Skip to content

Eligibility Threshold Review

Access-and-allocation review — instantiates Adaptive Threshold Recalibration

Re-examines a cutoff that decides who is in or out of a benefit, service, or protection, so the line still serves its purpose and treats groups fairly.

Eligibility Threshold Review reassesses a cutoff whose crossings fall on people: who receives a benefit, a service, an admission, a subsidy, a protection. That is what makes it distinct from its detection- and load-focused siblings — the central question is not "is the number accurate?" but "does this line still include the people it was meant to serve, and is the exclusion it produces fair?" Because the boundary allocates access, its two governing concerns are purpose (whom the line is for) and equity (who ends up unevenly on the wrong side of it). It reads the threshold as an act of inclusion and exclusion, and asks whether that act still lands where it should.

Example

A university sets need-based aid at a household-income cutoff and leaves it fixed for six years. Regional housing and childcare costs climb steeply over that period; families whose real disposable income hasn't budged now sit just above the frozen line and lose all aid — a hard cliff where a dollar of income costs thousands in support. Eligibility Threshold Review restates the purpose (enable students who otherwise couldn't enroll to attend), measures who now falls just inside and outside the line, and checks the impact by region and family structure. It finds the cliff is concentrated among students from high-cost areas. The remedy weighs the harm of wrongly excluding a needy family against the cost of admitting a marginal one, and replaces the cliff with an indexed, graduated phase-out.[1]

How it works

The review runs on three moves that a purely statistical threshold review skips. First it re-anchors the cutoff to its purpose — the outcome and population it exists to serve — so the debate isn't just "what number?" Second it runs a subgroup analysis of who is included and excluded now, because an eligibility line can drift into quietly excluding the very group it targeted. Third it applies an explicit consequence weighting: the harm of a wrongful denial is not symmetric with the cost of a wrongful grant, and the review names that asymmetry rather than optimizing a neutral error rate.

Tuning parameters

  • Cutoff level — where the line sits; the direct lever on how many are included, weighed against budget.
  • Cliff vs phase-out — a hard boundary or a graduated taper; tapering removes the notch but complicates administration.
  • Indexation — whether the cutoff auto-adjusts to inflation or cost-of-living so it doesn't silently drift.
  • Exception / appeal channel — a route for borderline cases the fixed rule mishandles.
  • Subgroup protections — explicit floors or adjustments to prevent uneven exclusion of a target group.

When it helps, and when it misleads

Its strength is catching the slow drift by which a frozen eligibility line stops serving its intended population, and surfacing the equity impact that an average-performance view hides. Its classic misuse is raising the bar to hit a budget target and presenting it as "tightening program integrity" — a purpose-free move dressed as a purpose-driven one. The discipline that guards against it is to force every proposed change back through the purpose statement and the subgroup review: a change that can't be justified by whom the line is meant to serve, and that worsens outcomes for that group, is a budget cut wearing a threshold's clothes.

How it implements the components

  • threshold_purpose_statement — it re-anchors the cutoff to the population and outcome it exists to serve, keeping the change from being a bare number-shift.
  • fairness_and_subgroup_review — it examines who is unevenly included or excluded across groups, sites, and case types.
  • consequence_weighting_policy — it names the asymmetry between wrongful denial and wrongful grant and weighs them explicitly.

It does not ratify the change through a governance body (stakeholder_review_panel) — Policy Threshold Update does — nor does it stage the rollout (rollout_and_guardrail_plan), which Staged Threshold Rollout owns.

References

[1] A benefit cliff (notch) occurs when crossing an eligibility cutoff by a small amount causes an abrupt, disproportionate loss of benefit, creating perverse incentives near the line. Indexation and graduated phase-outs are the standard structural fixes. ↩