Skip to content

Performance-Linked Renewal

Funding-renewal policy — instantiates Revenue–Accountability Coupling Design

Makes the renewal of an agent's funding or mandate conditional on demonstrated performance for the constituency it serves, so resource survival tracks answerable results rather than flowing automatically.

An agent drifts when its money arrives on a schedule that has nothing to do with whether it serves anyone. Performance-Linked Renewal is the rule that puts the revenue channel itself on the hook: the continuation of funding or mandate is gated on verified results for the served constituency, so that not performing costs the agent its resources. It is the renewal gate — the single point where "we have always funded them" is replaced by "we fund them because it demonstrably works for the people it is meant to work for." Its defining move is to make the default flow conditional: money that used to renew automatically now stops unless answerable performance is shown, which re-couples resource survival to responsibility. It says nothing about who the constituency is or how performance is measured — it consumes those — and everything about what happens to the funding when the results are in.

Example

An international development funder has for years renewed a service contractor's rural-clinic grant on the basis of spend and activity — money moved, buildings staffed. Performance-Linked Renewal rewrites the renewal terms: the next multi-year tranche is released only if independently verified outcomes clear a pre-set bar — clinics still operating and utilized ≈12 months on, measured with the served communities' own data rather than the contractor's reporting. A contractor that clears the bar renews. One that misses enters a defined correction window — an improvement plan with a shortened, conditional extension — and, if it still misses, funding is not renewed and service is transitioned to another provider under a wind-down plan.

The shift is not that the contractor is punished; it is that the villagers, who could never withhold a distant funder's grant, now sit at the hinge of the renewal decision. The contractor's revenue answers, at last, to whether the clinics work — not to the disbursement calendar.

How it works

  • State the renewal condition in the constituency's terms. The bar is written in outcomes the served actually experience, not activity or spend, and is agreed before the period so it can't be renegotiated under incumbency pressure.
  • Verify independently. Performance is judged on evidence the agent does not control — audit findings, drift indicators, the served constituency's own data — so the gate rests on facts, not self-report.
  • Graduate the response. Renew / renew-with-conditions / correction window / non-renewal-and-transition — a ladder, so a single noisy miss doesn't destroy good work and a persistent one still ends in exit.
  • Make the exit real. A credible wind-down and handover is part of the rule; without it the threat of non-renewal is a bluff and the coupling is nominal.

Tuning parameters

  • Threshold stringency — how demanding the renewal bar is. Set too low it becomes a rubber stamp; too high it churns capable providers on noise.
  • Outcome-vs-output basis — how far down the value chain the criteria sit. Outcomes align with the constituency but lag and are hard to attribute; outputs verify fast but are easier to game.
  • Correction-window length — how long a failing agent gets to recover before non-renewal. Longer absorbs noise and transition cost; shorter sharpens the incentive.
  • Renewal granularity — binary renew/cut versus funding that scales continuously with performance. Continuous is fairer but muddies the survival signal.
  • Term length — how long each renewal lasts. Long terms give stability but weaken the coupling by gating less often.

When it helps, and when it misleads

Its strength is that it turns "funded because we always have" into "funded because it works for whom it's meant to," and hands a bypassed constituency real leverage: the renewal gate is a lever they otherwise never hold over an automatically-funded agent. It blunts entrenchment and gives underperformance a defined, non-catastrophic path to correction or exit.

Its central failure mode is measure fixation: once survival hinges on a metric, effort migrates to the metric rather than the mission — teaching to the test, and cream-skimming the easy cases to hit targets while the hardest-to-serve are quietly dropped.[1] It can also drive short-termism (starving long-payoff work that won't score by the next gate) and, if thresholds churn good providers on noisy data, real instability and transition cost. The classic misuse is running the linkage backwards — setting or waiving the threshold after results are known so the incumbent renews regardless, which restores exactly the automatic flow the mechanism was meant to break. The discipline that keeps it honest is to pre-commit the criteria, pair the headline metric with counter-metrics that catch cream-skimming, verify independently, and keep a correction window so noise doesn't terminate good work.

How it implements the components

Performance-Linked Renewal realizes the renewal-and-exit side of the archetype's machinery — the components that decide whether resources keep flowing:

  • resource_renewal_condition — it is the condition: continuation of funding or mandate is gated on verified performance for the constituency.
  • correction_and_exit_path — the graduated response (correction window → non-renewal → managed transition) is the built-in route from underperformance to a real, orderly exit.

It does not map who the constituency is (principal_constituency_map — that's the Budget-to-Mandate Crosswalk) or generate the performance and drift evidence it consumes (drift_indicator_set / revenue findings — Rent-Stream Drift Dashboard, Revenue Source Audit); it acts on those signals rather than producing them.

Notes

Do not confuse this with the Sunset Reauthorization Vote. Both govern continuation, but they differ at the default: Performance-Linked Renewal defaults to continue-if-performing and gates on measured outcomes; the sunset defaults to expire-regardless and forces a political re-decision on a clock. They compose cleanly — the sunset forces the question, performance-linkage supplies the evidence to answer it.

References

[1] Goodhart's law — "when a measure becomes a target, it ceases to be a good measure." Tying survival to a metric is precisely the condition under which the law bites, which is why the discipline pairs the target with counter-metrics. ↩