The Field Guide › Paper
Donella Meadows's short, much-loved essay distilling decades of systems-dynamics work into a ranked list of the places where intervention in a complex system pays off, and the places where it does not. A leverage point is somewhere in a system (an organisation, an economy, a body, an ecosystem) where a small shift produces large changes throughout. Meadows lists twelve, ordered from least to most powerful: constants and parameters (subsidies, taxes, standards); the sizes of buffers; the structure of material stocks and flows; the lengths of delays; the strength of balancing (negative) feedback loops; the gain of reinforcing (positive) feedback loops; the structure of information flows (who has access to what); the rules of the system (incentives, punishments, constraints); the power to add, change, or self-organise structure; the goals of the system; the mindset or paradigm out of which the system arises; and the power to transcend paradigms. The essay's central, counterintuitive claim is that almost everyone pushes on the wrong points: as she puts it, about 99% of attention goes to parameters, where there is very little leverage, while the high-leverage points (information flows, rules, goals, and above all the governing paradigm) are neglected, and are frequently pushed in the wrong direction. The relevance to psychological safety is a diagnosis of why so many attempts to build it fail. Most organisational effort goes to the low-leverage parameters: the annual survey, the one-off training, the policy statement, the numeric target for 'engagement'. Meadows's hierarchy points instead to the high-leverage interventions: who is allowed to know and say what (information flows), the real incentives and sanctions around speaking up (the rules), what the group is actually optimising for (the goals), and, highest of all, the paradigm, whether psychological safety is conceived as a commodity to be installed and measured or, as this corpus argues, a commons to be cultivated, and whether the operating belief is blame or learning. It anchors the complexity strand of the corpus with a systems-thinking classic distinct from the safety-science framing, and gives precise language to the recurring critique of shallow, parameter-level intervention.