Incentive Structure Audit
Trace rewards and penalties before judging how a system behaves
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 92%
The audit starts with a repeated behavior, then works backward through the rewards, penalties, delays, and displaced costs surrounding it. In the episode, the host proposes this lens to explain why officers were discouraged from making crack arrests: Dowd says an arrest could create extensive overtime, remove officers from patrol, and add pressure elsewhere in the justice system. Those conditions did not justify his corruption, which he repeatedly accepts as his responsibility, but they help explain why the formal goal of enforcement diverged from street-level behavior. The mechanism is diagnostic: compare what an institution says it wants with what participants gain or lose by acting. Then change a relevant incentive and observe whether behavior follows. Treat the result as a testable explanation, not proof that incentives are the only cause.
Origin
Extracted from The Diary of a CEO
Core principles
- 01Observed behavior often follows practical incentives
- 02Formal goals can conflict with operational rewards
- 03Costs displaced onto workers or other units still shape behavior
- 04An incentive explains conduct without excusing misconduct
How to run it
- 1
Name the repeated behavior
Describe what people consistently do, using observable actions rather than assumptions about motives.
Pro tip Choose one behavior narrow enough to trace through the system.
Watch out A striking anecdote does not establish a general pattern.
- 2
Record the formal goal
State what the policy, role, or institution officially asks participants to achieve.
Pro tip Use the actual rule or target when it is available.
Watch out Do not assume the stated goal is the only goal the system manages.
- 3
Map practical consequences
List the immediate financial, social, workload, career, and safety consequences of following or resisting the desired behavior.
Pro tip Ask participants what happens after the action, not only what the handbook says.
Watch out Participant accounts are evidence to verify, not automatically established fact.
- 4
Trace displaced costs
Follow the action into adjacent teams, budgets, queues, and time periods to find costs hidden from the local metric.
Pro tip Map who pays now, who pays later, and who receives the benefit.
Watch out Optimizing one unit can worsen the whole system.
- 5
Test the incentive
Change one consequential reward, penalty, or bottleneck and monitor whether the target behavior changes.
Pro tip Use a bounded pilot with a before-and-after measure.
Watch out A correlation after the change may still have other causes.
In the wild
Dowd says a crack arrest could generate about 18 hours of overtime, take officers away from patrol, and add work across corrections and courts. He says officers were directly discouraged from making arrests because the department was managing those pressures. The host interprets this as an incentive structure that favored keeping officers on patrol, while Dowd describes exploiting the resulting lack of street control for personal gain.
→ The lens distinguishes an operational incentive from the personal crimes it neither required nor excused.
A service team is formally told to solve customer problems but is evaluated only on tickets closed per hour. An operator maps that reward, the penalty for long investigations, and the repeat contacts created by superficial closures. The team pilots a metric that combines resolution quality with speed.
→ The revised measure tests whether the original incentive was producing fast but incomplete work.
Common mistakes
Turning explanation into excuse
System pressures can help explain conduct without removing personal responsibility for illegal or harmful choices.
Stopping at the visible metric
A local reward may hide costs transferred to another team, budget, or future period.
Is it for you?
Best for
It is best for leaders, operators, and policymakers diagnosing persistent behavior across teams or institutions.
Not ideal for
It is not ideal as a substitute for individual accountability, legal analysis, or evidence about causes beyond incentives.
From the transcript
“If you look at any system, you'll understand why people behave they do if you understand the incentive structure.”
“The average amount of overtime for one crack arrest was 18 hours.”
From the episode
Corrupt Police Officer: I Arrested Drug Dealers, Then Sold Their Drugs! There’s Massive Corruption Going On Today! - Mike Dowd