A proposed principle of effort conservation in economic and organisational systems
Any effort to reduce work is presumed to displace that effort rather than eliminate it.
Ian's Law of Effort Displacement
Any effort to reduce work is presumed to displace that effort rather than eliminate it.
Any effort to reduce work is presumed to displace that effort rather than eliminate it.
Like the conservation of matter or energy, effort should be treated as conserved until demonstrated otherwise — not as a physical certainty, but as a practical presumption.
Work displaced from one part of a system may reappear elsewhere, at another time, or in another form. Genuine efficiency is established only when the total economic cost and quality of the outcome are considered.
Effort is presumed displaced, not eliminated.
Moving the work is not the same as eliminating it.
Effort is treated as conserved until demonstrated otherwise. The analogy to matter and energy is practical, not literal — and the presumption can be rebutted by evidence.
Whoever claims the work has been eliminated carries the burden of demonstrating it. Until then, assume it moved.
Moving work to a more capable actor, a better stage of the process or a more efficient resource can improve the system. The law asks you to show it, not to assume it.
Cheaper is not the same as better. Degradation that follows displaced effort belongs in the ledger, even when it never arrives as an invoice.
1 — The conservation principle
The central proposition is that work does not simply disappear because it is no longer visible where it was previously performed. An intervention that reduces local effort creates a presumption that the effort has been displaced.
This is not a claim that effort is a physically conserved quantity. Effort does not obey a conservation law, and nothing in economics or physics requires it to.
The analogy to the conservation of matter or energy is a practical one: it tells you where to look first, and it tells you not to treat a disappearance as proof.
It is a rebuttable presumption. The burden of demonstrating genuine efficiency rests with whoever claims the work has been eliminated.
Real efficiency gains are possible. The law challenges the assumption that a reduction in visible effort is sufficient evidence of those gains — nothing more, and nothing less.
No compensating work is created anywhere else in the system — not upstream, not downstream, not in another form.
Total economic cost falls, including the opportunity cost of whatever the released capacity is now used for.
Quality is maintained or improved — in reliability, usefulness, effectiveness and maintainability.
Meet all three and the presumption has been rebutted. That is a demonstration, not a contradiction of the framework.
2 — The three dimensions of displacement
Conceptual model: effort may move in either direction along each axis, or across multiple axes simultaneously.
| Dimension | Question | How effort moves |
|---|---|---|
| Horizontal | Who does the work? | Between people, teams, organisations, customers, suppliers or machines |
| Temporal | When is the work done? | Upstream into planning and preparation, or downstream into maintenance, correction and rework |
| Transformational | What form does the effort take? | Between human labour, capital, energy, computation, infrastructure and other resources |
The dimensions describe the location and form of displacement, not whether that displacement is beneficial. Displacement can reduce total cost when work moves to a more capable actor, a better stage of the process, or a more efficient resource.
3 — The ceteris paribus problem
Ceteris paribus means all other things being equal. It is useful when isolating variables, but potentially misleading when evaluating interventions intended to reduce work — because the intervention itself may change the things being held constant.
Reducing the effort required to write software might increase the effort required to specify, test, secure, maintain or operate it.
The effort excluded by the assumption may be precisely the effort that has been displaced.
If those consequences are excluded from the comparison, the apparent efficiency gain may be an artefact of the measurement boundary. This is the central methodological warning of the law.
4 — The economic test
Displacement is not inherently inefficient. Its consequences determine whether the intervention represents an improvement. A meaningful evaluation must consider total economic cost, opportunity cost and outcome quality — evaluated consistently, without counting the same consequence twice.
V is the value of the outcome, including quality. C is total economic cost, including opportunity cost. The ratio is a useful comparison when value and cost can be meaningfully quantified.
Direct and indirect costs across the system, including resources consumed by displaced work.
The value of alternative activities gained or foregone by the actors involved.
Changes in reliability, usefulness, effectiveness, maintainability and other relevant attributes.
Quality degradation is a cost because it has parts you can count. Split it, then sum the parts. If quality improved, the term is negative and works the same way.
Value lost because quality fell is booked here, not in V — otherwise the same loss is counted twice.
Value per unit of cost rises
The work moved somewhere it can be done more cheaply, and the ratio of value to total cost improved — including the effort it took to move it.
Ask Did value rise by more than cost rose?
Little net change
The work moved, but the savings at the new location roughly cancel the cost of the move and the new dependency.
Ask Do the gains and the new costs cancel out?
The outcome becomes worse
The work moved to somewhere more expensive to run, or the move created work that did not exist before — and the total cost rose.
Ask Did hidden costs outweigh the local saving?
The law does not prejudge which outcome occurs.
5 — Practical examples
The same instinct — reduce the work — produces displacement in three cases and a rebuttal in the fourth. The law does not tell you which is which in advance. It tells you what to check.
AI reduces the effort required to generate code. Some effort may move upstream into specification and prompting, downstream into verification and maintenance, and into computation and infrastructure. The resulting software may be cheaper and better. Or it may not. The reduction in coding effort alone cannot establish that.
Judge the software, not the volume of code.
A business reduces administrative labour by allowing customers to complete tasks themselves. The work has moved horizontally from the business to the customer. The business may report lower costs while the total effort across all participants remains unchanged or increases.
Count the effort of everyone the change touches.
A repetitive manual process is automated. Human effort is displaced into system design, equipment, energy, maintenance and exception handling. Automation can still produce substantial net savings. The law requires those savings to be demonstrated rather than inferred from the disappearance of the original manual task.
Savings must be shown, not inferred from the vanished task.
An unnecessary approval step is removed entirely, without creating compensating work or reducing quality. This may be genuine work elimination. It is a valid rebuttal to the presumption of displacement, not a contradiction of the framework.
This is what a rebuttal looks like. Show all three.
6 — Relationship to existing principles
Ian's Law draws on established ideas, but approaches them through a specific presumption. Its contribution is to combine a conservation-inspired presumption with a three-dimensional framework for investigating displacement.
If the work has disappeared here, where should we look for it?
Concerns the distortion of measures when they become targets.
Ian's Law is about a different distortion: the disappearance of the work itself.
Concerns the value of foregone alternatives.
Ian's Law adds the requirement to actually look for the displaced work before valuing it.
Explains why transferring work can improve efficiency.
Ian's Law asks where that transfer really ends, and who carries what it leaves behind.
Considers costs beyond the immediate transaction.
Ian's Law asks the prior question: whether the transaction's central work was displaced at all.
This is a conceptual and methodological principle, not an established physical or economic law. Its usefulness rests on whether it reveals costs and consequences that narrower assessments overlook.
In plain terms
The framework is easier to feel than to formalise. These stories are how the law shows up in ordinary conversations — at the office and at home.
The everyday reading If AI makes writing software nearly free, renting it stops making sense. Build it yourself and stop paying rent.
Writing it is the cheap part. Owning it is not.
AI may make applications cheaper to write without making them cheaper to own. Replacing a subscription transfers development, security, maintenance and operational responsibility from the vendor to the customer — and that bill arrives as your people’s time, not as an invoice.
AI may genuinely make custom software more economical in some cases. The law does not predetermine the winner; it requires you to add up the whole bill before declaring victory.
The everyday reading If a plumber can build an app with AI, software vendors are in trouble.
Capability was never the constraint. Cost is.
The question is not simply whether a plumber can build an application with AI, but whether building and maintaining it is a better use of their time and resources than buying software and concentrating on plumbing.
Comparative advantage and opportunity cost decide this, not what the tools are capable of. The displaced work does not disappear — it changes who carries it, and when.
The everyday reading The wash costs money every week. Wash it less often, scrub harder when you do, and the budget improves.
Less frequent is not less effort. It is later effort, and it is bigger effort.
Reduce the frequency of washes and the work does not disappear — it accumulates. The dirt stays where it was, so every wash now has more to remove and the effort per wash rises to offset the saving. That is the temporal axis in miniature: the work moved later, and it moved bigger. And if you trade the car wash for doing it yourself, it moved sideways too — from a Tuesday spent somewhere else to a Saturday at the driveway.
Quality is not far away either: a car washed rarely stays cleaner for longer, but its finish takes more punishment and its resale value takes the hit. Whether this is positive or negative displacement depends on the horizon and on what the car is worth to you. The law does not prejudge which.
The law is about cost, not capability.
Running the assessment
The presumption is cheap to state and expensive to discharge. These are the practical mechanics of discharging it — and the one way the law can be wrong in public.
Evaluate over the full life of the system, not the sprint in which the change landed. Deferred costs count, discounted as expected values — never ignored because they arrived later.
The same licence, hour or headline number must not appear on both sides of the ledger. If two teams price the same thing differently, the comparison is noise.
Every effort the intervention might have created elsewhere belongs inside the comparison. The excluded effort is the most likely place the displaced work is hiding.
Where the timing or size of a cost is unknown, estimate the expected loss and carry it. A risk you refuse to price is a cost you have chosen to forget.
If displacement genuinely lowers the unit cost of effort and nothing else constrains demand, total effort should rise, not fall. Cheap code becomes more code. A cheaper process produces more process. That is the testable consequence of the presumption.
Local effort falls, total system cost rises, quality holds.
The law is workingLocal effort falls, total cost falls, quality holds.
The presumption is rebutted — rare, and worth provingLocal effort falls, total cost holds steady, quality quietly drops.
Someone is paying in defects, rework or trustA team can cut its own bill while raising the system's. Before calling a change efficient, find out who is carrying the remainder of the work — and check whether they are in the room.
If the saving is real but the cost lands somewhere else, the change is redistribution, not efficiency.
In one sentence
Effort is presumed displaced, not eliminated.
Moving the work is not the same as eliminating it.
Ian's Law is a proposed conceptual heuristic, not an experimentally established conservation law. The analogy to the conservation of matter or energy is a practical presumption, not a physical equivalence — effort does not obey a conservation law, and nothing in economics requires it to. Its usefulness rests on whether it reveals costs and consequences that narrower assessments overlook.
A presumption can be over-applied. Genuine process simplification — removing an unnecessary task altogether — is real work elimination, and no one would call that displacement. The framework treats this as a rebuttal rather than a contradiction, which means the law's accuracy depends on being able to recognise a rebuttal when it appears.
The law tells us where to look. The economic assessment tells us whether the change was worthwhile.