Change what the system is optimizing for
If the goal of the system produces the behavior you want to change, change the goal.
Key takeaways
- What it is: If the goal of the system produces the behavior you want to change, change the goal.
- Why it works: In Meadows’ hierarchy, changing the goal of a system is higher-leverage than changing the rules, because the rules exist to serve the goal — if the goal stays the same, new rules will be gamed back toward the old behavior. Many intractable problems persist because everyone is actually succeeding at achieving a goal that produces the problematic behavior as a side effect; the solution is not to constrain the behavior but to change what success means.
- Evidence: Backed by observational / correlational evidence.
- Avoid: Designing a new goal metric without considering second-order effects — every goal metric will be optimized, including in ways you did not intend, so the new metric needs game-theory testing before implementation.
Why it works
In Meadows’ hierarchy, changing the goal of a system is higher-leverage than changing the rules, because the rules exist to serve the goal — if the goal stays the same, new rules will be gamed back toward the old behavior. Many intractable problems persist because everyone is actually succeeding at achieving a goal that produces the problematic behavior as a side effect; the solution is not to constrain the behavior but to change what success means.
How to do it
- 1Ask: "What goal is this system actually optimizing for?" — not the stated goal, but the one revealed by the behavior.
- 2Compare that to the goal you want. If they differ, identify who or what sets the current goal.
- 3Design a new goal metric that would make the desired behavior the natural path to success.
- 4Check for goal displacement: once the new metric is in place, is it producing unintended optimizations?
What the evidence says
ObservationalGoal displacement — systems optimizing for measurable proxies rather than actual outcomes — is extensively documented in organizational and public policy research. Goodhart’s Law ("when a measure becomes a target, it ceases to be a good measure") is the canonical formulation.
Honest caveat: Changing goals in social systems requires legitimacy and buy-in; technical correctness of a new goal does not guarantee adoption.
- — Goodhart (1975), cited in Marilyn Strathern (1997), "‘Improving ratings’: audit in the British university system"
Common mistake
Designing a new goal metric without considering second-order effects — every goal metric will be optimized, including in ways you did not intend, so the new metric needs game-theory testing before implementation.
IX Coach surfaces what your behavior over the past weeks suggests you are actually optimizing for — versus what you say you want — and helps you align the two by working on the implicit goal, not just the stated one.
Practice this with IX Coach →