Incentive Structures

In 2016, Wells Fargo disclosed that its employees had opened approximately 3.5 million fake bank and credit card accounts without customers' knowledge or consent. The fraud was not the work of a few rogue individuals. It was systemic — spread across thousands of employees in hundreds of branches over more than a decade. Congressional hearings and media coverage focused on corporate greed and moral failure. But the deeper cause was structural: Wells Fargo had implemented an aggressive cross-selling incentive system that rewarded employees based on the number of new accounts opened per customer. Employees who failed to meet quotas faced termination. Employees who exceeded them received bonuses, promotions, and public recognition. The bank's stated value was "putting customers first." The real incentive — the one attached to money and job security — was to open accounts, regardless of whether customers wanted or needed them. The employees did not fail the system. The system produced exactly the behavior it was designed to reward.

Incentive structures are the systems of rewards and punishments — both formal and informal — that shape how people actually behave. They include salaries, bonuses, promotions, recognition, social approval, autonomy, fear of punishment, and dozens of other forces that nudge behavior in particular directions. This is NOT the same as motivation, which describes internal psychological drives. Incentive structures are external: they are the architecture of consequences within which motivation operates. A highly motivated teacher working inside an incentive structure that rewards only test scores will produce test preparation, not deep learning — not because they lack motivation, but because the structure channels their motivation toward the measured outcome. Understanding incentive structures means understanding that the behavior you observe in any system is almost always a rational response to the incentives that system provides.

Why the Gap Between Stated and Real Incentives Explains Most Dysfunction

The foundational insight of incentive design is that every organization, relationship, and system contains two sets of incentives: the stated ones (what the system claims to value) and the real ones (what actually gets rewarded and punished). Economist Steven Kerr documented this gap in his landmark 1975 paper "On the Folly of Rewarding A, While Hoping for B," which became one of the most reprinted articles in the history of management science. Kerr observed that organizations routinely announce one set of desired behaviors while systematically rewarding the opposite. They hope for teamwork but reward individual achievement. They hope for long-term investment but reward quarterly results. They hope for candor but reward those who tell leadership what it wants to hear.

Kerr's insight was not that organizations are hypocritical. It was that incentive misalignment is the default state — it takes deliberate, ongoing effort to keep stated and real incentives aligned, and most organizations never do that work. The gap between what you say you want and what you actually reward is where dysfunction lives. Charlie Munger, Warren Buffett's long-time business partner, distilled this into his most famous behavioral principle: "Show me the incentive and I'll show you the outcome." Munger considered incentive analysis the single most reliable tool for predicting human behavior, more reliable than character assessments, personality profiles, or stated intentions. The reason is straightforward: when doing the right thing is also the rewarded thing, you get good outcomes reliably. When doing the right thing requires people to sacrifice their own interests, you get heroism occasionally and dysfunction consistently.

The Mechanics of Misalignment

Incentive misalignment produces dysfunction through three specific mechanisms, each of which compounds over time.

The first mechanism is that misaligned incentives punish good behavior. When the real incentives contradict the stated values, people who try to do the right thing pay a cost. The doctor who spends extra time with patients sees fewer of them, earns less, and may face pressure from hospital administrators optimizing for throughput. The software engineer who flags technical debt instead of shipping features gets labeled as slow while the colleague who ships brittle code gets promoted for meeting deadlines. Economist George Akerlof described this dynamic in his Nobel Prize-winning work on information asymmetry: when a system cannot distinguish between high-quality and low-quality behavior, it inadvertently drives out the high-quality actors — a process Akerlof called "adverse selection." Applied to organizations, this means that incentive misalignment does not just produce bad behavior. It selectively retains and promotes the people who are willing to engage in it.

The second mechanism is that misalignment creates cynicism. People are perceptive about incentive gaps. When leadership declares that innovation is a core value but promotes only those who avoid risk and deliver predictable results, everyone notices. Over time, this erodes belief in the organization's stated values entirely. People stop taking mission statements seriously. They learn to game the system instead of improving it. Organizational psychologist Edgar Schein observed that culture is not defined by what leaders say — it is defined by what gets rewarded, what gets punished, and what gets ignored. The gap between speeches and incentives is where trust goes to die.

The third mechanism is compounding. Small misalignments between stated and real incentives do not stay small. They attract people who thrive under the real incentives and repel those who believed in the stated values. Over time, the population shifts. The culture drifts. What began as a slight gap between aspiration and reward becomes a chasm. This connects directly to Compound Growth — the same mathematical principle that makes small positive investments grow exponentially also makes small negative misalignments deteriorate exponentially. Wells Fargo's fake account problem did not begin as a crisis. It began as a modest quota system that, compounded over years, produced systemic fraud.

The Cobra Effect and Its Cousins

The most dramatic form of incentive failure is what economists call the cobra effect, named after an apocryphal story from British colonial India. The colonial government, concerned about the number of venomous cobras in Delhi, offered a cash bounty for every dead cobra brought to a collection point. Initially the program worked — people killed cobras and collected bounties. Then enterprising residents began breeding cobras specifically to kill them for the reward. When the government discovered the breeding operations and cancelled the program, breeders released their now-worthless snakes into the wild, leaving Delhi with more cobras than it had before the intervention.

Whether or not the Delhi story is historically precise, the dynamic it describes is thoroughly documented. Economist Horst Siebert, who coined the term "cobra effect," catalogued dozens of verified examples across policy, business, and public health. Hanoi's colonial rat bounty program — which required citizens to bring rat tails as proof of kill — produced a thriving business in tail-cutting, with rats released alive to breed more tails. The Soviet Union's nail factory quotas, measured by weight, produced useless giant nails; when quotas switched to quantity, factories produced tiny useless nails. Modern examples are equally striking: paying software testers per bug found can incentivize writing buggy code; rewarding firefighters for the number of fires they respond to can discourage fire prevention; bonuses tied to quarterly targets reliably incentivize short-term thinking at the expense of long-term value.

The cobra effect is not a curiosity. It is the predictable outcome of designing incentives without modeling the behavioral response. Every incentive is a signal about what the system values, and people will optimize for that signal — including in ways the designer never imagined.

Designing Better Incentive Structures

Designing effective incentive structures requires specific practices that counteract the natural drift toward misalignment.

The first practice is alignment auditing: before implementing any incentive, ask the question "If people optimized purely for this metric, would I be happy with the result?" This test, which management theorist Eli Goldratt embedded in his Theory of Constraints framework, catches most cobra effects before they deploy. Wells Fargo's cross-selling quota fails this test immediately: if every employee optimized purely for accounts opened, you would get millions of unwanted accounts. The failure was not that no one asked this question. It was that the people who would have asked it were not in the room where the incentive was designed.

The second practice is measuring what matters rather than what is easy. The most important outcomes — customer loyalty, employee engagement, long-term health, genuine learning — are harder to quantify than the proximate metrics — tickets closed, hours worked, test scores, procedures performed. But measuring the easy thing instead of the important thing is one of the most common sources of incentive misalignment. This connects directly to Goodhart's Law, which states that when a measure becomes a target, it ceases to be a good measure. The moment you attach rewards to a metric, people begin optimizing for the metric rather than the underlying reality it was supposed to represent. Goodhart's Law is not a bug in human behavior. It is the predictable consequence of incentive design.

The third practice is using balanced scorecards — multiple metrics that create productive tension with each other. Single metrics are almost always gameable. If you reward only revenue, quality suffers. If you reward only quality, speed suffers. The solution is designing metric sets where optimizing one dimension at the expense of others is structurally difficult. Intel co-founder Andy Grove formalized this approach in his management philosophy: every output metric should be paired with a quality metric that prevents gaming. Revenue paired with customer retention. Speed paired with defect rates. Growth paired with profitability. The tension between paired metrics forces people toward genuine performance rather than metric manipulation.

The fourth practice is modeling second-order effects. Every incentive creates ripple effects beyond the immediate behavior it targets. This connects to Leverage Points — the most effective interventions in a system are often not the obvious ones, and the most destructive incentives are often those whose second-order consequences were never considered. Before introducing a new incentive, think through at least two levels of consequences: "If people respond to this incentive, what will they do? And if they do that, what happens next?" The question is not difficult. The discipline of actually asking it is what most incentive designers lack.

Where Incentive Thinking Breaks Down

Incentive analysis is powerful, but treating it as a complete theory of human behavior produces specific errors.

Intrinsic motivation can be destroyed by extrinsic incentives. Psychologist Edward Deci's research on the "overjustification effect" demonstrated that adding external rewards to activities people already find intrinsically meaningful can reduce their internal motivation. Volunteers who are paid to volunteer sometimes volunteer less than those who receive nothing, because the payment reframes the activity from an expression of values to a transaction. This means that in domains where intrinsic motivation is strong — creative work, caregiving, community service — adding incentives can make behavior worse, not better. The assumption that m