Precommitment

What it is

The distinguishing feature is timing. A resolution is made and then tested at the moment of temptation, which is the moment the person is least equipped to hold it. A precommitment is made at one moment and takes effect at a different one, when the person who made it is no longer in a position to overturn it cheaply. That is what separates a Ulysses pact from a plan. The pact is designed by a self who expects to be overruled, and it deliberately disempowers the self doing the overruling.

Institutions already run on it, which is a decent argument that the structure is sound even where the psychology is unsettled. Advance healthcare directives exist precisely because the future person may not be competent to decide. Early-withdrawal penalties on retirement accounts make a rational future decision expensive on purpose. Marriage is a legal contract rather than a standing intention. None of these are motivational techniques. They are arrangements that make one option costly or unavailable before anyone wants it.

Nir Eyal sorts the personal versions into three types in Indistractable, and the taxonomy is the useful part of his treatment. Effort pacts make the unwanted behaviour harder, adding friction rather than prevention. Price pacts make it cost money, staking something the person already feels they own. Identity pacts make it inconsistent with who the person takes themselves to be, so that the individual decisions stop registering as decisions.

Eyal is emphatic that precommitment is a last resort rather than a first move, and the sequencing claim is the part most readers skip. In his model it is the fourth step, and precommitments made before the internal triggers, the traction and the external triggers have been dealt with "will be useless." A timer-locked box will not hold against a person using their phone to avoid a feeling they have not looked at. He also concedes, straightforwardly, that researchers do not fully understand why precommitment works as well as it does.

That concession is the right level of confidence for this page. Precommitment is a design pattern with a long institutional record and a thin experimental one, and the vault has not assessed the underlying evidence for the personal versions. What earns it a place in an argument about willpower is not the strength of its evidence base but its independence from a contested one. When the mechanism of self-control is unsettled, the interventions worth reaching for first are the ones whose value survives whichever account turns out to be right.

In effect

The writer who removed the option

Jonathan Franzen wrote on an obsolete Dell with the wireless card physically removed, an Ethernet cable glued into the port and then sawn off at the head. Jhumpa Lahiri drafts longhand. Quentin Tarantino works in notebooks. None of them requires the writer to want the internet less on any particular morning. They are effort pacts, and the honest description of what they achieve is friction rather than prevention. Eyal's own example concedes the limit: the timer-locked container can be opened with a hammer, and the point of the design is that you probably will not fetch one. The commercial versions are now everywhere, from the Kitchen Safe to SelfControl, Freedom, Forest, iOS Downtime and Android Digital Wellbeing, all selling a few seconds of inconvenience placed between a person and a habit. The social version is more interesting and less sold: co-working in timed sprints with a friend, or services like FocusMate that pair strangers over video, manufacturing back the ambient social pressure a visible desk and a visible screen used to supply for free. Eyal discloses that he invested in FocusMate.

Money already felt to be yours

A price pact puts the person's own money at stake, forfeited if they do not follow through, which is why it is usually explained through loss aversion rather than reward. The prospective loss of something already owned does work that an equivalent prospective gain does not. Eyal's personal version is a hundred-dollar note taped to a wall calendar, to be burned on any week he does not exercise, and he reports three years without burning it. He used a larger pact with a friend to finish the book itself. He also names four pitfalls, and they matter more than the technique: price pacts fail against external triggers a person cannot escape, they suit short defined tasks rather than open-ended ones, they punish people who are already struggling, and they should not be used before the earlier steps are in place. The trial usually cited for deposit contracts reaches this site only through the trade literature, without a named study, so its figures are not repeated here.

The pact that changes the person

An identity pact binds a self-image rather than a behaviour. The claim is that identity works as a shortcut: once something is who you are, the individual refusals stop being refusals. "I don't" beats "I can't," because the first describes a person and the second describes a rule being obeyed under protest. The research anchor is Christopher Bryan's work on voter framing, in which registered voters answered pre-election surveys phrased either around the verb, how important it is to vote, or the noun, how important it is to be a voter, and were then checked against public voting records. The noun framing raised measured turnout, and the result held in a second election. Whether that generalises from a civic act to a snack is a step the study does not take, and one part of the literature Eyal draws on here has since been retracted.

What it does not say

It does not say that removing options is always benign. A device that binds a future self is defensible when the present self is the one setting it, and the same machinery in someone else's hands is choice architecture at best and coercion at worst. The ethics sit in the question of who is allowed to bind whom, and the answer does not follow from the technique working.

It does not say that precommitment substitutes for the rest of the work. Eyal's sequencing claim is a limit rather than a flourish. A pact placed in front of an unexamined discomfort tends to produce an inconvenienced person with the same discomfort, and the tools are sold hard precisely to people who have not done the earlier steps, because those are the people looking for something to buy.

It does not carry the evidence base its confident presentation implies. This page is tagged unreviewed, which is a statement about this publication rather than about the finding: the definition is sound and nobody here has assessed the experimental support for the personal versions. Eyal's own admission that the mechanism is not well understood should be read as the ceiling on any claim made in this area. The three-part taxonomy is a useful sorting device that one author proposed, not a result.

One specific caution about the identity strand, because it is the strand most likely to be repeated uncritically. The ritual argument in that chapter of Indistractable rests on a Harvard Business School study in which an arbitrary three-step pre-meal ritual outperformed mindful eating on calories, fat and sugar. That paper, Tian and colleagues in the Journal of Personality and Social Psychology, was retracted in 2024 after the authors' own reanalysis found discrepancies against the underlying data that invalidated the self-control conclusion. Do not cite the ritual study. The Bryan voter work and the "I don't" framing are separately evidenced and are not affected by the retraction.

---