Over the last two years, I’ve repeated run into the concept of “systems thinking.” From context clues, I was able to cobble together a rudimentary understanding, but I never had an formal grounding it in. A month or so ago, I heard several people on Twitter mention Thinking in Systems: A Primer, so I downloaded the Kindle sample and took a look. That was quickly followed by buying the book and devouring it.
From a structural perspective, the critical features of systems are stocks (the amounts of something you have, like temperature or population or electrical charge), flows (the ways that stocks change), and feedback loops (which change flows based on the quantity of a stock). Think about the temperature system in your house. Heat flows in from the furnace, which raises the temperature of the room. Heat flows out of the room to the outside, which lowers the temperature. So you’ve got a single stock with one inflow and one outflow. The rate at which heat comes in to the room is controlled by your thermostat, which compares the current temperature to the desired temperature. The rate at which it flows out is control by the laws of thermodynamics, which tries to equalize the inside and outside temperatures. So in this system you have two competing feedback loops, and which one will dominate depends on initial conditions and the specific values of certain variables.
There are a couple of nuances here, of course. Feedback loops come in two varieties: balancing and reinforcing. Balancing feedback loops try to drive a stocks to a specific value (like the two loops in the temperature example), while reinforcing loops cause a stock to increase exponentially over time (like compound interest). But as was reading the first couple of chapters, I thought, “These concepts are pretty intuitive.”
Then I hit the car dealership inventory example, where delays in the balancing feedback loop cause oscillations in the dealer’s inventory. That piqued my curiosity. On the next page I saw how reducing some of these delays — which seems like it should make the problem better — actually increases the amplitude of the oscillations. And that’s when I knew had to keep reading.
My initial assessment was correct in this respect: Most of these concepts are easy to understand, at least in static analysis. The two-chapter overview of the components of systems and some well-chosen examples from the “systems zoo” gave me a reasonable grasp on the sorts of things I might encounter. Systems, of course, are dynamic things. One of my two favorite sections of the book is on “sources of system surprises,” an explanation of why we so often fail to apprehend the dynamic behavior of systems. These include the multitude of non-linear relationships in systems, the problem of bounded rationality, and that the limits on a system exist in layers around it. Systems are less simple than they look, and that’s were we get in trouble.
Trouble is the underlying theme of my other favorite part of the book, on “systems traps and opportunities.” This explains problems like arms races, the tragedy of the commons, even drug addiction, from a systems perspective and offers ways out. This was the most resonant section for me, because it caused me to start seeing systems everywhere. As I listened to an NPR story about the economic impacts of Alabama’s new immigration enforcement law, I started drawing stocks, flows, and feedback loops in my head. When I read about solutions to the “success to the successful” trap, I thought, “Oh, this is how Power Grid prevents the runaway leader problem by giving the last place player the most advantageous position.”
This is all useful information, but the key is that it is presented in one of the most human voices I’ve read. The book was published from a manuscript Donella Meadows had been working on when she passed away unexpectedly. Her decades of experience and understanding come through in way that is remarkably humble. As such, while the book is technical, it’s surprisingly applicable and astonishingly approachable. Which makes it the kind of book I would like to write someday.