Quality
and the Prisoner’s Dilemma: When Every Department Optimizes for Itself —
and the Whole Organization Loses
There’s a scene that plays out in manufacturing plants every single
day, and if you’ve spent any time in quality, you’ve watched it
happen.
Production wants to hit their numbers. Purchasing wants to cut
material costs. Engineering wants to push the design envelope. Logistics
wants to minimize inventory. And quality? Quality wants zero
defects.
Each department, acting perfectly rationally from its own
perspective, makes decisions that seem smart. Production skips a
secondary inspection to save twelve minutes. Purchasing switches to a
cheaper supplier who’s “basically the same.” Engineering releases a
drawing with a tolerance that looks great in CAD but is impossible to
hold on the shop floor. Logistics pushes for smaller safety stocks that
leave no buffer when something goes wrong.
And then the customer finds the defect. Not the supplier. Not the
drawing. Not the schedule. The customer.
What happened wasn’t a failure of intelligence. It wasn’t
incompetence or malice. It was something far more dangerous because it’s
far more subtle: every department played the game correctly — but they
were all playing different games, and the organization
lost.
This is the Prisoner’s Dilemma, and it’s eating your quality system
from the inside out.
The Classic Problem,
Reimagined
The original Prisoner’s Dilemma is elegant in its simplicity. Two
suspects are arrested. Each is interrogated separately. Each can either
cooperate (stay silent) or defect (betray the other). If both cooperate,
they each get a light sentence. If both defect, they each get a heavy
sentence. But if one defects while the other cooperates, the defector
goes free and the cooperator gets the maximum penalty.
The tragic logic is that from each prisoner’s individual
perspective, defecting is always the rational choice — regardless of
what the other does. Yet when both follow this logic, they both end up
worse off than if they had both cooperated. Individual rationality
produces collective irrationality.
Now replace “prisoners” with “departments” and “confessing” with
“optimizing local metrics,” and you have the single most common
structural failure mode in manufacturing quality.
How It Plays Out on Your
Shop Floor
Let me show you what this looks like in practice, because it’s not
theoretical. It’s happening right now.
The Production-Quality Dilemma. Production is
measured on throughput. Quality is measured on defect rates. These
metrics exist in tension. When production pushes speed and quality
pushes caution, each department is acting on the incentives it was
given. But the optimal outcome — high throughput and low
defects — requires cooperation: production investing in setup
discipline, quality investing in prevention rather than detection, and
both sharing information in real time.
Instead, what usually happens is production hides early warning signs
(“it’s probably fine, let’s keep running”) and quality reacts by adding
more inspection steps that production then resents. Both defect. Both
lose. The customer loses most.
The Purchasing-Quality Dilemma. Purchasing saves 15%
on a critical component by switching suppliers. The new supplier’s
material has a slightly higher defect rate — nothing dramatic, just a
few percent more. But those defects don’t show up in purchasing’s
scorecard. They show up in scrap, rework, customer complaints, and
warranty claims — costs that belong to quality, production, and the
customer. Purchasing “wins” its metric while the organization pays far
more than it saved.
The Engineering-Quality Dilemma. Engineering designs
to the theoretical ideal, pushing tolerances tighter because modern CAD
software makes it easy to specify ±0.01mm on everything. Quality
inherits a drawing that’s technically achievable under laboratory
conditions but impossible to maintain during a Friday night shift with
tooling that’s approaching its wear limit. Engineering met its design
intent. Quality inherited a manufacturing nightmare.
In every case, the pattern is identical: local optimization produces
global suboptimization. Each department made the choice that maximized
its own outcome. And the total system performed worse than if any single
department had simply cooperated.
Why This Is So Hard to See
The Prisoner’s Dilemma is insidious because the people involved
aren’t doing anything wrong. They’re doing exactly what their
performance reviews, their KPI dashboards, and their managers are asking
them to do. They’re being good employees — and that’s precisely
the problem.
When a production supervisor pushes to keep a line running despite a
minor quality concern, he’s not being reckless. He’s responding to the
fact that his bonus depends on OEE and his boss will ask why the line
was down. When a purchasing manager negotiates a cheaper price, she’s
not being irresponsible. She’s responding to a cost-reduction target
that her performance review depends on.
The system is perfectly designed to produce the outcomes it’s
getting. And the outcomes it’s getting are the worst possible collective
result from individually rational decisions.
This is why exhorting people to “work as a team” or “think about the
big picture” rarely works. You’re asking people to act against their own
rational self-interest based on nothing more than good intentions. The
moment pressure increases — and in manufacturing, pressure
always increases — people revert to their incentives. Every
time.
The Iterated Version:
Why It Gets Worse
In game theory, there’s an important distinction between a one-shot
Prisoner’s Dilemma and an iterated one — where the same players
face the same choice repeatedly. Manufacturing is always iterated. The
same departments interact every day, every shift, every week.
In iterated dilemmas, a famous strategy called “tit-for-tat” tends to
emerge: you start by cooperating, and then do whatever the other player
did last time. If they cooperated, you cooperate. If they defected, you
defect. It’s simple, and it’s remarkably effective — in theory.
In practice, what happens in organizations is worse. Because
departments don’t just respond to what happened last time — they respond
to what they expect will happen next time. And once trust
erodes, expectations become self-fulfilling.
Production expects quality to slow them down, so they stop flagging
early issues. Quality sees production hiding issues, so they add more
gates. Engineering sees both departments fighting, so they stop
consulting either one during design reviews. Each round of interaction
reinforces the belief that the other side won’t cooperate, so nobody
cooperates, and the cycle deepens.
This isn’t a personality problem. It’s not a leadership problem
(though bad leadership makes it worse). It’s a structural
problem — and it requires a structural solution.
Escaping the Trap: What
Actually Works
The research on the Prisoner’s Dilemma gives us a clear roadmap for
how to shift from defection to cooperation. The solutions aren’t
mysterious. They’re just hard to implement because they require changing
the structure, not just the people.
1. Change the Payoff Matrix
The most powerful intervention is also the most obvious: make
cooperation more rewarding than defection. If production’s bonus depends
partly on defect rates and quality’s bonus depends partly on throughput,
both departments suddenly have skin in each other’s game. The payoff
matrix shifts so that the individually rational choice is the
collectively optimal choice.
This is what balanced scorecards and shared KPIs are supposed to do.
But they only work when the shared metrics carry real weight — when
they’re tied to compensation, recognition, and career advancement, not
just printed on a poster in the break room.
I’ve seen plants where production and quality share a single “right
first time” metric that determines 40% of both departments’ variable
compensation. The cultural shift is dramatic. Conversations change from
“your defect rate is killing my throughput” to “what do you need from us
to hit our shared target?” Same people. Same equipment. Different
incentives. Different outcome.
2. Increase the Shadow of the
Future
In game theory, cooperation increases when players expect to interact
many more times in the future. The “shadow of the future” — the
expectation of continued relationship — makes short-term defection less
attractive because there will be consequences tomorrow.
In organizations, this means building stable, long-term
cross-functional relationships. Rotating people through different
departments. Creating permanent cross-functional teams rather than ad
hoc problem-solving groups. When the production manager used to be a
quality engineer, cooperation isn’t abstract — it’s personal.
Job rotation programs, cross-functional project teams, and joint
Gemba walks all extend the shadow of the future. They transform the
interaction from a series of one-shot games into a genuine iterated
relationship where reputation matters and cooperation pays dividends
over time.
3. Make
Communication Cheap and Defection Visible
The classic Prisoner’s Dilemma assumes the players can’t communicate.
In reality, your departments can communicate — but the
communication is often filtered through reports, dashboards, and
meetings that obscure more than they reveal.
Making communication direct, frequent, and transparent reduces the
information asymmetry that drives defection. Daily stand-ups that
include production, quality, and maintenance. Shared digital dashboards
that show real-time quality and production data side by side. Visual
management boards that make problems visible to everyone
simultaneously.
When production can see a defect trend building in real time — not in
a weekly quality report that arrives three days after the trend started
— they can adjust before it becomes a crisis. When quality can see that
a machine is approaching its maintenance window, they can plan
inspections around it rather than being blindsided by a spike in
variation.
Transparency doesn’t solve the incentive problem, but it makes it
visible. And visible problems are far easier to address than invisible
ones.
4. Create Enforcement
Mechanisms
In the academic literature, the most reliable way to sustain
cooperation in a Prisoner’s Dilemma is to change the game entirely — to
add costs to defection. In organizations, this means governance
structures that make local optimization at the expense of the whole
system visible and consequential.
Escalation paths that bring cross-functional conflicts to senior
leadership quickly. Regular cross-departmental reviews where trade-offs
are made explicit. A quality council with genuine authority to resolve
disputes between departments, not just recommend improvements.
This isn’t about punishment. It’s about creating a structure where
the costs of defection are borne by the defector, not shifted to the
downstream department. When purchasing knows that a cheaper supplier’s
defects will be traced back to the sourcing decision and reflected in
their performance review, the calculation changes.
5. Build Identity Beyond
the Department
Perhaps the most powerful long-term intervention is also the most
subtle: shifting people’s primary identity from their department to the
organization. When people see themselves as members of the plant first
and members of their department second, the calculus of the Prisoner’s
Dilemma changes. Cooperation stops being a favor you do for another
department and starts being a natural expression of shared purpose.
This is what “quality culture” actually means — not posters and
slogans, but a genuine shift in how people identify with the outcomes of
the whole system. It’s built through shared experiences, shared
successes, shared failures, and shared accountability. It’s slow. It’s
hard to measure. And it’s the single most powerful defense against the
Prisoner’s Dilemma in manufacturing.
The Diagnostic: Is
Your Organization Trapped?
Here’s a simple test. Walk through your plant and listen to the
language people use. If you hear “your problem” more than “our problem,”
you’re in the dilemma. If department meetings spend more time assigning
blame than solving problems, you’re in the dilemma. If metrics are
discussed in isolation — production numbers without quality context,
quality numbers without production context — you’re in the dilemma.
Another signal: look at your escalations. When a cross-functional
conflict reaches senior leadership, what happens? If the resolution is a
compromise that satisfies neither party but keeps the peace, you’re
managing the symptoms. If the resolution addresses the structural
incentive that drove the conflict, you’re solving the real problem.
And perhaps the most telling signal of all: ask yourself what happens
when a defect escapes to the customer. Is the first question “whose
fault was this?” or “what in our system allowed this to happen?” The
former is the language of defection. The latter is the language of
cooperation.
The uncomfortable truth
The Prisoner’s Dilemma teaches us something most quality
professionals don’t want to hear: your biggest quality problems aren’t
caused by bad people making bad decisions. They’re caused by good people
making rational decisions inside bad systems.
The operator who skips a check isn’t lazy — he’s responding to
pressure to hit a rate. The engineer who over-specifies a tolerance
isn’t arrogant — she’s responding to a design review process that
rewards precision over manufacturability. The purchasing manager who
switches suppliers isn’t reckless — he’s responding to a cost target
that ignores downstream quality costs.
If you want different outcomes, you need a different system. Not
different slogans. Not different people. A different structure of
incentives, information, and identity that makes cooperation the
rational choice — not just the noble one.
The Prisoner’s Dilemma has a mathematical proof that individual
rationality can produce collective catastrophe. Your organization is
living that proof every day. The question isn’t whether you’re caught in
it. The question is whether you have the courage to change the game.
Peter Stasko is a Quality Architect with 25+ years
of experience transforming organizations across automotive, aerospace,
and pharmaceutical industries. He has spent his career helping companies
see the systems behind their defects — and building structures that make
excellence the path of least resistance.
