Friction decides more than motivation does

Effort is not linear in its effects

A small obstacle placed between a person and an action reduces how often the action happens by far more than the obstacle would seem to justify. An extra click, a password to retrieve, an item in another room, a form requiring a piece of information you have to look up — each is trivially surmountable and each reliably reduces the rate at which people proceed.

The same asymmetry runs the other way. Removing a step raises uptake disproportionately. This is not because people are lazy in any interesting sense; it is because most behaviour is decided in a low-attention mode where the relevant comparison is between doing this thing now and doing something else that is already easier. In that comparison a small cost is decisive, because the alternative is right there and costs nothing.

Defaults are the strongest form of it

The most powerful version of a friction effect is the default: whatever happens if the person does nothing. Defaults are enormously influential across many domains, and this is one of the better-established findings in applied behavioural work, having been observed repeatedly in real administrative systems rather than only in laboratories.

Several mechanisms are proposed. The default requires no action, so inertia favours it. It may be read as an implicit recommendation from whoever designed the form. And changing it requires a decision, which means confronting a trade-off that most people would rather not confront. These are not competing explanations so much as forces operating together, and their relative weight probably varies with the domain.

The consequence is that whoever sets defaults exercises real influence, whether or not they intended to. There is no neutral option — some state of affairs must obtain when nothing is chosen — which is why the ethics of default-setting is a live argument rather than a technicality.

The effect sizes are contested

It is worth being careful here, because this area has its own evidence problem. A large literature on behavioural interventions accumulated over the past two decades, and recent examinations of it have found signs of the familiar difficulties: publication favouring successes, small studies producing large estimates, and effects that shrink substantially when the whole body of work including unpublished trials is accounted for.

The reasonable summary is that the average intervention of this kind is real but considerably smaller than the enthusiastic early literature suggested, and that some categories perform much better than others. Defaults and changes to the physical arrangement of options have held up comparatively well. Interventions relying on informational messages or social comparison have looked weaker under scrutiny. Anyone quoting a headline figure from a decade-old study is probably quoting a number that has since come down.

Using it on yourself

The personal application is straightforward and unglamorous. Increase the number of steps between you and what you do too much of; reduce the number between you and what you do too little of. Neither requires resolve at the moment of temptation, which is the point, because that is the moment when resolve is least available.

The most effective version operates at the level of what is present rather than what is nearby. An item that is not in the house cannot be reached for; an application that is not installed cannot be opened absent-mindedly. This is a decision made once, in a deliberate state, that then governs many later moments without further input. Compare that with a rule requiring a fresh act of restraint every evening.

Why this reframes the willpower argument

A great deal of advice treats self-control as the central variable and the environment as background. The evidence points the other way: people who appear consistently self-controlled seem, when studied, to encounter fewer difficult moments rather than to win more of them. Some of the research supporting that picture is observational and correlational, so the causal direction is not fully established, and it should be held loosely.

But the practical implication survives the uncertainty, because it costs nothing to act on. Arranging your surroundings so that the thing you want is the path of least resistance is available to anyone, works while you are not paying attention, and does not depend on any contested theory about how self-control operates. That last property is worth something on its own in a field with this much unsettled in it.