The team works perfectly together.
Targets are met.
Roles are clear.
Information flows.
Everyone inside the group is pleased.
Outside the group, somebody else pays the cost.
A cooperation boundary is the line around whose goals, costs, risks and outcomes are represented inside the group’s shared success model—and whose consequences remain outside it.
This is the third pillar beneath How Cooperation Works. The master owns cooperation broadly. This page owns a harder question: can cooperation be internally successful while externally harmful?
How The World Works | Externalities remains the canonical owner of the general mechanism by which consequences escape a decision. Cooperation Boundary is narrower: it asks whether a cooperating group has defined its shared objective so narrowly that affected outsiders disappear from the coordination model.
Quick Read
Cooperation is not automatically good simply because coordination is strong. People can cooperate to cheat, exclude, cartelise, bully, dump costs, hide risk or optimise a local target at the expense of the larger system. The key diagnostic is the cooperation boundary: who counts inside the shared goal, who supplies resources, who bears risk, and who receives downstream consequences? Strong cooperation therefore needs an external receiver check. Before declaring success, ask whether the group created value, merely captured value, or shifted part of the cost onto people who had no voice in the decision. A healthy cooperative system can defend its shared goal at more than one zoom level: team, institution, community and wider world.
shared goal → define participating group → map affected outsiders → trace displaced costs/risks → test larger-system outcome → widen goal or add constraints → cooperate within the corrected boundary → observe world return
Cooperation Describes Coordination, Not Moral Goodness
Two people can cooperate to help a neighbour.
Two people can also cooperate to deceive someone.
The cooperative mechanism—shared goal, role allocation, trust, communication—can operate in both.
Evaluation therefore needs another layer:
what does the shared outcome do to people and systems beyond the cooperating group?
The Boundary Often Hides Inside the Metric
Metric:
Finish the project as fast as possible.
The group succeeds by:
- sending incomplete documentation downstream;
- creating cleanup work for another team;
- using a shortcut that shifts risk to users;
- ignoring accessibility or maintenance.
The local metric says “success.”
The larger system says “cost moved elsewhere.”
Local Optimisation Is a Cooperation Failure at the Larger Scale
A department hits its target by making another department’s work harder.
A class group gets a high grade because one student quietly carries the entire load.
A firm increases margin by creating environmental cleanup costs outside the transaction.
Inside the local boundary, cooperation may look efficient.
Zoom out and the definition of success changes.
The Receiver Outside the Group Must Be Represented
Ask:
- Who receives the output?
- Who maintains it?
- Who absorbs the error?
- Who loses time?
- Who has no voice but still bears consequence?
A cooperative system becomes more honest when the receiver is present in the success definition even if the receiver is not physically present in the room.
Externalities Are the General Mechanism
When consequences escape the decision-maker, economists describe positive or negative externalities.
The general concept belongs to How The World Works | Externalities.
This Cooperation child asks a narrower question:
did cooperation make the internal group so effective at its narrow goal that it became better at exporting the cost?
Collusion Is Cooperation With the Wrong Boundary
Competitors may cooperate secretly to protect themselves at the expense of customers or the wider market.
Within the cooperating subgroup:
- trust may be high;
- communication may be excellent;
- roles may be clear;
- incentives may align.
The cooperation architecture is functioning.
The boundary is illegitimate because harmed outsiders are excluded from the objective.
Group Loyalty Can Hide Boundary Problems
“Protect the team” can become:
- hide a mistake;
- do not report misconduct;
- defend a member regardless of evidence;
- exclude an outsider’s complaint.
Loyalty supports cooperation when it protects legitimate shared commitments.
It becomes dangerous when it shields the group from the world return.
A Strong Internal Culture Can Intensify Boundary Blindness
Everyone shares language.
Everyone trusts one another.
Everyone knows the internal priorities.
Now outsider criticism feels like misunderstanding rather than evidence.
The first sibling, Common Knowledge, owns shared internal state. Cooperation Boundary asks what that shared state leaves out.
Complementary Roles Can Optimise the Wrong Goal Efficiently
Researcher finds the loophole.
Analyst models the advantage.
Writer builds the justification.
Operator executes.
Role complementarity has increased capability.
It has not guaranteed a good outcome.
The second sibling, Role Complementarity, owns the fit among different contributions. Boundary analysis tests the larger objective those contributions serve.
Stakeholder Mapping Expands the Field of View
Before action, map:
- direct participants;
- direct receivers;
- people carrying downstream maintenance;
- people exposed to failure;
- people unable to opt out;
- future users;
- environment or shared resources where relevant.
Not every stakeholder gets equal decision authority.
But relevant consequences should not disappear simply because the affected party is outside the meeting.
The Boundary Can Be Temporal
Today’s group succeeds.
Tomorrow’s maintainer inherits technical debt.
Current students enjoy an easy shortcut.
Future students inherit a weakened system.
Future receivers belong in the boundary when current decisions create delayed consequences.
The Boundary Can Be Spatial
A local system reduces cost by moving waste, congestion or risk elsewhere.
The local dashboard improves.
The larger map deteriorates.
Zoom level is part of evaluation.
The Boundary Can Be Organisational
Sales closes the deal.
Operations inherits an impossible promise.
Sales metric rises.
Company-level reliability falls.
Good cooperation requires a success metric that spans the handoff, not one that stops at the departmental border.
The Boundary Can Be Ethical Even When Law Is Satisfied
An action may be legally permitted yet still produce avoidable costs for people with little bargaining power.
Legal compliance is one boundary.
Responsible cooperation may require a wider one.
The Shared Goal Should Include Constraints
Weak goal:
Maximise speed.
Stronger:
Maximise speed while preserving safety, accuracy, accessibility and maintainability above defined floors.
Constraints bring absent receivers back into the objective.
Boundary Checks Should Occur Before and After Action
Before: who might bear displaced cost?
After: who actually bore it?
The second check matters because systems often create consequences nobody predicted.
The world return should be allowed to correct the cooperation model.
Feedback From Outsiders Is Boundary Evidence
Affected receiver reports a problem.
The group says:
But our internal numbers are excellent.
The complaint may be evidence that the metric is incomplete.
How Feedback Works owns correction signals broadly. Cooperation Boundary owns whether the affected receiver had a route into the group’s definition of success.
Boundary Expansion Is Not Infinite Responsibility
No group can model every remote consequence perfectly.
The goal is proportionate scope:
- material effects;
- reasonably foreseeable effects;
- high-consequence uncertainty;
- affected parties with little ability to protect themselves.
Boundary awareness should improve decision quality without making action impossible.
Reconfiguration Can Shift Boundary Costs
A contributor leaves.
The group redistributes the work.
One quiet member absorbs almost everything.
The shared outcome survives, but internal fairness collapses.
The fourth sibling, Cooperation Reconfiguration, owns the continuity mechanism; Boundary checks whether the emergency solution simply exported the load to an invisible person.
A Practical Cooperation Boundary Map
SHARED GOAL: Finish group presentation by Friday INSIDE GROUP: 4 students DIRECT RECEIVER: Teacher/class audience DOWNSTREAM COSTS TO CHECK: - one student carrying hidden extra work - citation errors passed to audience - accessibility of visuals - late-night work caused by poor handoffs SUCCESS CONSTRAINTS: - no single member >40% of total workload without explicit agreement - sources verified - slides readable from classroom distance - blockers reported 24h before final deadline
A 30-Lens Cooperation Boundary Audit
- Group: who is inside the cooperative unit?
- Goal: what is being optimised?
- Metric: what counts as success?
- Direct receiver: who uses the output?
- Outsiders: who is affected without participating?
- Cost: what is shifted outside?
- Risk: who bears failure?
- Benefit: who captures gain?
- Voice: who lacks a route into the decision?
- Consent: who cannot opt out?
- Externality: what consequence escapes the group?
- Local optimisation: does one metric damage the larger system?
- Department boundary: is downstream work invisible?
- Temporal boundary: do future people pay?
- Spatial boundary: is cost moved elsewhere?
- Ethical boundary: legal but still harmful?
- Collusion: is cooperation protecting insiders at outsiders’ expense?
- Loyalty: is group solidarity hiding evidence?
- Common knowledge: what internal state dominates attention?
- Complementarity: is capability amplifying a narrow objective?
- Constraint: what safety/fairness floor belongs in the goal?
- Feedback: can outsiders report effects?
- Measurement: are external effects counted?
- Uncertainty: what high-consequence effect is unknown?
- Proportionality: how far must the boundary reasonably extend?
- Repair: what if external harm appears?
- Reconfiguration: does emergency continuity overload someone?
- Zoom: team, institution, society?
- Value creation: did cooperation create or merely capture value?
- World return: does the larger system improve when the group succeeds?
Laboratory 1: Zoom Out One Level
Take one successful group action. Re-evaluate it from the viewpoint of the next department, customer, family member or community affected by the output.
Laboratory 2: Add One Constraint
Choose a narrow target such as speed, marks or cost. Add one minimum constraint for fairness, safety, quality or maintainability and observe how the cooperative strategy changes.
Laboratory 3: Outsider Receipt
After a project, ask one downstream receiver what extra work, confusion or benefit the group created for them. Compare that receipt with the internal success metric.
For Primary Readers
A group can work together very well to make a mess for somebody else. Good cooperation asks not only “Did we succeed?” but also “What happened to everyone affected by our success?”
For Secondary Readers
Take one group success metric and identify at least two people outside the group who may carry costs or risks that the metric does not show.
For Advanced Readers
Model cooperation as optimisation over a bounded objective function. Internal coordination quality can rise while global welfare falls if costs are excluded from the objective. Robust cooperative design therefore requires stakeholder mapping, externality recognition, receiver feedback and cross-boundary constraints that prevent local success from being purchased through hidden displacement.
Common Misconceptions
- “Cooperation is always prosocial.” cooperation describes coordinated joint action; the shared goal still needs evaluation.
- “If everyone inside agrees, the outcome is fair.” outsiders may carry costs without voice or consent.
- “Externalities and cooperation boundary are the same topic.” Externalities owns the general escaping-consequence mechanism; this page owns how group boundaries omit affected receivers.
- “Zooming out means every group is responsible for everything.” use proportionate boundaries focused on material, foreseeable and high-consequence effects.
- “A good local metric proves a good system result.” local optimisation can improve the dashboard while worsening the next layer.
Research Corridor
- UNESCO — Recommendation on Education for Peace, Human Rights and Sustainable Development — cooperation, shared responsibility and wider social consequences.
- UNESCO — Global Citizenship Education — working with others on challenges that exceed one individual or nation.
- eduKateSG — How The World Works | Externalities — canonical general owner.
- eduKateSG — How Cooperation Works.
Frequently Asked Questions
Can cooperation be harmful?
Yes. A group can coordinate effectively around a goal that excludes or harms outsiders. Cooperation quality and social desirability are separate questions.
What is a cooperation boundary?
It is the boundary around whose goals, costs, risks and outcomes are represented inside the group’s shared success model.
How can a team avoid local optimisation?
Include downstream receivers, material external effects and minimum system constraints in the success definition, then inspect the world return after action rather than trusting the internal dashboard alone.
Final Thought: A Group Is Not the Whole World
Cooperation becomes mature when a group can coordinate strongly without confusing the edge of its membership with the edge of responsibility for the consequences it creates.
COOPERATION · FOUR PILLAR LEGS
Return to How Cooperation Works, or continue through Common Knowledge, Role Complementarity and Cooperation Reconfiguration. Return to the How X Works Hub.