Rules are shared expectations made explicit enough to guide behaviour repeatedly.
In one line: rules work when people can understand what is expected, believe the rule has a legitimate job, predict what happens when it is followed or broken, and see that the rule is applied consistently enough to coordinate behaviour.
Evidence boundary: Rules are studied across law, governance, education, psychology, economics and organisations. This article explains common mechanisms—clarity, legitimacy, incentives, enforcement, exceptions, feedback and review—without claiming that every good rule must have the same form or penalty.
People often think a rule works because it has been written down.
But a written sentence does nothing by itself. A rule becomes operational only when people notice it, interpret it similarly enough, connect it to a real purpose, expect meaningful consequences and adjust behaviour accordingly.
What Is a Rule?
A rule reduces uncertainty by telling people what should happen under recurring conditions.
Situation → rule → interpretation → expected consequence → choice → action → enforcement or confirmation → feedback → revised expectation.
1. Rules Begin With a Coordination Problem
Rules usually exist because repeated behaviour must become more predictable.
Traffic rules help strangers use the same road. Examination rules protect comparability. School rules reduce uncertainty about attendance, conduct and safety. Financial rules define what institutions may do with other people’s money.
The first question should therefore be: What problem would become harder if this rule did not exist?
2. Clarity Determines Whether People Can Follow the Rule
A rule that is vague creates inconsistent interpretation.
“Behave properly” leaves many decisions unresolved. “During the test, no communication is allowed once papers are distributed” is more operational because the condition and prohibited action are visible.
Clarity reduces the gap between written rule and actual behaviour.
3. Legitimacy Affects Voluntary Compliance
People are more likely to follow rules voluntarily when the rule appears connected to a legitimate purpose and applied by an authority they regard as reasonably fair.
A safety rule can be strict and still legitimate if the risk is real and the burden proportionate. A minor arbitrary restriction can create more resistance because people cannot see the job it serves.
Legitimacy matters because systems cannot inspect every action continuously.
4. Rules Change Incentives
A rule changes what happens after a choice.
It may attach a penalty, reward, delay, requirement, approval step or reputational consequence to particular behaviour.
This does not mean every rule needs punishment. Some rules work because people value the shared order the rule creates. Others need credible enforcement because the private incentive to break them is strong.
5. Enforcement Makes the Rule Credible
If a rule is repeatedly broken without consequence, people update their model of the real system.
The written rule may say one thing while the operating rule becomes: “This is not really enforced.”
Enforcement does not need to be maximal. It needs to be credible, proportionate and sufficiently consistent for expectations to stabilise.
6. Consistency Prevents Status From Replacing the Rule
A rule applied differently to similar cases weakens predictability.
People begin to ask whether outcome depends less on the stated rule and more on who you are, who you know or who is enforcing it.
Perfect uniformity is neither possible nor always desirable, but similar cases should be treated similarly unless a relevant difference justifies the exception.
7. Exceptions Prevent Rules From Becoming Blind
General rules compress complexity. Real life then produces unusual cases.
Emergency medical needs, disability accommodations, force majeure, safeguarding concerns and genuinely unforeseen conditions may require exceptions.
A good exception process protects the purpose of the rule without turning every inconvenience into an exemption.
8. Procedures Translate Rules Into Action
A rule says what should happen. A procedure says how it happens.
Who reports the problem? Which evidence is required? Who decides? What time limit applies? Can the decision be appealed?
Without procedures, rules can become dependent on personality and improvisation.
9. Informal Rules Can Be Stronger Than Written Rules
Every group develops norms about what really happens.
A workplace may officially encourage dissent while informally punishing people who disagree with senior leaders. A classroom may officially welcome questions while socially rewarding silence.
The operating rule is revealed by repeated consequences, not only policy documents.
10. Power Shapes Which Rules Exist
Rules are not created in a vacuum.
The World Bank’s governance work emphasises that actors with unequal power interact within changing rules and that who participates, who is excluded and who can shape enforcement affects outcomes.
This means rule analysis should ask both: “What does the rule say?” and “Who had the power to shape the rule and its application?”
11. Rules Can Produce Gaming
When people know the rule, they adapt to it.
If the rule rewards a metric rather than the real outcome, people may optimise the metric. If a threshold matters, effort may cluster around the threshold. If compliance is measured only by paperwork, the paperwork may improve while reality does not.
Strong rule design anticipates adaptation rather than assuming the rule will be followed in exactly the spirit intended.
12. Rules Need Proportionality
The burden imposed by a rule should make sense relative to the risk or outcome it protects.
Using the same heavy procedure for trivial and catastrophic risks wastes attention and can reduce compliance. Rules become stronger when effort and enforcement scale with consequence.
13. Feedback Reveals Whether the Rule Works
A rule should be judged by the behaviour and outcomes it creates, not by how elegant it looks on paper.
Are incidents falling? Are people finding workarounds? Are delays increasing? Is the rule generating unfair burdens? Has a new technology made part of the rule obsolete?
Rules need observation after implementation.
14. Review Prevents Rule Fossilisation
Conditions change while old rules remain.
A rule that once protected a real risk can become unnecessary. A weak rule may need strengthening when new evidence appears. A new technology may create a risk the old rule never anticipated.
Strong institutions therefore keep a correction route: retain, revise, narrow, broaden or remove.
The Whole Rule Chain
Coordination problem → purpose → rule → clear scope → legitimate authority → procedure → expected consequence → behaviour → enforcement → observed outcome → review → retain, repair or remove.
A Useful Metaphor: Rules Are Lane Markings
Lane markings do not drive the car. They make many independent drivers more predictable to one another.
If markings are invisible, contradictory or ignored without consequence, coordination weakens. If every road uses excessive restrictions regardless of conditions, movement becomes unnecessarily difficult.
Rules at Three Zoom Levels
Micro: one choice
Does the person understand the rule and its consequence?
Meso: one organisation
Do written rules, informal norms, procedures and incentives produce the behaviour the organisation actually needs?
Macro: institutions and society
Do rules support commitment, coordination and cooperation while remaining legitimate and correctable?
How Rules Fail
- Ambiguity: people cannot tell what the rule requires.
- Illegitimacy: the rule appears arbitrary or disconnected from a defensible purpose.
- Selective enforcement: similar cases receive different treatment for irrelevant reasons.
- Paper-only rule: enforcement is so weak that the real operating rule is different.
- Gaming: people satisfy the measurement while defeating the purpose.
- Exception collapse: no path exists for unusual cases, or exceptions become so easy that the rule disappears.
- Fossilisation: the rule survives after the world that justified it has changed.
How Rules Are Repaired
Return to the original coordination problem. Rewrite the rule in observable language. Remove unnecessary burden. Clarify who decides and what evidence matters. Make exceptions explicit. Match enforcement to consequence. Measure the actual outcome after implementation.
If the rule no longer serves a real purpose, removal can be better than repair.
What Parents and Students Should Notice
- What problem is this rule trying to solve?
- Can the child explain the rule in plain language?
- Is the consequence predictable and proportionate?
- Are exceptions based on relevant differences?
- Does the rule build responsibility or merely obedience?
- What behaviour appears when adults are not watching?
- When evidence shows the rule is not working, can it change?
Principles, Rules, Standards, Policies and Procedures Are Different Layers
A principle expresses a protected reason or value. A rule specifies what is required or prohibited under defined conditions. A standard defines a level or criterion that performance should meet. A policy states an organisation’s chosen approach across recurring decisions. A procedure specifies the operational steps.
Confusing these layers creates brittle systems. A procedure can change while the underlying principle remains. A standard may stay fixed while several procedures are acceptable routes to meet it.
Principle explains why → rule defines boundary → standard defines acceptable level → policy selects approach → procedure defines execution.
Bright-Line Rules and Principle-Based Rules Solve Different Problems
A bright-line rule uses a relatively clear threshold: no entry after a specified time, a maximum load, a minimum age. It improves predictability and reduces discretion.
A principle-based rule asks decision-makers to apply a broader standard such as reasonableness, proportionality or acting in the receiver’s best interests. It adapts better to varied cases but demands stronger judgement and review.
The design choice is therefore partly a trade-off between predictability and context sensitivity.
Ex Ante and Ex Post Rules Act at Different Times
Ex ante rules constrain behaviour before harm occurs: licensing, safety requirements, access controls. Ex post rules impose liability, compensation or sanction after an event.
Ex ante controls can prevent high-consequence failures but may burden safe actors. Ex post systems preserve more freedom beforehand but may be too late when harm is irreversible.
High-risk domains often combine both.
Discretion Is Neither the Opposite of Rules Nor Automatically a Loophole
Rules cannot encode every relevant fact. Discretion lets an authorised person adapt the general framework to unusual conditions.
But discretion creates variability and can create bias or favouritism when criteria are hidden. Strong systems therefore bound discretion through purpose, relevant factors, documentation, review and appeal.
Good discretion is constrained judgement, not unrecorded personal preference.
Rule Hierarchies Resolve Conflicts Between Rules
Real organisations contain many rules. They can conflict unless authority and precedence are defined.
Constitutions, laws, regulations, contracts, organisational policies and local procedures often occupy different levels. A lower-level procedure should not silently override a higher-level obligation.
When rules conflict, the system needs a recognised method for determining which authority governs and who resolves ambiguity.
Rule Density Creates Compliance Load
Each rule may look reasonable in isolation while the total rule set becomes impossible to remember, interpret or administer consistently.
High rule density can produce checkbox behaviour, dependence on specialists and selective enforcement because nobody can monitor everything.
Rule quality therefore depends partly on the portfolio: overlap, contradiction, burden and whether lower-value rules consume attention needed for high-value ones.
Compliance Cost Is Part of the Rule’s Real Cost
A rule can achieve a worthwhile goal yet impose paperwork, waiting time, training, verification, technology and opportunity costs.
These costs should be compared with the risk or coordination benefit produced. If compliance cost rises while the protected outcome barely improves, redesign may be better than stronger enforcement.
Enforcement Probability Can Matter More Than Maximum Penalty
A severe penalty that is almost never applied may produce less compliance than a smaller, predictable consequence.
Credible enforcement depends on detection, evidence, decision capacity and procedural legitimacy. Increasing the headline penalty cannot compensate indefinitely for a system that rarely detects or processes violations.
Detection Rules Shape What Can Be Enforced
Rules depend on observability. Some violations are easy to detect automatically; others require judgement, reporting or investigation.
If detection focuses only on easy-to-measure behaviour, the enforcement system can become biased toward visible violations while serious but less observable harms receive less attention.
Measurement architecture is therefore part of rule architecture.
Due Process Protects the Rule From Its Own Enforcement Errors
Enforcement can be wrong. Evidence can be incomplete, identity mistaken, context misunderstood or authority misapplied.
Notice, an opportunity to respond, reasoned decisions, independent review and appeal are correction mechanisms where consequences are serious enough to justify them.
Due process is therefore not merely friction around a rule. It is a method for reducing false positives and arbitrary exercise of power.
Rules Can Create Loopholes Because Language Has Boundaries
A rule must define scope. Once scope is explicit, actors can search for behaviour that technically sits outside the wording while producing a similar effect.
This is not always bad faith; boundaries are unavoidable. But repeated loophole exploitation signals that the rule’s wording and underlying purpose have separated.
Repair may require broader principles, anti-avoidance provisions, revised definitions or a different measurement object.
Rules Can Cause Moral Hazard When Protection Removes Too Much Consequence
Rules sometimes protect people from harm, liability or failure. Protection may be necessary, but if actors no longer bear any consequence from risky behaviour, incentives can change.
The design problem is to preserve legitimate protection while keeping enough feedback for prudent behaviour and responsibility.
Sunset Clauses Force Some Rules to Re-Earn Their Existence
A sunset clause causes a rule or extraordinary power to expire unless it is deliberately renewed. This can be useful when evidence is uncertain, conditions are changing or emergency authority should not become permanent by inertia.
Sunsets are not appropriate for every stable rule. Their value is strongest where automatic continuation would hide whether the original justification still exists.
Grandfathering Trades Uniformity for Transition Stability
When a new rule applies only to future cases while existing arrangements continue under old terms, the system is grandfathering prior conditions.
This can reduce transition shock and protect reliance on previous rules, but it creates two classes operating under different regimes and can prolong outdated arrangements.
The design question is whether transition fairness and stability justify the temporary inconsistency.
Rule Changes Need Transition Architecture
A new rule can fail even when its final design is sound if people are not given enough notice, training, system support or time to adjust.
Effective dates, transition windows, safe harbours, pilot periods and staged enforcement can separate learning mistakes from deliberate non-compliance while the new regime is installed.
Rules Should Preserve Protected Values While Allowing Implementation to Evolve
A good rule system distinguishes what must remain invariant from what may change with technology or context.
For example, privacy, safety, fairness or examination integrity may be protected values. The procedure used to achieve them can evolve as new tools appear.
This prevents institutions from confusing loyalty to purpose with loyalty to one historical implementation.
Rule of Law Is a System Property Beyond Having Many Rules
A society can have numerous written rules without those rules reliably constraining power. Rule-of-law properties include publicity, relative stability, prospective application, consistent administration, lawful authority and meaningful routes for challenge.
The deeper test is whether people can reasonably predict how public power will operate and whether the powerful are themselves constrained by the legal system.
A High-Resolution Rule Audit
- Purpose: What recurring coordination or protection problem does the rule solve?
- Layer: Is this a principle, rule, standard, policy or procedure?
- Authority: Who legitimately created it?
- Hierarchy: Which higher-level obligation constrains it?
- Form: Is a bright-line boundary or principle-based standard more appropriate?
- Timing: Does the system need ex ante prevention, ex post liability or both?
- Scope: Which situations and actors are covered?
- Clarity: Can ordinary users understand what is required?
- Discretion: Which judgement is permitted and how is it bounded?
- Detection: How will violations become visible?
- Evidence: What proof is needed before consequence follows?
- Enforcement: Is consequence credible, proportionate and consistent?
- Due process: How are enforcement errors corrected?
- Exceptions: Which relevant differences justify departure from the general rule?
- Loopholes: Can actors defeat the purpose while satisfying the wording?
- Compliance cost: What time, money and attention does obedience consume?
- Rule density: Does the wider rule portfolio remain comprehensible?
- Incentive effect: What behaviour changes because the rule exists?
- Moral hazard: Does protection erase too much consequence feedback?
- Transition: What happens to existing arrangements when the rule changes?
- Review: Is there a sunset, review date or correction route where appropriate?
- Protected value: Which purpose must remain even if implementation evolves?
- World return: Did the rule improve real behaviour and receiver outcomes enough to justify its burden?
Connect Rules to the Wider eduKateSG Mechanism Estate
- How Power Works — who creates, applies and remains constrained by rules.
- How Incentives Work — how enforcement probabilities, thresholds and consequences change behaviour.
- How Change Works — how new rules migrate from announcement into stable practice.
- How Trust Works — why consistency, fairness and predictability reduce the need for constant monitoring.
- How Responsibility Works — how obligations, authority and answerability should remain aligned.
Causal Gateway Handoff
- How the World Works — place rules inside the wider causal map.
- How Countries Work — see rules embedded in territorial, institutional and administrative systems.
- How Borders Work — follow rules into permissions, crossing procedures, customs and human consequences.
- How International Governance Works — follow shared rules into treaties, institutions and implementation across states.
- How Commercial Systems Work — see rules structure promises, orders, acceptance, payment and dispute resolution.
Continue Through eduKateSG
Evidence and Further Reading
The World Bank’s World Development Report 2017: Governance and the Law explains how actors with unequal power interact within changing rules and why policy effectiveness depends on credible commitment, coordination and cooperation rather than formal rules alone.
Frequently Asked Questions
Why do people break rules?
Reasons include weak enforcement, unclear wording, low legitimacy, stronger competing incentives, lack of capability, social norms or deliberate opportunism. The reason matters because different failures need different repairs.
Are stricter rules always better?
No. Stronger restrictions can improve safety or coordination when the risk justifies them, but excessive burden can create avoidance, gaming and unnecessary loss of agency.
What is the difference between a rule and a norm?
A rule is usually more explicit and tied to a recognised authority or procedure. A norm is a socially shared expectation that can operate informally through approval, disapproval and belonging.
Final compression: Rules work when they make expectations legible, consequences credible and coordination easier—while remaining proportionate, consistently applied and open to correction when the world or the evidence changes.