lev-os Schelling Fence
Overview
Strategic precommitment to defend arbitrary boundaries on slippery slopes: Establish a conspicuous, defensible line that takes on significance as a coordination point, preventing gradual erosion toward undesirable outcomes.
Core insight: When facing slippery slopes, the strongest defense isn't always at the most justifiable position—it's at the most obvious, unambiguous line where everyone can coordinate to hold the boundary.
Source: Scott Alexander (Slate Star Codex), 2012, building on Thomas Schelling's concept of focal points
The Framework
The Slippery Slope Problem
You face a continuum from acceptable to unacceptable:
- No restrictions → reasonable restrictions → excessive restrictions → totalitarian control
- Never lie → white lies → significant lies → pathological lying
- Staying up reasonable hours → slightly late → very late → all-nighter → ruined sleep schedule
Challenge: Each marginal step seems small and justifiable in isolation, but the cumulative effect is disastrous.
The Schelling Point Solution
Schelling point: A salient, obvious boundary that serves as a natural coordination point even without communication.
Schelling fence: A deliberately established Schelling point used as a precommitment device to resist incremental erosion.
Key properties:
- Conspicuous: Easy to recognize, hard to mistake
- Binary: Clear which side of line you're on
- Symbolic value: Takes on significance beyond inherent properties
- Coordination: Multiple parties/your future self can align on defending it
- Credible commitment: Failing to hold this line means all positions up to next Schelling point are vulnerable
Why Arbitrary Lines Work
Not defending the optimal position—defending the most defensible position:
- "No alcohol before 5pm" is more defensible than "moderate drinking only" (what's moderate?)
- "Zero tolerance for lying in engineering specs" is more defensible than "don't exaggerate too much"
- "Bedtime at 11pm always" is more defensible than "get enough sleep"
The first example in each pair is more arbitrary but clearer—that clarity makes it holdable.
When to Use
Personal commitment devices:
- Habit formation (gym 5x/week, not "regularly")
- Avoiding temptation (no social media on phone, not "use less")
- Sleep/work boundaries (hard stop at 6pm, not "reasonable hours")
Interpersonal boundaries:
- Professional ethics (never accept gifts from vendors, not "small gifts okay")
- Relationship limits (certain topics off-limits, not "be respectful")
- Management decisions (all expenses require approval, not "use good judgment")
Policy and governance:
- Constitutional rights (absolute free speech, not "reasonable restrictions")
- Safety regulations (zero tolerance, not "minimize incidents")
- Institutional norms (mandatory recusal with conflicts, not "be aware of bias")
Slippery slope prevention:
- Whenever incremental movement toward bad outcome is easy
- When future you will face temptation to rationalize "just this once"
- When multiple parties need to coordinate on a boundary
- When erosion happens gradually and imperceptibly
Implementation Steps
1. Identify the Slippery Slope
What's the undesirable end state you're trying to avoid?
- Complete sleep deprivation
- Totally compromised professional ethics
- Feature bloat that kills product simplicity
- Technical debt that makes codebase unmaintainable
Map the continuum from current state to disaster state.
2. Find Natural Schelling Points
Look for conspicuous boundaries along the continuum:
Good Schelling points:
- Round numbers (bedtime at 11pm, not 10:47pm)
- Categorical distinctions (weekdays vs. weekends)
- Bright lines (zero instances, not "very rare")
- Existing norms/traditions (industry standards)
- Simple rules (always/never, not "usually")
Bad Schelling points:
- Vague thresholds ("reasonable" "moderate" "appropriate")
- Subjective judgment calls (case-by-case basis)
- Complex conditions (if X and Y but not Z...)
- Invisible boundaries (hard to verify compliance)
3. Choose Conservative Fence Location
Critical insight: Place the fence before the truly problematic region.
Why: If you try to defend right at the edge of disaster, small violations are catastrophic. If you defend early, small violations are just small violations.
Example:
- Too close: "No drinking and driving" (violation = immediate danger)
- Better: "No drinking within 8 hours of driving" (violation = buffer remains)
- Schelling fence: "Never drink on nights before early meetings" (simple, defensive)
Build in margin for error.
4. Establish Symbolic Significance
Make the boundary mean something beyond its literal position:
Explicit precommitment: "I'm setting this boundary and I will defend it"
Public commitment: Tell others—social cost of violation reinforces boundary
Link to values: "This represents my commitment to X principle"
Create ritual: Make crossing the boundary require deliberate ceremony/justification
The boundary's power comes from its significance, not just its position.
5. Defend Vigorously
Zero tolerance for violations: Every exception weakens the fence.
Resist "just this once": If it's justified this time, it'll be justified next time.
Escalate defense: Treat boundary violations as more serious than the marginal harm—you're defending the fence itself, not just the immediate issue.
Make exceptions explicit: If you genuinely need to move the fence, do it deliberately and establish a new one—don't just let it erode.
6. Repair Breaches Immediately
When the fence is violated:
- Acknowledge clearly: Don't pretend it didn't happen
- Re-establish boundary: Return to fence position, don't stay at violation point
- Strengthen if needed: Add accountability or make boundary more conspicuous
- Learn: What pressure led to breach? Reinforce those defenses
Every violation you tolerate becomes precedent for the next.
Common Pitfalls
Defending indefensible positions: Putting the fence at an awkward, hard-to-explain place. Pick obvious boundaries even if they're not theoretically optimal.
Too many fences: If you have 20 bright-line rules, none are bright. Choose the critical boundaries.
Justifying exceptions: "This situation is special" is how fences erode. Exceptions should be extremely rare and involve moving the fence deliberately, not ignoring it.
Arguing the fence position: Getting into debates about whether the specific line is optimal. The fence's value is in its defense, not its specific location—don't undermine it by admitting it's arbitrary.
Not making boundaries visible: A fence you can't see or verify compliance with doesn't work. Make violations obvious.
Gradual movement: Slowly shifting the fence 5 minutes at a time. Defeats the purpose—fence should stay put or move in deliberate, discrete jumps.
Real-World Applications
Product development: "No features that require >2 clicks" prevents gradual UX complexity creep. Each individual extra click seems fine, but the fence holds simplicity.
Engineering standards: "All code requires tests" prevents gradual erosion to "most code has tests" to "important code has tests" to "tests are optional."
Personal productivity: "No phone use before 9am" prevents checking "just one thing" spiral. Arbitrary but defensible.
Free speech policy: "No content moderation based on viewpoint" is clearer than "moderate extreme content"—the fence prevents incremental expansion of "extreme."
Commitment devices: "Zero alcohol during work week" easier to maintain than "moderate drinking"—no decisions, just a rule.
Behavior change: "Gym every weekday morning" beats "exercise regularly"—the fence is conspicuous, violations are obvious.
Power Moves
Pre-establish fences before temptation: Set bedtime policy when well-rested, not when tired at midnight. Set expense rules when solvent, not when vendor offers "special deal."
Use existing Schelling points: Don't create arbitrary boundaries from scratch—anchor to existing round numbers, natural categories, social norms. Leverages existing salience.
Make fence public: Social pressure reinforces personal commitment. "I don't drink on school nights" is easier to maintain when friends know the rule.
Celebrate fence defense: Each time you successfully hold the boundary in face of temptation, reinforces its significance. "I wanted to stay up, but I held the 11pm line."
Create friction at fence: Add verification, reporting, or ceremony required to cross. Makes violations deliberate, not accidental.
Multiple parties share fence: Team agreement on boundary makes defense easier—"we don't ship without tests" is stronger than individual commitment.
Advanced Technique: Fence Maintenance
When to move a fence:
- New information shows fence is in wrong place (too conservative or too permissive)
- Context change makes old fence irrelevant
- Successfully holding fence for long period allows moving closer to optimal
How to move a fence without eroding it:
- Make movement explicit, deliberate decision
- Establish new fence at specific new position
- Communicate clearly: "Old rule was X, new rule is Y as of [date]"
- Defend new fence as vigorously as old one
- Don't move incrementally—jump to new stable position
Never: Just ignore the fence when inconvenient and hope no one notices. That destroys the fence's credibility.
Related Frameworks
- Commitment Devices: Schelling fences are a type of commitment device using salience
- Sorites Paradox: The philosophical problem Schelling fences solve (heap of sand: when does removing grains stop being a heap?)
- Precommitment Strategies: Ulysses tying himself to mast—limiting future options
- Trigger-Action Plans: "When X, then Y" rules similar to fence-defense protocols
- Bright-Line Rules: Legal/ethical boundaries that are unambiguous
- Chesterton's Fence: Before removing a boundary, understand why it was erected
Why "Fence" Metaphor Works
Physical fence properties:
- Visible: You know when you're crossing it
- Binary: You're on one side or the other
- Defended: Crossing requires effort/intention
- Symbolic: Represents property boundary beyond just physical barrier
Schelling fence inherits these: The rule isn't just a guideline—it's a defended boundary with significance beyond its specific position.
Sources
- LessWrong - Schelling Fences on Slippery Slopes (Scott Alexander, 2012)
- Beeminder Blog - Schelling Fences
- Thomas Schelling - The Strategy of Conflict (original focal point concept)