Skip to main content

Commitment Devices: How to Bind Your Future Self

A commitment device locks in a choice before temptation arrives. What 7 field experiments show about who signs up and who quits, plus soft versions to try.

Commitment Devices: How to Bind Your Future Self

A commitment device is a choice you make now that makes a tempting choice later harder, costlier or impossible. Odysseus had himself tied to the mast so he could hear the Sirens without steering toward them. Modern versions include a savings account you cannot touch until a date, a cash deposit you lose if you skip the gym, or a work block you tell a colleague about in advance.

Key Takeaways

  • These arrangements restrict your own future options on purpose. Hard ones carry a real penalty; soft ones mostly carry a psychological cost.
  • Field experiments show they can work for the people who take them up: savings rose, smokers quit, and gym habits lasted years after the program ended.
  • Most people turn them down. In one large smoking trial, 90.0% accepted a reward offer but only 13.7% accepted a refundable deposit.
  • A commitment that is too weak can leave you worse off. In a second Philippine savings study, 55% of clients broke their own contracts and paid the penalty.
  • A famous study on self-imposed deadlines was retracted in September 2026, and a replication found that changing the deadlines had a negligible effect on performance, so do not lean on it.
  • Start with a soft version, aim it at one specific moment of temptation, and only add money once you know where you slip.

What is a commitment device?

It is an arrangement you set up in advance to limit what your future self can choose. You know that later, in the moment, you will be tempted to do the thing you would rather not do, so you make that option more expensive now, while you still see clearly. The cost can be money, effort, embarrassment or plain impossibility.

In their review of the economics literature, Gharad Bryan, Dean Karlan and Scott Nelson describe it as an arrangement that restricts a person’s future choice set by making certain choices more expensive, perhaps infinitely expensive. They separate two kinds:

  • Hard commitments carry a real penalty or remove the option entirely: money you forfeit, a savings account locked until a date.
  • Soft commitments carry mainly a psychological cost: a promise to a friend, a goal you announce, a block on your calendar you have labelled.

The two are ends of a spectrum rather than separate boxes, and many everyday versions sit in between.

The oldest example is in Homer. In Book XII of the Odyssey, Odysseus tells his crew: “If I beg and pray you to set me free, then bind me more tightly still.” He also stopped his crew’s ears with wax, which is the other classic move: removing the temptation instead of resisting it.

The idea goes by several names. “Precommitment” means nearly the same thing and is common in economics papers. A “commitment contract” is a version with money at stake.

Commitment device examples

Common examples are telling a colleague when a draft is due, booking a focused work block, deleting a social app during a project, a savings account locked until a date, and a cash deposit you lose if you miss a goal. The table below runs from the softest to the hardest. The last column says what breaking each one costs you, which is what makes it work, or fail.

ExampleTypeWhat breaking it costs
Telling a colleague you will send the draft by FridaySoftEmbarrassment
Booking a focused work block in your calendarSoftBreaking a plan you made
Keeping tempting audiobooks only on your gym playlistSoft to hardNo audiobook today
Deleting a social app from your phone during a projectSoft to hard (effort)Reinstalling it
A savings account locked until a date or amountHardAccess to your own money
A cash deposit you lose if you miss your goalHard (money)The deposit
A Ulysses contract in psychiatric careHard (legal)Not possible by design: you agreed in advance that a later objection, made without capacity, would not count

Most everyday productivity versions sit in the top half of the table. The bottom half is where most of the research is.

Do commitment devices work?

For the people who choose them, the best field evidence says yes, often for longer than the program itself. The catch is that only a minority choose them, so the effect across everyone offered one is much smaller than the effect on the people who signed up.

Five field experiments show the pattern, and two more, below, show its limits:

Savings. Nava Ashraf, Dean Karlan and Wesley Yin offered a locked savings account to 710 clients of a Philippine bank. 202 (28.4%) opened one. After twelve months, average savings balances in the group offered the account had risen by 81 percentage points relative to the control group.

Smoking. Xavier Giné, Dean Karlan and Jonathan Zinman offered smokers an account they paid into for six months, after which they took a urine test; failing it sent their money to charity. Only 11% took it up, but smokers offered it were 3 percentage points more likely to pass the six-month test, and the effect held in surprise tests at twelve months.

Exercise. Heather Royer, Mark Stehr and Justin Sydnor ran a four-week gym incentive at a Fortune 500 company. The incentive alone left only small lasting effects. Offering a self-funded commitment contract at the end produced significant long-run changes, still detectable several years later.

Healthy food. In a study by Janet Schwartz and colleagues, 36% of households offered the chance to stake their 25% grocery discount on buying more healthy food agreed, and they then raised their healthy purchases by an average of 3.5 percentage points in each of the six months.

Temptation bundling. Katherine Milkman, Julia Minson and Kevin Volpp let gym-goers listen to page-turner audiobooks only at the gym. They visited 51% more often than the control group at first, though the effect faded over time, especially after Thanksgiving. Afterwards, 61% said they would pay for gym-only access to the audiobooks, which suggests people wanted the restriction.

The fading effect in the audiobook study is a warning. A device that works for a few weeks buys you a few weeks in which to build a habit, and the habit still has to form.

Why most people turn down a hard commitment

Most people offered a hard commitment say no, and the ones who say yes are not a random sample. That is the biggest limit on how much good these arrangements can do, and it explains why a device that works brilliantly for some people barely moves the average.

The clearest evidence is a 2015 trial by Scott Halpern and colleagues in the New England Journal of Medicine. They randomly assigned 2,538 CVS Caremark employees and their friends and relatives to smoking-cessation programs. One kind paid rewards of about $800 for quitting. The other asked for a refundable $150 deposit plus $650 in rewards. 90.0% accepted the reward programs; 13.7% accepted the deposit programs.

Across everyone assigned, the reward programs won: 15.7% were still abstinent at six months, against 10.2% for deposits and 6.0% with usual care. But among the estimated 13.7% who would accept either kind of program, the deposit produced 13.2 percentage points more abstinence than rewards. The penalty worked very well for the few people willing to risk it.

Economists explain the refusal with naivety about present bias, the tendency to weigh what you want now far more heavily than what you will want later. People who know they have it are the ones who seek out a binding plan. People who do not see it in themselves see no reason to bind anything.

When a commitment backfires

A commitment can hurt you when the penalty is real but too small to change what you do. You pay the price of breaking it and still do the thing you were trying to avoid. Choosing the penalty yourself does not protect you, because the person choosing it is the same person who underestimates the temptation.

Anett John tested this in a field experiment published in Management Science in 2020. Low-income clients of a Philippine bank were randomly offered an installment-savings account and chose their own savings plan and default penalty. Average savings went up, but 55% of clients defaulted and lost money. John’s suggested explanation is that the penalties people picked were too low to overcome their self-control problems: people who partly understood their present bias committed, but too weakly.

The practical rule: if you would rather pay the penalty than do the task, all you have set up is a fee.

What about self-imposed deadlines?

A self-imposed deadline is a commitment many knowledge workers reach for, and the study usually cited for it no longer stands. In September 2026, Psychological Science retracted Dan Ariely and Klaus Wertenbroch’s 2002 paper “Procrastination, Deadlines, and Performance: Self-Control by Precommitment”, which had reported that people set themselves costly deadlines and that those deadlines improved their work, though less than some deadlines imposed from outside.

A replication of Study 2, the proofreading experiment, by Kyle Hyndman and Alberto Bisin was published in the same journal in July 2026. Days later, the Data Colada researchers privately sent the authors and the journal’s editor analyses that, in Data Colada’s assessment, showed the data in Studies 1 and 2 had been tampered with. According to Retraction Watch, Wertenbroch asked for the retraction on July 23; the analyses appeared as blog posts on August 31 and September 2, the date of the retraction notice. The replication itself found that changing the deadlines had a negligible effect on all three performance measures, and evenly spaced external deadlines did not stand out either.

Deadlines may still help you; the problem is that the best-known evidence for them is gone, and a deadline nobody else sees is a very soft commitment. If a deadline matters, attach it to something outside your head: a person waiting for the work, or a meeting already booked. Our guide to why tasks run over explains why the deadline itself also needs a realistic estimate behind it.

Precommitment and the Ulysses pact

“Precommitment” and “commitment device” mean nearly the same thing: committing in advance to limit a later choice. “Ulysses pact” or “Ulysses contract”, after the Latin name for Odysseus, is the same idea with a formal agreement behind it, and in medicine it has a narrower, serious meaning.

A 2023 integrative review in Bioethics by Connor Brenna and colleagues describes these clinical agreements: in many jurisdictions a patient can make an advance medical decision that includes a special provision giving up their ability to object to it later, should they lose decision-making capacity. They appear under several names, including Ulysses contracts, Odysseus transfers and psychiatric advance directives with Ulysses clauses. A 2024 review in Transplantation by Daniel Hurst and colleagues notes that they are most often used in psychiatry. That clinical sense is a legal and ethical matter for patients and clinicians, well beyond a productivity habit.

In everyday use, a Ulysses pact is any agreement you make with yourself, or with someone you trust, to hold you to a plan when you later want out.

How to set up a commitment device

Pick one moment of temptation, choose the softest arrangement that would actually stop you in that moment, and test it for two weeks before raising the stakes. The research above points the same way: aim it narrowly, size the cost honestly, and expect to adjust.

  1. Name the moment, not the goal. “Exercise more” is a goal. “On weekday evenings I sit down on the sofa at 18:30 and do not get up” is a moment. Commitments work on moments.
  2. Start soft. Tell someone, book the time, or move the temptation out of reach. Odysseus’s wax, removing the temptation, is often easier than his rope.
  3. Make the cost bigger than the temptation. If you would happily pay the penalty, raise it or choose a different one. Anett John’s study is the warning here.
  4. Add another person. A friend who checks in, a colleague expecting the draft, or a referee who decides whether you met the goal. Other people make soft commitments harder to quietly drop.
  5. Write an exit rule. Decide in advance what counts as a legitimate reason to break the commitment, such as illness, so you do not renegotiate it every time you feel tempted.
  6. Review after two weeks. If you never broke it, keep it. If you broke it repeatedly, it was too weak or aimed at the wrong moment. If your real trouble is starting at all, the 5-minute rule is a lighter tool for that.

Pairing a “want” with a “should”, as in the audiobook study, is a gentle way to begin, and it can work alongside the extrinsic rewards you already use.

Can a planning app be a commitment device?

A planning app is a soft commitment at most. It can make your plan visible and specific, and make breaking it a conscious act, but it cannot fine you, and most of the time it should not. Its value is in turning a vague intention into a scheduled, time-boxed promise you can see.

FocusBox works this way. Its morning ritual, Plan My Day, has you pick today’s tasks, give each a time estimate drawn from your own history, and order them; a capacity meter warns you when you plan more than the day can hold, but it never blocks you. Daily Shutdown, at the end of the day, shows what you finished and rolls the rest to tomorrow, with a streak that rewards closing the day. Both are described on the planning rituals page.

None of that is a hard commitment, and that is deliberate. If you need money or a referee on the line, add them outside the app. If what you need is a plan you can see and a block of time you have already agreed to, a timeboxed schedule is often enough. The FocusBox glossary entry on incentive systems covers rewards in more detail.

Frequently Asked Questions

What is an example of a commitment device?

A locked savings account is a classic example: you cannot withdraw the money until a date or amount you chose, so spending it on impulse is no longer an option. Others include a cash deposit you lose if you miss a goal, deleting a distracting app for a project, or telling a colleague when they will get your draft.

What is a commitment device in behavioral economics?

In behavioral economics, the term describes an arrangement a person chooses that restricts their own future choices, usually by making some options more expensive. Economists study such arrangements as a response to present bias, the tendency to favor immediate rewards over later ones, and as evidence that some people know about their own self-control problems.

What is the difference between hard and soft commitments?

A hard commitment carries a real penalty or removes the option, such as forfeited money or a locked account. A soft commitment carries mainly a psychological cost, such as breaking a promise or a plan you announced. Hard ones are stronger but fewer people accept them; soft ones are easier to adopt and easier to drop.

Do commitment devices actually work?

For the people who take them up, field experiments show real effects: higher savings after a year, smokers still abstinent at twelve months, and gym habits lasting years. But most people decline them, and a commitment with a penalty that is too small can backfire, as when 55% of clients in one savings study defaulted and paid.

What is a commitment contract?

A commitment contract is an agreement in which you put money or something else of value at stake on reaching a goal, and lose it if you fail. In one smoking study, participants paid into an account for six months and forfeited the money to charity if a urine test showed they still smoked.

What is a Ulysses pact?

A Ulysses pact is an agreement made in advance to hold yourself to a plan when you later want to abandon it, named after Odysseus binding himself to the mast. In medicine, a Ulysses contract is a formal advance decision in which a patient gives up the ability to object to a treatment later, should they lose decision-making capacity.

Sources

Bogdan

Founder at FocusBox

Bogdan builds and runs the tools this blog is about. He writes from what the products actually do in production, including the parts that break.

Follow along on Instagram at @focusboxio for campaign breakdowns.

AI To-Do List, Focus Timer, and Timeboxing App for ADHD

Designed for ADHD and Timeboxing: AI To-Do Lists, Timers, and More.

Designed for ADHD and Timeboxing: AI To-Do Lists, Timers, and More.