GBN in Conversation
presents
Paul de Ruijter on Systems Thinking, 7/97
Systems thinking is not new, but today I'd like to describe one technique we've used. We have used systems thinking on a project in which we didn't do scenarios; we went straight from systems thinking to options generation without scenarios in-between.
The project I'm going to describe is one we did for the Economy, Ecology and Technology Steering Group; a Steering Group consisting of people of the Dutch ministry of Education and Science, the ministry of Economic Affairs and the ministry of the Environment.
The principle of scenario thinking is thinking about structure. An example is: if you were from Mars, and you saw a bird's-eye view of three men in a boat and wanted to turn the boat left, what would you do? You might apply pressure from the front right, from the rear left, etc. But if you knew the structure of the boat, and knew it had a rudder, you'd know where the leverage points were and you'd know how to move it most effectively.
Senge's book contains a chapter called If I had a lever long enough, I could move the world. That's systems thinking.
People often concentrate on the outside forces, but it's the internal structure that makes things move. Alain Wouters demonstrates it this way: he grasps a slinky from the top and puts his other hand underneath. Then takes his hand out from underneath and the slinky oscillates up and down. He asks a group why that happened. He gets answers like "because you took your other hand away" or "gravity". Then he holds a pen the same way and lets go from underneath; his other hand is still holding it and it moves. It's not the external forces that make the slinky behaves like it does, it's the internal make-up of the slinky.
It's like people saying the profitability of their chemical plant is going up and down because of the economy. But it's not the economy, it's because of the structural inadequacy of the company to adapt to small changes in demand.
So systems thinking focuses on the structural linkages between internal factors and factors in the environment.
There are two important concepts in systems thinking: reinforcing loops and balancing loops (and a combination of the two).
Reinforcing loops are a cycle, either virtuous or vicious. For example, if your expenses go up, your debt goes up, so your interest payments go up, so your expenses go up…and it grows exponentially.
A balancing loop would be this: if your demand for sales goes up, your contact hours with clients goes up, so your time for acquisition goes down, so your demand for sales goes down…it balances out.
That's it for systems thinking. Look at a system, identify the variables, think about the relationship of the variables and what the relations between them are, then you can see how the system is behaving.
Q: Based on the slinky and boat examples, systems thinking takes care of the internal forces only; Scenario Planning works in external factors; how do they feed into each other?
A: You could look at the external as one big system, and identify the whole system of which the company is a part. Do one or a couple broad systems diagram(s), to find your leverages.
Q: Could you have external forces that are one-way linked? They affect you, you don't affect them.
A: Yes, you draw those as a cloud with arrow pointing into your system.
We were asked by the ministries of Economic Affairs and Education and Science, both in charge of technology development in Holland, to work with their steering group to think of new themes/target areas for them to put money into.
Goal of the program was not to come up with a strategy for survival for the ministries; rather, it was to change the world.
When you're working for the public sector, everyone reads everything you write. Therefor we started with a literature review to see what everyone else had said. Our project was to look at the links between the economy, ecology, and technology, so we took that apart and put them back together in pairs.
First we looked at the economy and ecologyóthe whole government is steered by GNP, but that doesn't measure what you remove from nature and what you dump back into it. The incentive structure of the government is wrong; what is valued is wrong. We put a question to the two ministries: what do you want, to increase the GNP or increase the real value people are perceived to have?
Then we tried the relation between the economy and technology; 50% of the growth of the economy comes from technology. And because the GNP doesn't measure quality, just quantity, it might just even be higher. Do we need technology to improve things for quality or for quantity?
The third couple we tried was technology and ecology…lots has been done there, lots of positive things. Companies produce a lot cleaner than, say, a year ago. But there still remain many lock-ins like in the car- and gas-industry were there are very dominant players with dominant technologies; the problem is breaking the lock-ins.
Then we went to the interviews. The interviewees scared us a bit because there's lots of anger towards the government. Some blame the government, because they tax people and not energy or material which is spoiling the environment. If you calculate the burden on human capital, even with a high emissions tax, you're not balancing it. This results in a replacement of people for machines, which is bad for the environment.
Also, regulations. If you want to change a car, you have to change the regulations around it, and that takes a long time, so the government is actually part of the lock-in. The government wanted to change industry to become more environmentally friendly, but the government itself is a big part of the problem.
The interviews showed us how people saw the glass half full or half empty things differently: industry said we've come a long way and environmental groups said we're not even close to our goal.
We took 25-30 people from industry, environmental groups, universities and the government. It was hard to get these people together; the subject is quite broad so people thought they could stay out; that's where Jaap stepped in and pulled some strings.
We did a one-day workshop. We started the morning with a plenary session and asked people what the prime blockages were between the environment and the economy and clustered the hexagons. We came up with:
1. Decoupling of culture and nature
2. Separate decisionmaking process (economy v. ecology)
3. Suboptimal knowledge management
4. Deteriorating carrying capacity of nature
Then we split up into four groups of about seven people each; each group took one blockage and were given the assignment of answering the five whys on each blockage; go really deep.
Facilitators were in the room, so whenever they heard an important variable, they put it on a hexagon and placed it on the board. Whenever there was a relation between two variables, an arrow was drawn, and slowly we came up with systems diagrams without asking for it.
We came up with a classic one which is a bit similar to the Limits of Growth model (s=same, o=opposite):
Then we came up with another, that showed that the EET program is part of the problem, because it works on ecological fears; the EET program tried to "disconnect" people from nature by stimulating new "stand-alone" technology.
These models could also have been used to generate several scenarios about nature being integrated in the economy via higer prices, a high tech world were we all live in "artificial biospheres" or a "back to nature" kind of scenario.
So, we worked on blockages, identified structures, then built the system. The last step in the workshop was: if you know the system, what would you do?
EET is now in the upper right "artificial biosphere"-loop; and it should redefine itself to work to move to the left "back to nature"-loop. New technology programs which are put forward for EET-funds should work together with nature, instead of against it. EET should avoid "context-loose" solutions. If you design a factory so you could put it anywhere, you are not in sync with ecological thinking. EET should take Industrial Ecology as an example, to integrate your environment in the design of production systems. Secondly, the value of projects which are entered for EET-funds should be expressed in new indicators, integrating economy and ecology from the start.
We did the analysis and option development in one day. We spent approximately one and a half hours for blockages, 3 hours for systems, 1 hour for option development.
Everyone in the workshop was happy, but we were working with the civil servants and not the politicians. So we are not sure yet if politics will choose to work with the new insights for the EET-program.
Q: How did you use your pre-analysis in designing the workshop?
A; It was very important to know where the tensions were in the group and what the issues were, before designing the workshop. And the literature study was done to satisfy the need for a "substantiated" report. Since everything you do for the public sector is public, the requirements for everything that you write are higher.
Q: Isn't this more of a strategic conversation than a scenario workshop; it seems high-level, not a detailed, quantifiable resource.
A: Yes, it wasn't a scenario workshop, and it was a strategic conversation. The result is "actionable" for the steering group, they could choose to implement the suggested themes and criteria in the EET-program.
Q: How you integrate this into scenario projects?
A: Shell's 1992 scenarios were called Barricades and New Frontiers; they were actually narratives written in support of the systems diagrams:
Q: Do you typically write a systems diagrams in every scenario project?
A: No, depends on the project.
Q: How do you decide which one you'll do?
A: It emerges, and if you see an interesting structure emerging, you say "wow, I can explain this better with system diagrams". In the EET project, we wanted to change something in the outside world, so we didn't want to do scenarios which separate the internal structure from the outside world.
Q: There's something about this that's making me uncomfortable. You draw this diagram, it's looks very rational, it's one view of the system. if you're not careful people could think that the model is the truth. The most important discovery in some of our projects are things that aren't on their radar screen.
A: Well, the system diagrams emerged in the conversation...they weren't thinking about all the surfaced variables at the same time, they didn't know how they were connected. Those new insights were synthesized during the collective conversation.
I think it's important to know the group you're working with; some like narratives, some like systems diagrams.
Q: How do you make sure you've included all of the important variables?
A: You never know.
I've simplified the diagrams for this purpose; in the actual workshop, the diagrams are much more complex. After the initial conversation, you can boil it down.
Q: This is different from Kees' business idea; Kees does the organization only. He uses systems to tie the organization to its marketplace. I don't think we have an easy explanation of how to use this in GBN context.
A: There are inductive and deductive ways to do scenarios, but you could also try to directly model the world you are working in and let the narrative come afterwards (if needed). Using modelling makes your thinking internally consistent; and it gives insight about how things could happen over time; that's how it was done in Shell in 1992. Narratives came after systems thinking.
Q: There may be some confusion in GBN between systems and logics diagrams.
A: Scenario Logic Diagrams tend to sketch a linear time flow, from left to right. In here I've used a circulair time dimension. Some people put numbers in these system diagrams, but that's tricky. Quantification and simulation using these diagrams work only in certain cases, and have only a limmited value. You need to know what you are doing, otherwise you end up making an expert-model which is a black box for the users, reducing the learning value. But we in GBN tend to shy away from everything that is quantitative, but we shouldn't overshoot that way either!
Q: How does this relate to that other way of systems thinking: Checklands Soft Systems Thinking?
A: To model human activity systems you could use VOCATE. There's more than one system, but given only one Viewpoint one can identify the Owner of the system, and the Client of the system, and the Actors in the system, for Transformations in the system, within the Environment of the system.
At Shell we did this to define a learning organization. A real learning organization must be able to change its owner, able and willing to change its client, actors, transformation processes, and should be willing and able to change its environment if needed. An organization defined this way defines it self; the organization is what the organization thinks it is. So, if you're changing the mental map of an organization, you're actually changing the organization. But that would bring me into the theory of self-referential systems....
