The manuscript of ‘Thinking in systems‘ has been around in draft since the early ’90s, but never completed. Now her colleague Diana Wright has edited it for publication. In the circumstances, it ought to be something of a publishing event, even if a niche one. It is, I’d say, the best single introduction to systems work that is available, especially for non-specialists. But the book seems to have surfaced with little fanfare, and barely a review.

Thinking in Systems effectively unfolds in three parts. A first section runs through the main systems concepts (stocks and flows, time graphs, and reinforcing and balancing loops). A second section looks at how systems work, but told through examples and stories. And the third section explores what we should do to use some systems knowledge to improve policy and management decisions.

Why use systems at all? In a good passage, which I’ve shortened here, she writes about the way we understand the world:

1. Everything we think we know about the world is a model. Every world and every language is a model. … So are the ways I picture the world in my head – my mental models. None of these is or ever will be the real world.
2. Our models usually have a strong congruence with the world. That is why we are such a successful species …
3. [But] our models fall far short of representing the world fully. That is why we make mistakes … We often draw illogical conclusions from accurate assumptions, or logical conclusions from inaccurate assumptions. …

And to give two examples of the last type, we aren’t very good, for example, at thinking about non-linear systems – ‘trends that bend’ – or at understanding exponential growth. We’re also poor at estimating the effects of delays within a system, or how the relationship between stocks and flows affects overall outcomes.

Systems, then, are a way to create different – one hopes, better – models to inform the way we perceive the world. And successful systems have a number of core characteristics: they are resilient (they can absorb shocks without collapsing); they are capable of self-organisation (if you leave them to their own devices they can can structure and re-structure themselves, creating diversity as they go), and they are also hierarchical – which creates stability. enables resilience, and reduces the amount of information that the system needs to track. But hierarchies serve the bottom, not the top; they evolve from the lowest level upwards.

Although it’s implicit in the text, it is evident that Meadows regards systems thinking as a having a radical purpose. This becomes clearer as she moves towards the policy and public implications. Indeed, our public policy would probably be a lot better if every civil servant had to read chapters 5 and 6, on systems traps and leverage points respectively, before they were let loose with public resources.

To take one ‘systems trap’ by way of an example, ‘policy resistance’ describes systems where each actor has a competing interest, but the overall effect is to maintain the system just as it is. In the drugs market, for example, if the police are effective at reducing the drugs supply. the effect is to drive up price and encourage suppliers to find new sources. In the meantime, robbery levels are likely to rise as drugs users have to find more money in the short-term to finance their habit (“goals of subsystems are different from, and inconsistent with, each other”.) The policy solution is either to ‘let go’, since the system will re-balance itself, but with less public effort being expended, or to redefine the policy goal so it is richer. (Instead of Romania’s policy, which failed, of simply increasing population – regardless of the financial ability of families to raise the extra children – Sweden had a policy that every child should have excellent care, which was much more successful at increasing birth rates and population levels).

Chapter 6, about leverage points within a system, is a version of a paper that has been available online in earlier drafts for some time.It’s stronger for being in this context. Since there are 12 leverage points, I’m not going to list them here. But it is striking that those which are more effective are more about how systems monitor and manage themselves, rather than about their physical structure. Numbers five and six are about adjusting rules and improving information flows, number three is about the goals of the system, while numbers one and two are both about how we frame the way we understand the purpose of the system; the underlying paradigm. Since there’s never a correct system boundary (it is shaped by the question you ask) systems thinking can be rich in challenging these. It is still depressing that last week’s G20 meeting was mostly messing about at the bottom end of the range of leverage points, fiddling with the numbers (or “diddling with the details”, as Meadows puts it), trying to manage stocks and buffers, and improving the feedback loops.

Diana Wright and the publishers (Chelsea Green) have done a good job in keeping things straightforward and comprehensible, with clear diagrams (too clear in one respect, since my systems-colleagues would want to see labels on the causal links within diagrams), ‘interludes’ for interesting examples, and text boxes summarising core ideas and main points. There’s also a surprising amount of humour.

For me, one of the most basic ideas remains the most resonant: that you don’t judge a system by its rhetoric, by what it says it does – you have to judge it by what it produces. It’s a valuable sanity check when thinking about quite a lot of organisations, from hedge funds to the Home Office.