Manu's Vault

Work-the-Policy-Not-the-Exception

4 min readStatus:Seedling

Work the policy, not the exception is Will Larson’s principle that engineering managers should respond to exception requests by improving the underlying policy rather than granting individual workarounds. The core insight: if an exception seems reasonable, that’s evidence the policy needs updating — not that this person deserves special treatment.

The Core Principle

When someone asks for an exception to a process or rule, the manager faces a choice:

  • Grant the exception — fast, helps this person now, but creates a precedent
  • Work the policy — slower, but fixes the root cause and helps everyone systematically

Larson’s guidance: default to working the policy.

Why Exceptions Are Harmful

  • Precedent and scaling: Every exception is an implicit commitment to grant it to anyone in similar circumstances. With 10 engineers, informal negotiation is manageable; with 100, it collapses.
  • Manager-as-bottleneck: Exception-granting concentrates decision power in the manager. People learn to advocate with the manager rather than follow the system — creating a dependency loop that doesn’t scale.
  • Systematic unfairness: Only those who ask (or ask well) receive exceptions. Quieter team members, newer joiners, and those less comfortable self-advocating systematically receive less — even when they deserve equal treatment.
  • Exception debt: Accumulated exceptions form an invisible parallel policy — inconsistent, informal, and resentment-generating. Each exception that isn’t codified creates a ghost rule that is felt but never articulated.
  • Cognitive load: Each exception requires re-litigating context, tracking precedent, and defending consistency — all costs that compound with team size.

When Exceptions Are Appropriate

  • Genuine emergencies that no policy could anticipate — unique, one-time circumstances
  • Discovery mechanism: grant the exception AND immediately update the policy to reflect the new case
  • Explicit one-time events — name them as non-repeating when granting, and track them

Four-Step Response Framework

  1. Ask: Is this request exposing a gap in the policy?
  2. If yes: Fix the policy first, then apply the updated policy to this person
  3. If no: Grant explicitly as a named, one-time exception
  4. Track: If you grant the same exception twice, that’s a policy gap — update the policy

Procedural Justice Angle

Research on procedural justice (Thibaut & Walker, 1975; Tyler, 1990) shows that people experience consistent rule application as fair — even when the outcome is unfavorable — because the process signals respect and impartiality. Discretionary exceptions undermine this: they create a perception that outcomes depend on who you know and how assertively you advocate, not on merit or circumstance. Policy consistency is not just operationally efficient — it is perceived as more just.

Connection to Systems Thinking

The principle applies Deming’s distinction between common cause variation (systemic issues requiring policy change) and special cause variation (unique circumstances justifying individual response). Exception-granting treats every case as special cause; working the policy diagnoses whether the variation is actually systemic. See Systems-Thinking-Stocks-Flows-Feedback.

Sources

  • Larson, Will (2019). An Elegant Puzzle: Systems of Engineering Management. Stripe Press. ISBN: 978-1-7322651-8-9. Chapters 3.3 and 4.1.

    • Original articulation of the “work the policy” principle and its connection to exception debt.

  • Thibaut, John W. and Walker, Laurens (1975). Procedural Justice: A Psychological Analysis. Lawrence Erlbaum Associates. ISBN: 0-470-85803-6.

    • Foundational research establishing that perceived fairness depends as much on process consistency as on outcome; uniform rule application perceived as legitimate.

  • Tyler, Tom R. (1990). Why People Obey the Law. Yale University Press. ISBN: 978-0-300-04820-7.

    • Demonstrates that procedural fairness (consistent, unbiased process) drives voluntary compliance and institutional trust more than favorable outcomes.

  • Deming, W. Edwards (1986). Out of the Crisis. MIT Press. ISBN: 978-0-262-54115-7.

    • Distinguishes common cause variation (systemic, requiring structural change) from special cause variation (unique circumstances); argues treating common causes as special causes worsens the system.

  • Kahneman, Daniel, Rosenfield, Andrew M., Gandhi, Linnea, and Blaser, Tom (2016). “Noise: How to Overcome the High, Hidden Cost of Inconsistent Decision Making.” Harvard Business Review, October 2016.

    • Research on “noise” in human discretionary judgment: the same expert makes different decisions about identical cases at different times, demonstrating the unreliability of exception-granting over systematic policy.

Note

This content was drafted with assistance from AI tools for research, organization, and initial content generation. All final content has been reviewed, fact-checked, and edited by the author to ensure accuracy and alignment with the author’s intentions and perspective.

Graph View

Backlinks