Jevons Paradox
Making a resource cheaper or more efficient often increases total consumption of it.
“It is wholly a confusion of ideas to suppose that the economical use of fuel is equivalent to a diminished consumption.”
To an executive
Making something cheaper rarely cuts the bill. People find more uses for it, so total spend can climb even as the price of each unit falls.
To an engineer
Track total consumption as well as unit cost. When you make compute, environments or AI calls cheaper or easier, add quotas, budgets and alerts.
The idea
In The Coal Question (1865), William Stanley Jevons argued that more efficient steam engines would not reduce Britain's coal consumption. They would increase it, because cheaper energy made coal worth using for many more purposes. In his words, it is "the very economy of its use which leads to its extensive consumption."
Technology follows the same pattern. When compute, storage, environments or AI tokens get cheaper and easier to obtain, people find many more uses for them, and total consumption and spend can rise even as unit cost falls.
Why it works
Lower cost unlocks demand that was already there. Uses that were not worth it at the old price become worth it at the new one.
Lower friction matters as much as lower price. Self-service removes the queue and the approval step, which were quietly rationing demand.
Unit metrics hide totals. A team celebrating a falling cost per transaction may not notice that transactions have grown tenfold.
Where it applies
- Cloud and FinOps strategy, where self-service and cheaper compute drive up total spend.
- Internal platforms, where making environments or pipelines trivial to create multiplies their number.
- AI adoption, where cheaper models invite more calls, longer prompts and more automated use.
- Business cases for efficiency programmes, which often assume savings that rising demand will absorb.
How to apply
- Expect demand to rise whenever you make something cheaper or easier, and estimate by how much.
- Track total cost and total use alongside unit cost.
- Tie consumption to value, with unit economics such as cost per customer or per order.
- Add light guardrails, such as budgets, quotas, tagging and automatic clean-up, rather than reintroducing gates.
Example scenario
A self-service platform made spinning up a test environment a matter of minutes instead of weeks. Within a year the number of environments had grown many times over, and cloud spend had risen sharply, even though each environment cost less than before.
The enterprise architect did not bring back the approval queue. Instead, environments got owners, cost tags and an expiry date by default, and spend per team was shown on a shared dashboard. Unused environments disappeared, and the teams kept the speed.
Questions to ask
- If this gets cheaper or easier, how much more will people use it?
- Do we track total spend as well as unit cost?
- What guardrails keep consumption intentional without killing the benefit?
When not to use it
- Demand that is genuinely capped, such as a fixed number of regulated reports, where efficiency gains do translate into savings.
- When the extra use is the goal, as with a platform built to encourage experimentation.
How it misleads you
Adding friction to stop efficiency gains. The extra consumption may be creating real value, so measure what it delivers before treating it as waste.
In the wild
Further reading
Last reviewed · Suggest an improvement