Why Cooperation Breaks Down in the Prisoner’s Dilemma

Hello again. Last time, you learned to distinguish a stable outcome from a good shared outcome: people can be individually “stuck” in a pattern that leaves everyone worse off. The Prisoner’s Dilemma is the clearest version of that trap.

In this lesson, you will play and interpret a short dilemma, then diagnose why cooperation failed. The aim is not to label people as selfish or untrustworthy. It is to notice how a particular setup can make withholding help look sensible to each person, even when mutual help would have been better for both.

---

A one-round dilemma: contribute or withhold?

Imagine two people are briefly assigned to improve a shared document before a meeting. Each can either:

• Contribute: spend a small amount of time adding a useful section that helps the other person.
• Withhold: save that effort and add nothing.

The interaction happens once. They choose privately and at the same time. Neither can guarantee what the other will do.

Here are their point outcomes. Higher is better.

Your choice	Other person contributes	Other person withholds
You contribute	$2,2$	$-1,3$
You withhold	$3,-1$	$0,0$

The first number is your result; the second is the other person’s.

The scores reflect a simple structure:

• Contributing costs you $1$ point of effort but gives the other person $3$ points of benefit.
• If both contribute, each gives $3$ and pays $1$, for a net result of $2$.
• If neither contributes, neither gains or loses anything.

Run the mini-game

If you have a willing partner nearby, each of you should privately write Contribute or Withhold, without discussing which action you will choose. Reveal your choices simultaneously, locate the matching cell, and record the result.

If you are studying alone, do a simulation: first choose the action you think you would take, then inspect all four possible results as though you had just learned the other person’s choice.

Before interpreting the result, separate two questions that are often blended together:

1. What happened? Which cell did you reach?
2. What pressure did the game place on each person? Given the other person’s action, was changing your own action attractive?

The second question is the game-theory diagnosis.

---

The two comparisons that create the trap

Start by assuming the other person contributes.

You compare:

• Contribute: $2$
• Withhold: $3$

Withholding gives you a better personal result. You still benefit from the other person’s effort, but avoid paying the effort cost yourself. This is the temptation to free ride.

Now assume the other person withholds.

You compare:

• Contribute: $-1$
• Withhold: $0$

Again, withholding is better. Contributing alone means paying a cost without receiving the reciprocal benefit. This is the desire to avoid being the only contributor, sometimes called the sucker’s payoff.

So the same action, withhold, looks better in both situations:

If the other person…	Your better-scoring action	What drives it
Contributes	Withhold	Gain from their effort without paying your cost
Withholds	Withhold	Avoid contributing alone
Is unknown	Withhold can seem safest	You cannot rely on reciprocal effort

When one action is better regardless of what the other person does, it is a dominant strategy. In this mini-game, withholding is dominant for each participant.

That does not mean either person prefers the joint outcome of mutual withholding. They do not: $0,0$ is worse for both than mutual contribution, $2,2$. But each person can see an individual reason not to be the contributor.

This gives the characteristic Prisoner’s Dilemma result:

• Mutual contribution is the better shared outcome.
• Mutual withholding is the stable outcome.
• A person who expects cooperation has an incentive to exploit it.
• A person who expects withholding has an incentive to protect themselves.

The result is not a mysterious collapse of goodwill. It is a predictable response to the combination of temptation and fear of being left carrying the cost alone.

---

See the logic played out visually

The video uses “spare” and “sacrifice” rather than “contribute” and “withhold,” but the structure is the same. Do not get distracted by the story’s dramatic language. A Prisoner’s Dilemma is defined by the incentives in the table, not by betrayal in a literal or moral sense.

---

Interpreting your mini-game result

Use the outcome you observed, or one of the simulated outcomes, as evidence about the game’s pressures.

Result	What it shows	Careful interpretation
Both contribute: $2,2$	Cooperation occurred this round.	It does not prove the setup is safe for cooperation. Each person could still improve their individual score by withholding while the other contributes.
You contribute, other withholds: $-1,3$	One person bore the cost while the other captured the benefit.	Cooperation broke down asymmetrically. The withholder had an immediate incentive to free ride; the contributor either expected reciprocity or chose to help despite the risk.
You withhold, other contributes: $3,-1$	The same asymmetry, with roles reversed.	Do not infer bad character from one choice alone. The payoff structure rewards withholding when the other contributes.
Both withhold: $0,0$	Neither person took the risk of contributing alone.	Cooperation failed through mutual self-protection. Neither can improve their outcome by contributing alone, even though both would prefer $2,2$.

The last result is especially important. It is tempting to say, “They both failed to cooperate because neither trusted the other.” That may be partly true, but it is incomplete.

A fuller diagnosis is:

Each person faced an uncertain, one-time choice. Contributing created value if reciprocated, but imposed a loss if unreciprocated. Withholding avoided that loss and also offered an extra gain if the other contributed. Since both could reason this way, mutual withholding became stable.

This diagnosis identifies incentives, uncertainty, and simultaneous choice. It does not assume anyone is irrational, hostile, or incapable of cooperation.

---

Compare with the original prisoners

In the classic story, “cooperate” means staying silent and “defect” means testifying against the other prisoner. The labels can sound morally loaded, but treat them as neutral game actions.

The decisive reasoning is exactly the same as in the document-contribution mini-game:

• If the other prisoner stays silent, testifying is personally better than staying silent.
• If the other prisoner testifies, testifying is still personally better than staying silent.
• Because each prisoner faces that same comparison, both testify.
• Both testifying is worse for each prisoner than both staying silent.

The symbols often used in this kind of matrix are useful shorthand:

• $R$: the reward for mutual cooperation
• $T$: the temptation payoff for defecting while the other cooperates
• $P$: the punishment payoff for mutual defection
• $S$: the outcome for cooperating while the other defects

A genuine Prisoner’s Dilemma has the ranking:

$$T>R>P>S$$

In plain language: exploiting a cooperator is individually most attractive; mutual cooperation is better than mutual withholding; mutual withholding is better than being the sole cooperator.

---

A practical diagnosis: is this really a Prisoner’s Dilemma?

Not every disappointing group outcome has this structure. Before declaring “a Prisoner’s Dilemma,” check three features.

1. Mutual cooperation must be better for everyone than mutual withholding.
   If people would not jointly benefit from cooperation, you may be looking at a conflict of interests rather than a cooperation problem.

2. Withholding must be individually attractive when the other cooperates.
   This is the free-rider temptation.

3. Withholding must also be individually attractive when the other withholds.
   This is the protection against being the sole contributor.

If either of the last two conditions fails, the situation may instead be a coordination problem. For example, two colleagues choosing between meeting platforms may simply need to choose the same platform; neither necessarily gains by choosing a different one.

OpenStax’s account of the original dilemma draws the individual-versus-group distinction sharply:

---

Don’t over-explain a result with personality

A mini-game result tells you something about a strategic environment, not the whole inner life of the people in it.

Suppose a team repeatedly fails to maintain a shared knowledge base. “People do not care” may be one possible explanation, but it is weak unless you have evidence. A Prisoner’s Dilemma analysis looks first at the design:

• Does an individual contributor pay a visible time cost?
• Can a non-contributor still receive the shared benefit?
• Is it easy for others to notice who contributed?
• Is the interaction truly one-off, or will the same people meet again?
• Can people communicate, make commitments, or repair a breakdown?

For this lesson’s one-round game, the first two questions do most of the work. The individual benefit of withholding is built into the payoff structure; no amount of vague encouragement removes it by itself.

The remaining questions point toward ways real relationships can differ from the stripped-down mini-game. Those differences matter, but they change the game rather than refute its logic. The next module will examine one especially important change: what happens when people expect to interact again.

---

Key takeaways

A Prisoner’s Dilemma is not simply “people refusing to cooperate.” It is a situation where:

• both participants would prefer mutual cooperation to mutual withholding;
• each can gain by withholding when the other cooperates;
• each can avoid an even worse outcome by withholding when the other withholds;
• those individual incentives make mutual withholding stable.

To interpret a mini-game result, identify the outcome first, then explain the conditional incentives each person faced. Mutual withholding usually reflects both temptation to free ride and fear of contributing alone, amplified by uncertainty and the absence of a guaranteed response.

Next, you will compare one-shot and repeated interactions. That comparison explains why cooperation can become more realistic when today’s choice affects tomorrow’s relationship.