A system can improve the number it measures while quietly damaging the purpose that number was supposed to represent.
Incentives become dangerous in an unusual way. They do not always fail because people ignore them. Sometimes they fail because people respond to them exactly as the system encouraged.
A target is established to improve performance. A metric makes the target visible. Rewards are attached so that people take it seriously. Behavior changes, the number improves, and the intervention appears successful.
Then something underneath the number begins to deteriorate.
Quality gives way to speed. Learning gives way to test performance. Service gives way to throughput. Judgment gives way to compliance. The system continues producing evidence of success while the mission that justified the system becomes harder to see.
This is the problem of perverse incentives: a structure rewards behavior that works against its stated purpose. The incentive itself may be logical. The response may be rational. The outcome can still be wrong.
This entry sits inside The Rational Field framework , which examines how perception, interpretation, structure and judgment shape what people believe and do. Here the question is structural: what happens when the measurement used to represent a mission becomes easier to optimize than the mission itself?
The Trouble Begins When Purpose Needs a Proxy
Most important goals are difficult to measure directly. Organizations want quality, trust, learning, safety, responsiveness, good judgment or meaningful progress. Those ideas are real, but they are not always easy to place on a dashboard.
So systems use proxies.
A school may use test performance as one signal of learning. A service organization may use response time as one signal of responsiveness. A business may use engagement as one signal that people find a product useful. A manager may use completed tasks as one signal of productivity.
There is nothing inherently wrong with this. Measurement often requires simplification. The problem appears when the proxy quietly stops being treated as a signal and starts being treated as the purpose itself.
A proxy is useful when it helps us see part of the mission. Distortion begins when improving the proxy becomes more important than improving the mission the proxy was designed to represent.
Once rewards attach to the proxy, the distinction becomes more important. People are no longer simply being measured. They are learning which visible outcome carries consequence.
That changes behavior.
People Learn the Game the System Actually Scores
Incentives communicate priorities more concretely than statements of value. If one outcome receives money, promotion, status, protection or relief from scrutiny, people have reason to direct effort toward it.
This is not necessarily cynicism. It is adaptation.
Consider a team told to solve customer problems but evaluated primarily on how quickly tickets are closed. Closing tickets and solving problems may often support each other. Under pressure, however, they can separate. The team can improve the visible measure by resolving easy cases first, closing ambiguous cases prematurely or discouraging complex requests.
From the dashboard’s perspective, performance improves. From the customer’s perspective, the purpose may be weakening.
The same dynamic can appear anywhere a narrow measure stands in for a broader objective. What gets rewarded receives attention. What receives attention becomes optimized. Once optimization becomes skilled, the system begins discovering every gap between the measure and the mission.
This is where the distinction from Incentives Explain Behavior becomes useful. That entry asks what operating incentives reveal about behavior. This one asks what happens after people become good at responding to those incentives.
How a useful target can begin competing with its mission
The problem rarely appears all at once. Distortion often develops as a reasonable measurement system becomes an increasingly powerful behavioral system.
The system begins with a broader outcome it genuinely wants to improve or protect.
A measurable signal is selected because the full purpose is difficult to observe directly.
Consequence attaches to the proxy, increasing the importance of moving the visible number.
People become increasingly skilled at producing the rewarded result, including through methods the original design did not anticipate.
Unmeasured parts of the mission begin losing time, attention, resources or status to the rewarded proxy.
The system gradually normalizes the rewarded behavior even when it no longer serves the purpose that justified the measure.
The System Can Be Working and Still Be Wrong
This is the uncomfortable part of perverse incentives. Failure does not always arrive as malfunction.
A school can raise scores while narrowing the education students receive. A company can increase engagement while making attention harder to control. A workplace can report greater productivity while encouraging employees to prioritize visible activity over consequential work.
In each case, the system may be functioning according to its design. People are responding to the reward. The metric is moving. Formal expectations are being met.
What has changed is the relationship between the scored outcome and the larger purpose.
This distinction matters because organizations often respond to disappointing outcomes by increasing pressure on the existing metric. If the measure is already distorting behavior, stronger incentives can intensify the distortion.
More pressure does not automatically produce more alignment.
When performance on the metric improves while the mission deteriorates, the first question should not be how to push the metric harder. It should be whether the metric is still carrying the meaning assigned to it.
That question also protects against another mistake: assuming people must be cheating whenever an incentive produces an unwanted result. Gaming can happen, but harmful outcomes do not require dishonesty. A person can follow the stated rule carefully and still contribute to an outcome the system did not intend.
Sometimes the deeper flaw is that the rule asked too little of judgment.
Good Incentive Design Looks Beyond the First Win
First-order thinking asks whether a reward produces the behavior being targeted. That question matters. If the incentive does not move behavior at all, the design may already be ineffective.
Second-order thinking asks what happens when people become excellent at responding to the reward.
What gets crowded out? Which behaviors become rational but undesirable? Who carries the hidden cost? What does the system stop seeing because the dashboard still looks healthy?
These questions move incentive analysis beyond intention.
A designer may have no desire to encourage shortcuts, superficial compliance or metric manipulation. That intention is relevant, but it does not control how the incentive functions once people begin navigating it.
Therefore, stronger design examines the behavior created at the edges, not only the behavior expected at the center.
Before strengthening the reward, inspect what the system may teach
Name the underlying outcome the system exists to improve, protect or produce before discussing the metric.
Identify the measurable behavior or outcome carrying money, status, approval, access or consequence.
Look deliberately for shortcuts, substitution, neglect and behaviors that improve the score without improving the underlying purpose.
Identify important outcomes the measurement system does not capture and determine whether optimization could damage them.
Examine whether one group receives the benefit of the metric while another absorbs the cost created by achieving it.
Establish evidence that would cause the incentive, target or measurement system to be reconsidered rather than defended automatically.
This check is useful precisely because people adapt. A fragile incentive assumes users will behave as designers imagined. A stronger incentive assumes people will learn the system, find its boundaries and direct effort toward whatever the system makes consequential.
That is not pessimism. It is design realism.
The Real Test Is Alignment
The solution to perverse incentives is not to remove every target, metric or reward. Organizations need ways to coordinate effort. Standards matter. Accountability matters. Measurement matters.
The deeper requirement is alignment.
Alignment coordinates purpose, priorities, authority, incentives, resources and action. Applied here, it asks whether the behavior being rewarded continues serving the purpose the system claims to value.
That relationship should be inspected repeatedly rather than assumed permanently. Conditions change. People learn. New workarounds appear. A useful measure can become less useful as the system around it evolves.
This is also why Structure Builds Freedom remains the governing principle. Better structure does not eliminate incentives. It keeps them inside a larger design that preserves judgment, exposes side effects and makes correction possible before distorted behavior becomes normal.
That may require more than one measure. It may require qualitative review, countermetrics, protected discretion, direct observation or periodic examination of outcomes the dashboard does not capture.
The objective is not measurement perfection. No metric can contain the entire reality it represents.
The objective is to prevent the measurable part from quietly replacing the whole.
That is the discipline going forward.
The Rational Field does not reject incentives because they can distort behavior. It asks whether the structure around those incentives can detect distortion before distortion becomes the operating norm.
Good systems ask what behavior they want. Better systems also ask what behavior their rewards will produce after people become skilled at pursuing them.
Every system teaches.
Rational design requires checking whether the lesson still matches the purpose.
Continue Through The Rational Field
The Principle and Condition Beneath This Work
This article applies one Groundwork Daily governing principle and one structural condition to incentive design, reward distortion and second-order consequences.
Structure Builds Freedom
Strong structure keeps incentives inside a wider operating design so that targets, measures and rewards remain answerable to the purpose they were created to support.
Alignment
Alignment coordinates purpose, priorities, authority, incentives, resources and action so that improving the visible target does not quietly undermine the underlying mission.
See the full Groundwork Daily Core Principles and Conditions architecture .