Week 8 Session: Mapping a priority rule

Reading one real, published priority policy for its class, its criterion, and what it actually changes — without calculating whether it's fair.

Week 7 asked whether you could verify a queue’s rule. This week the rule is already visible and stated outright — the task is to read exactly what it changes, and to be honest about what it doesn’t tell you.

Before the session#

Read the Week 8 lecture first. Keep its distinction in mind: a priority rule’s criterion — what earns it — is separate from what it actually changes for the people who don’t have it.

Finding a policy#

Find one real, publicly published priority policy — stated outright, in writing, by a real organisation: a supermarket express-lane sign, an airline’s boarding-priority tiers, a retailer’s loyalty checkout tier, a call centre’s stated priority category, or a government service’s priority-processing category. Don’t invent one. Don’t use a medical, emergency, clinical, or triage setting.

Building the priority map#

Answer five questions, in order, from the policy’s own published wording wherever you can:

1. Who is the priority class? Quote or closely paraphrase how the policy itself describes who qualifies.
2. What exact criterion creates that class? Something stated and checkable — fewer than a stated number of items, existing account, membership tier — not your guess at the organisation’s reason for it.
3. What does the rule change about who moves first? Describe the mechanical change to ordering: who now goes ahead of whom, and under what circumstance.
4. Who remains under the ordinary rule? State plainly what governs everyone not in the priority class.
5. What information would we need to evaluate the resulting redistribution? Name the specific missing information — each class’s typical service time, class sizes, arrival rates — that a formal analysis like Raz et al.’s would require before anyone could say whether this redistribution is even favourable by that measure. This is a request to name what’s missing, not to estimate it.

Do not calculate anything. No fairness score, no RAQFM, no verdict on whether the policy is mathematically fair — that data doesn’t exist for a publicly posted policy, and this exercise isn’t asking you to invent it.

The closing distinction#

End your entry with two separately labelled statements:

• What this policy tells us — items 1–4, read directly off the published policy.
• What we’d need more information to establish — item 5, plus anything about how people in either class actually experience the wait, which no published policy states and this exercise cannot determine.

Keep these visibly separate. Don’t let one bleed into the other.

What’s out of bounds#

Don’t decide whether the policy is fair. Don’t redesign it. Don’t render a moral verdict. Don’t claim the policy is or isn’t “mathematically fair” or “biased” — you don’t have the data, and neither does anyone reading a public sign.

In the session#

Bring your priority map. We compare across the room: whose criterion was crisp and checkable versus vague, whose “ordinary rule” was itself hard to state, and where everyone’s missing-information list converged.

What this can and can’t show#

One policy’s public wording can tell you what a system says it’s doing and roughly what it mechanically changes. It cannot tell you whether that change is fair, efficient, or resented — those need data or evidence this exercise doesn’t have. That gap is the point, not a flaw in the exercise.

Afterwards#

Keep the log running. Week 9 asks what the whole waiting system is actually exchanging for whatever waiting it reduces.