Skip to content

Rebound-Leakage Boundary Review

Boundary-sensitivity method — instantiates Rebound-Aware Efficiency Governance

Re-runs the efficiency outcome at successively wider category, supply-chain, geographic, and time boundaries to expose rebound that was merely exported or delayed past the original accounting line.

A saving that looks real inside a tightly drawn box may only have pushed the resource use across the box's edge — to another product category, another country, or another year. Rebound-Leakage Boundary Review attacks that edge directly. Rather than refine the estimate inside the accounting boundary, it redraws the boundary outward — wider category, wider supply chain, wider geography, longer time horizon — and re-computes the outcome at each frame. Its defining insight is that the gap between the narrow result and the wide result is the leaked rebound, the part that was exported or delayed rather than eliminated. It is a sensitivity analysis on where you drew the line.

Example

A country runs an energy-efficiency policy for heavy industry. Inside the national books it works: energy per tonne of steel falls, and total industrial energy drops an illustrative ~12%. The Rebound-Leakage Boundary Review re-runs that outcome at wider frames. Widen the geographic boundary and some production has simply relocated to less-efficient plants abroad — the tonnes, and their energy, reappear as imports (the carbon-leakage pattern[1]). Widen the supply-chain boundary and the embodied energy of those imports was never on the domestic ledger. Widen the time boundary and mothballed domestic capacity is poised to restart in the next upcycle. Re-computed at the widest defensible frame, the crisp 12% domestic drop shrinks toward ~3% globally, and the difference is logged in a leakage register as exported rebound. The policy's headline saving is corrected for the resource use that merely moved.

How it works

The method's distinguishing procedure is to hold the intervention fixed and move the boundary, one dimension at a time — category, supply chain, geography, time — re-running the same outcome calculation at each widening. Each expansion can reveal resource use that the narrower frame excluded by construction; the difference between narrow and wide is the exported or delayed rebound, which it records rather than lets vanish. It does not trace the causal channels of rebound (an audit does that) or attach prices to what it finds; it establishes where the boundary should honestly sit and how much the answer depends on that choice.

Tuning parameters

  • Which boundaries to widen — category, supply chain, geography, or time; each dimension exposes a different kind of leak.
  • How far to widen — global and all-time is the most honest frame but the least tractable to compute.
  • Attribution strictness — how confidently displaced activity is assigned to this intervention versus background trends.
  • Materiality cutoff — how large a leak must be before it is worth registering.

When it helps, and when it misleads

Its strength is catching the flattering result that exists only because the boundary was conveniently narrow — the single most common way rebound hides — with carbon leakage as the canonical worked case.[1] Its failure mode is that widening the frame adds noise and weakens attribution the further out it goes, so a very wide boundary can be honest and unusable at once. The classic misuse is to keep the boundary narrow precisely because widening it would spoil the number. The discipline that guards against it is to predeclare the widest defensible boundary up front and to report the narrow and wide results side by side, so the reader can see how much of the saving depends on where the line was drawn.

How it implements the components

Rebound-Leakage Boundary Review fills the archetype's boundary and exported-cost components — the sensitivity of the answer to the accounting frame:

  • boundary_leakage_sensitivity_check — its core procedure: re-running the outcome across widening boundaries and reading the leak off the gap between frames.
  • externality_register — the log of exported and delayed resource use the widening reveals, recorded so it is not silently dropped.

It does not trace the within-frame rebound channels (that is the Direct and Indirect Rebound Audit) or price the externalities it logs (that is the Full-Cost Accounting sibling) — it establishes where the boundary should sit before either of those runs.

Notes

A leak found here is not double-counted with the audit's within-frame rebound — it is precisely the portion the audit's boundary excluded. The two are complementary: the audit traces where rebound went inside the box, and this review finds what slipped outside it. A saving that survives both is one you can actually trust.

References

[1] Carbon leakage: the standard climate-policy case in which emissions cut in one jurisdiction are offset by increases elsewhere — often because production relocates — so the global total barely moves. It is used here as the archetype's canonical instance of rebound exported across an accounting boundary. ↩