult to regulate food intake. Overeating is a common result. In the centre of the diagram there is another special symbol, a 'B' inside a small curved arrow, a loop identifier to indicate a balancing feedback loop. Generally speaking a feedback loop can be either balancing or reinforcing. The names give a clue about the way the feedback process operates. In a balancing loop a change in the condition of a given variable leads to a counteracting or balancing change when the effects are traced around the loop. A simple thought experiment illustrates the idea. Imagine you take a long walk and return home feeling hungry. Hunger rises and the feedback loop swings into action. Amount eaten rises and eventually hunger declines. The feedback effect of the loop is to counteract the original rise in hunger, which is a balancing process. By comparison a reinforcing loop amplifies or reinforces change. In a realistic multi-loop system, such as the transport example mentioned earlier, behaviour through time arises from the interplay of balancing and reinforcing loops. It is useful, when interpreting a web of causal connections, to identify the main loops as a way of telling a story of what might unfold. At the same time, it is a good discipline to name each loop with a mnemonic for the underlying feedback process. Hence, in Figure 2.6, the balancing loop is called 'control of food intake'. Similarly, in Figure 2.3, a feedback view of road congestion is depicted vividly as the interplay of balancing loops for 'capacity expansion', 'discretionary trips', 'extra miles' and 'take the bus'.