Causal Loops

A causal map is a specific type of model focusing on causal factors. It is an abstract model that uses cause and effect logic to describe the behavior of a system. Causal Models have found empirical realizations in many different fields and have been integral to theories of cognitive balance proposed by Heider in 1946 and Kelly in 1955, and since then have been applied in psychology, international relations, management sciences, game theory, group decision, and electric circuits’ analysis. They have a natural application to integrated management communications.

To construct a causal map, start by naming the variables of the relationship. Next, if one variable affects another draw a line from the “cause” to the “effect.” Place a plus or minus sign next to each line drawn according to the following: + means the two connected events move together, when a cause increases so does effect; - means the two connected events move in opposite directions, when a cause increases the effect decreases.

Consider Maruyama’s model of urban growth (see Weick for a fuller explanation of this model.) Maruyama defined the variables as:

• Number of people in city
• Modernization
• Migration into a city
• Sanitation facilities
• Number of diseases
• Bacteria per area
• Amount of garbage per area

His analysis resulted in the following relationships:

Any variable in the map that only has lines coming into it is a dependent variable. Any variable in the map that only has lines going away from it is an independent variable. A variable that has both lines in and lines out is an interdependent variable. In Maruyama’s solution above, all variables are interdependent.

To analyze with a causal map, find causal loops. To do this, start with an interdependent variable. See if there is a pathway leading out on the from-lines eventually returning on the to-lines, i.e. start with interdependent variable and see if it has paths back to itself. Here are the causal loops in the Maruyama map:

If there is an even number of minus signs (0 is even) in a loop it is a deviation amplifying loop, aka vicious circle or virtuous circle depending on your viewpoint. Trace what happens when a variable in one of the loops increases. Take loop P-M-C-P, image (1) above. An action to increase P (number of people in a city), goes through the loop, and at the end, yet a further increase in P is produced.

If there is an even number of minus signs (0 is even) in a loop it is a deviation amplifying loop. This means there is NO REGULATION OR CONTROL. Once a variable starts in a direction, it will continue that way until system is destroyed or a dramatic change occurs.

If there are an odd number of minus signs in a loop it is a deviation counteracting loop. These impose stability on the system, self regulation. Once again, trace what happens when a variable in one of the loops in our example increases. Take loop P-G-B-D-P, image (4) above. An action to increase P (number of people in a city), goes through the loop, and at the end, a decrease in P is produced. This effect, opposite from the cause results in a stability for the loop. A change eventually dampens into an equilibrium.

When we examine relationships with complex interactions, such as integrated marketing communications, we can use a causal map. After defining the varibales and their realtionships, we would

• Look for interdependent variables
• Causal loops
• Presence or absence of control
• Deviation counteracting loops mean system is basically stable

Is stable always good? Change may be an objective, in which case working a deviation amplifying loop would be advantageous.


References

Weick, Karl (1979). The Social Psychology of Organizing, 2nd Edition. McGraw-Hill.