Rebound Effect

Definition

The rebound effect is the micro-mechanism beneath Jevons Paradox: saved resources do not quietly disappear into nature or a warehouse; they turn into capital that finances the next gear of consumption. Per-unit cost falls, new uses open, populations that did not previously demand the resource are pulled in, and total consumption overruns the saving that efficiency produced.

Why it matters

If the rebound term is left out of a projection, efficiency looks like a policy tool. Once it is in, efficiency looks like an accelerator that still needs a policy tool to steer it. The difference is decisive for climate and AI-energy planning.

How it appears in AI systems

Direct rebound: an efficient car is driven further. Indirect rebound: money saved is spent on other emitting activities. Economy-wide rebound: cheap AI compute enables entirely new sectors of demand, expanding the perimeter of consumption faster than any single deployment saves.

Examples

• Cheaper LED lighting expands the total hours and area lit.
• Efficient AI inference lowers the cost per query, and organisations respond by running orders of magnitude more queries.
• Cheaper delivery logistics expand the range of goods delivered daily, not the volume delivered per household.