See Leverage Points Process. You will already have a CLD so start from step 3.
In a very simple model a quick method is as follows: Identify which variables you can influence. Circle them.
In the CLD below which is about my exploration of the factors affecting my horse's headshaking, I have used the color green to highlight the variables I think I can influence. I used yellow for the one I can partially influence.
My horse's headshaking and variables I can influence
We then ask: What will be the system impact of intervening there?
We create driver trees of the consequences using Vensim or by hand.
How do you feel?
Where are the root causes?
If we create a full depth 999 Driver Tree in Vensim. The root variables are on the far left (the very tips of the roots.)
999 Driver Tree for the Headshaking CLD
Some of the root variables we can influence and others we can't. In the case of this CLD we can influence: Interesting work Fit of Tack Breeding Quality of Conformation Rider's Weight Rider's body awareness Rider anxiety Rider's hands Spookiness of environment Riding in fly areas Pollen Use of nets Wind
Example: Driver tree of "Delight" - Unleashing Genius
There is a difference between a root cause fix and symptom masking and the consequences. Sometimes we need both. A good example would be taking pain killers to mask the symptom of pain while we change our diet (the longer term fix.) Remember that some quick fixes can be addictive! and they could have other consequences.Be conscious of your choices and how you will experiment with them.
Think again about whether we have defined the boundary correctly. See Three KEY Questions
We can also create a matrix with the variables vertically listed and the stakeholders horizontally listed. Tick which stakeholders influence which variables. These groups need to be aware that there may be missing conversations.
I created a CLD like this for the Grenfell Tower fire and used it to identify leverage points.
There are two other structured approaches to identifying leverage points:
1. See Leverage Points Process.
2. We can explore the consequences of our proposed interventions by converting our CLDs to Stock Flow Models (see System Dynamics Structure.) We can then calibrate the relationships (sounds simple but it is not!) and create microworlds to simulate the system.
See Leverage Points