Skip to content

Rebound Scenario Stress Test

Scenario model — instantiates Rebound-Aware Efficiency Governance

Runs the efficiency intervention through a spread of rebound scenarios — from negligible to full backfire — before scaling, to see whether the intended saving survives the bad cases.

Before an efficiency intervention is committed at scale, the honest question is not "how much will it save?" but "how much might it not save if people respond badly?" Rebound Scenario Stress Test answers that by refusing a single rebound number and instead spanning a spread of scenarios — low, central, high, delayed, and structural backfire — and mapping how the response could unfold in each. Its defining move is to treat rebound magnitude as a variable to be ranged, not a point estimate, and to check whether the intended saving still survives the pessimistic cases. It is ex-ante and model-based: a pre-commitment gate, not a measurement of anything that has happened yet.

Example

A government is about to set a national minimum-efficiency standard for air conditioners in a hot, fast-growing region; the engineering says ~35% less electricity per unit. The Rebound Scenario Stress Test declines to bank that figure and spans the response instead. In the low scenario people simply pocket the saving. In the central scenario cheaper cooling nudges thermostats down, taking back an illustrative ~20%. In the high scenario the cheaper running cost accelerates AC ownership itself — the induced-adoption path. In the delayed scenario rebound only appears in year three as incomes rise. In the structural backfire scenario cheap cooling reshapes construction toward glass towers that demand still more cooling, and total use climbs above the pre-standard baseline. Mapping each response path to a total-use outcome, the test reports that the standard nets ~15% under central assumptions but backfires under the structural one — and recommends pairing it with a demand guardrail before national rollout, rather than discovering the backfire in the field.

How it works

The test's distinguishing structure is the deliberate spread. It defines scenario axes — behavioral elasticity, adoption response, timing, and structural feedback — and for each combination traces a response path from the efficiency gain through behavior to a total-use outcome. It then compares the whole spread against the intended saving and the tolerance band, and surfaces the specific scenario that would break the plan, so that scenario can be guarded against rather than gambled on. Because it is ex-ante, it runs on plausible behavior ranges rather than measured ones — the mirror image of the ex-post audit.

Tuning parameters

  • Scenario span — how pessimistic the worst case is, and whether outright backfire is included. A wider span catches more but can paralyze a decision.
  • Elasticity range — the width of the behavior-estimate band fed in. Wider is more honest but less decisive.
  • Structural feedbacks — whether to include second-order paths like induced adoption and redesign. Realistic, but the further out, the more speculative.
  • Decision rule — whether the intervention must clear the central case or the high case to proceed — the conservatism dial.

When it helps, and when it misleads

Its strength is revealing backfire[1] before a costly commitment at scale, and naming the exact scenario that would defeat the plan so it can be pre-empted with a guardrail. Its failure modes are the familiar ones for scenario work: the set is only as good as the imagination behind it, and false precision can attach to invented probabilities dressed as analysis. The classic misuse is to pick a flattering central case and wave the project through on it. The discipline that keeps it honest is to predeclare the scenario axes, carry the pessimistic case explicitly into the decision, and update the ranges from real data once the audit starts reporting.

How it implements the components

Rebound Scenario Stress Test fills the archetype's behavioral-projection components — the ranged estimate and the mapped paths, before commitment:

  • elasticity_and_behavior_estimate — it takes the behavioral response as a range spanning low to backfire, rather than a single elasticity, and stress-tests across it.
  • response_path_map — for each scenario it traces how the response propagates from the efficiency gain through behavior and investment to a total-use outcome.

It does not measure the actual elasticity (that is the Elasticity Experiment) or the realized rebound after the fact (that is the Direct and Indirect Rebound Audit) — it stress-ranges their inputs before any commitment is made.

Notes

The stress test is a pre-commitment gate, not a monitor. Once the intervention is live, the ex-post audit and the recalibration loop take over from its scenarios — the test's job was to make sure the plan could survive a bad response before the response was real, and to hand forward the one scenario worth watching for.

References

[1] Backfire: the extreme case of the rebound effect in which take-back exceeds 100%, so an efficiency improvement raises total resource use above its pre-improvement level. It is the scenario a stress test exists to catch before scaling, because it inverts the intervention's entire purpose. ↩