Leverage Points

TL;DR

Leverage Points: In any complex system, some places are much more powerful intervention points than others. Donella Meadows ranked 12 leverage points from weakest (adjusting numbers) to strongest (changing the paradigm). The most powerful leverage points are often counterintuitive — and most problem-solving targets the weakest ones.

What Is a Leverage Point?

Donella Meadows, one of the most important systems thinkers of the 20th century, defined a leverage point as "a place within a complex system where a small shift in one thing can produce big changes in everything." The concept comes from Archimedes — give me a lever long enough and I can move the world. The challenge in systems thinking: finding the lever.

Meadows published her landmark paper "Leverage Points: Places to Intervene in a System" in 1999. Her key insight was not just that leverage points exist, but that they have a hierarchy — some are far more powerful than others — and that systems thinkers and policy designers routinely target the weakest leverage points while avoiding the most powerful ones.

The weak leverage points feel comfortable: they involve adjusting numbers (changing a tax rate, changing a budget line item) or changing parameters within an existing structure. These rarely produce lasting change because the system structure remains intact.

The powerful leverage points are uncomfortable: they require changing the rules of the system, changing the goals that the system is designed to achieve, or — at the highest level — changing the paradigm or worldview from which the system emerged. These are fought by every institution that benefits from the current system.

How It Works

Meadows' 12 Leverage Points (weakest to strongest):

LOW LEVERAGE:
12. Numbers (constants and parameters)
    "Change the tax rate by 2%"
    → Rarely changes system structure; most political fights occur here
    
11. The size of buffers and stocks
    "Build larger strategic reserves"
    → Useful but often expensive and slow
    
10. Structure of material flows
    "Change the physical layout of the supply chain"
    → Physical changes are costly and slow
    
9. Length of delays
    "Improve information lag in the supply chain"
    → High leverage for specific oscillation problems

MEDIUM LEVERAGE:
8. Strength of negative feedback loops
   "Strengthen regulation/monitoring mechanisms"
   
7. Gain of positive (reinforcing) feedback loops
   "Reduce the rate at which successful systems grow"
   
6. Structure of information flow
   "Who has access to what information, and when?"
   → Often very high leverage despite being mid-ranked
   
5. Rules of the system (incentives, constraints)
   "Change the laws that govern behavior"

HIGH LEVERAGE:
4. The power to change system structure
   "Change who has authority to modify the rules"
   
3. The goals of the system
   "Change what the system is designed to optimize for"
   
2. The mindset or paradigm from which the system arises
   "Change the shared beliefs that gave rise to the system's rules"
   
1. The power to transcend paradigms
   "Holding no paradigm as absolute; remaining open to change"

Three Real-World Examples

Carbon Taxes vs. Information Flow in Climate Policy

Carbon taxes are a #5 intervention (changing rules/incentives). Economists generally favor them, and they can be effective. But the #6 intervention — information flow — may be more immediately powerful: requiring mandatory disclosure of carbon emissions on products, making the environmental cost visible to consumers at the point of purchase, changes decision-making without changing prices.

Neither targets the highest leverage points. The #3 intervention (goals) would mean changing the system's optimization target from GDP growth to some measure that includes environmental sustainability. The #2 intervention (paradigm) would mean changing the widely shared belief that economic growth and environmental protection are in tension — a paradigm shift that would change the rules, goals, and behaviors simultaneously.

Goodhart's Law as a Leverage Point Failure (Information Flow)

When a metric becomes a target (Goodhart's Law), it ceases to be a good measure. This is a failure at leverage point #6: information flow. The information that was supposed to drive good decision-making gets corrupted because agents now optimize for the information rather than the underlying goal.

The fix is a higher-leverage intervention: change the goal (#3) so it can't be easily gamed, or change the paradigm (#2) so that gaming metrics is seen as the failure it is rather than rational optimization. Trying to fix Goodhart's Law by choosing better metrics (#12, adjusting numbers) produces a cycle: the new metric gets gamed too.

The Power of Changing Information Flow (Glassdoor)

Before Glassdoor, salary information was highly asymmetric: employers knew what everyone earned; employees had almost no information about market rates. This information asymmetry benefited employers in salary negotiations. Glassdoor changed the information flow structure (#6), giving employees access to salary benchmarks.

The effect on the employment market was significant and largely irreversible once the information was public. No individual hiring manager decided to implement it; no regulation required it. A change in information structure (who knows what, when) shifted power relationships across the entire market.

When to Use It

✅ Use Leverage Point thinking when:

Designing a systems intervention and wanting to maximize impact per unit of effort
Diagnosing why repeated interventions in a system haven't produced lasting change
Prioritizing among multiple potential interventions
Understanding why entrenched problems are hard to solve

❌ Limit when:

The problem is genuinely simple and linear — not all problems are systems problems requiring leverage point analysis
Quick, reversible interventions are needed — high-leverage interventions often require sustained political effort

Pairs well with	Why
Feedback Loops	Leverage points 6–9 target specific loop structures
Unintended Consequences	High-leverage interventions produce larger unintended effects
Second Order Effects	Leverage point thinking is extended second-order analysis
Goodhart's Law	Goodhart failures often occur at leverage point #6

Common Misuses

Assuming the highest-leverage intervention is always right. Paradigm shifts (#2) and changing goals (#3) are very powerful but may be wrong. The history of utopian political movements is full of people who believed they had the right paradigm. High leverage points require high epistemic humility.

Confusing leverage with ease. High-leverage interventions are often politically and practically difficult precisely because they threaten the systems that current power-holders benefit from. Meadows is explicit: the most powerful leverage points are the most contested.

Using leverage point language without doing the systems analysis. Claiming something is "high leverage" without mapping the system's feedback loops, goals, and information structures is just assertion. The framework requires the prior analytical work.

Feedback Loops — loops are the structural elements that leverage points operate on
Goodhart's Law — a specific leverage point failure (information flow corruption)
Theory of Constraints — finding the binding constraint is finding the leverage point for a specific throughput problem

FAQ

Why do people keep targeting low-leverage points?

Meadows identified this as one of the central paradoxes: the low-leverage points (#12, #11) are exactly where most political, organizational, and policy fights happen — because they're visible, quantifiable, and technically easier to change. Numbers can be adjusted without disrupting power structures. Changing the rules (#5) or goals (#3) requires confronting the people who benefit from the current ones — which generates fierce resistance.

Is the leverage point hierarchy always in this order?

Meadows offered it as a general guide, not an absolute law. She acknowledged that sometimes a parameter change can be surprisingly high-leverage (if it affects a key threshold), and that paradigm changes don't always work as expected. The hierarchy is a strong heuristic for identifying where to look, not a guarantee of outcomes.

What does "transcending paradigms" (#1) mean in practice?

It means holding your own worldview lightly — recognizing that every paradigm is a model of reality, not reality itself, and remaining genuinely open to paradigm shifts rather than defending your current one. Meadows observed that this is both the highest leverage and the hardest for most people: it requires epistemic humility about the frameworks that constitute your identity and professional training.

Apply with AI

🚀 Find leverage points in your system with MindMax →

What Is a Leverage Point?​

How It Works​

Three Real-World Examples​

Carbon Taxes vs. Information Flow in Climate Policy​

Goodhart's Law as a Leverage Point Failure (Information Flow)​

The Power of Changing Information Flow (Glassdoor)​

When to Use It​

Common Misuses​

Related Models​

FAQ​

Apply with AI​

Further Reading​