In one line: A system’s behavior comes from its structure (stocks, flows and feedback loops), so to change the behavior, understand and change the structure rather than reacting to events.
Core idea
Donella Meadows’s Thinking in Systems [1] is a primer on how systems work and why they surprise us. A system is a set of elements, interconnected in a way that achieves some purpose. The elements are usually the most visible part and the least important. The interconnections and the purpose drive behavior. Change the players on a sports team and it is still the same team; change the rules and it is a different game.
How it works
- Stocks are accumulations you can measure at a moment: water in a bathtub, inventory, cash, trust, knowledge, carbon in the atmosphere.
- Flows change stocks over time: inflows and outflows such as production, sales, hiring, learning, emissions. A stock changes only through its flows, which is why stocks act as buffers and create delays.
- Feedback loops connect a stock back to its own flows:
- Balancing loops seek a goal and resist change, like a thermostat or the way inventory triggers reordering.
- Reinforcing loops amplify change, like compound interest, a viral product, or a skills gap that keeps widening.
- Delays in feedback loops cause oscillation and overshoot: think of a shower that’s too cold, then too hot.
- Bounded rationality: people act sensibly on the information they have locally, yet together produce outcomes nobody wants.
- Resilience, self-organization and hierarchy are what let systems survive, adapt and grow.
The carbon “bathtub” is the classic illustration. Atmospheric CO₂ is a stock; emissions are the inflow and natural uptake the outflow. As long as inflow exceeds outflow, the stock keeps rising, even if emissions fall.
In practice
Systems thinking changes the questions engineers ask:
- not who caused this failure? but what structure made it likely?
- not how do we fix this event? but what pattern does it belong to?
The book’s second half turns from how systems work to how they go wrong (see System Traps) and where to intervene (see Leverage Points). The practical tool for drawing a system is the causal loop diagram.
Questions to ask yourself
- What are the key stocks in the system I’m trying to change, and what flows feed them?
- Which feedback loop is currently dominating its behavior?
- Where does a delay in feedback cause us to overreact or underreact?
Connections
- Applications: System Traps, Leverage Points, Causal Loop Diagrams, Iceberg Model.
- Related: Complexity, and Deming’s “appreciation for a system” (see Deming’s Quality Philosophy).
- Capabilities exercised: Inquiry, Vision, Connecting Across Disciplines.