Other Mental Models · OM-18
Not every place you could intervene in a system produces equal effect — a small, well-chosen change at the right point (a system's rules, its goals, its underlying paradigm) can produce far more total change than a much larger effort spent adjusting a parameter that isn't actually where the system's behavior originates.
A framework identifying places within a complex system where a relatively small intervention can produce a disproportionately large shift in the system's overall behavior, and ranking different types of intervention points by how much systemic leverage they typically provide — from low-leverage numeric parameters up to high-leverage shifts in a system's underlying goals, structure, and paradigm.
Developed and most fully articulated by systems scientist Donella Meadows in her widely cited 1999 essay 'Leverage Points: Places to Intervene in a System,' building on decades of prior systems-dynamics research including her work on the Limits to Growth studies.
The Mechanism
Where you intervene matters at least as much as how hard you push
Meadows explicitly ranked numeric parameters (tax rates, subsidy levels, specific standards) as among the lowest-leverage intervention points in a system — despite being the type of intervention policymakers and managers most frequently reach for, because they're the easiest to adjust — while identifying a system's underlying goals and paradigm as far higher-leverage points that are simultaneously much harder to identify and to change.
01 · MOST REAL-WORLD INTERVENTIONS DEFAULT TO THE LOWEST-LEVERAGE POINTS BECAUSE THEY'RE THE EASIEST TO ADJUST
There's a systematic mismatch between where leverage actually is and where intervention effort typically goes
Numeric parameters are the easiest type of intervention to identify, measure, and adjust — a tax rate, a subsidy amount, a specific numeric standard — which is precisely why they receive a disproportionate share of real-world intervention effort, even though Meadows's framework ranks them among the lowest-leverage points available, producing a systematic pattern where a great deal of effort goes toward comparatively low-impact interventions.
02 · HIGHER-LEVERAGE POINTS ARE HARDER TO IDENTIFY AND OFTEN MEET MORE RESISTANCE
Changing a system's underlying goals or paradigm threatens existing power structures more directly
Because a system's goals, rules, and underlying paradigm are frequently tied to the interests and worldview of whoever currently benefits from the system's existing behavior, high-leverage interventions targeting those elements tend to meet substantially more resistance than low-leverage numeric adjustments, which explains part of why organizations and policymakers gravitate toward lower-leverage interventions even when they're aware higher-leverage options exist.
03 · MEADOWS HERSELF CAUTIONED THAT THE RANKING IS APPROXIMATE, NOT A PRECISE FORMULA
The framework is a way of thinking about leverage, not a rigid, universally applicable ordering
Meadows explicitly noted in her original essay that the specific ordering of leverage points is not a strict, universally applicable hierarchy that applies identically to every system — it's intended as a way of training intuition about where to look for higher-leverage interventions, and the actual relative leverage of any specific intervention still needs to be assessed within its own specific system context.
Where It Fails / Inversion
Where it fails / inversion
Not every situation calling for change actually has accessible high-leverage points available to the person trying to intervene — a mid-level manager, for instance, may have genuine, meaningful influence only over lower-leverage parameters within their own scope of authority, and correctly identifying a higher-leverage point that is nonetheless entirely outside your actual ability to influence doesn't produce useful change; realistic assessment of your own actual leverage over a given point matters as much as the point's theoretical leverage.
How To Use It
Worked example · improving an organization's chronically missed deadlines
A manager could adjust a low-leverage parameter (adding more explicit deadline-tracking software, a numeric metric) or instead examine whether the organization's underlying incentive structure or goals (what people are actually rewarded for) is the true source of the pattern — addressing the higher-leverage goal-and-incentive-structure level, though harder to identify and change, is more likely to produce a durable shift in behavior than repeatedly adjusting the lower-leverage tracking parameters alone.
How to use it
When trying to change a persistent pattern of behavior within a system, resist defaulting immediately to the easiest-to-adjust numeric parameter, and instead ask whether the system's underlying goals, incentive structure, or governing paradigm is the true source of the behavior — intervention at that higher level, though harder to identify and implement, tends to produce a far larger and more durable shift.
See Also