What is a commitment device?
A commitment device is any structure that lets you decide today what you will be allowed to do tomorrow, with a real consequence if you back out. The point is not to be in the moment, fighting yourself with willpower. The point is to settle the decision when your prefrontal cortex is doing its best work, then make it costly for future-you to renegotiate.
Three elements have to be present for it to count as a commitment device:
- A goal you set in advance, when you are not under the influence of the urge you are trying to manage.
- A real consequence attached to failing it. Money, social cost, automatic restriction, or removal of an option.
- A mechanism that makes removing the consequence harder than honouring it. If you can opt out cheaply when the urge hits, it is not a commitment device. It is a New Year's resolution.
The third element is the one that fails most often. A daily journal entry promising to exercise is not a commitment device. There is no consequence and no enforcement. A scheduled gym appointment with a friend who will charge you $50 if you no-show is.
Examples in everyday life
You probably already use a few commitment devices without calling them that:
- Automatic savings transfers on payday. The money leaves before you can spend it.
- Putting your alarm clock across the room. You have to physically get out of bed to silence it.
- Deleting Instagram from your phone. The friction of re-downloading is the commitment device.
- Telling friends about a project publicly. The social cost of failing becomes the consequence.
- Prepaying a personal trainer for the month. The sunk cost makes you show up.
- StickK contracts. You commit to a goal and put money on the line, going to a charity (or anti-charity) if you fail.
- Forfeit. You stake money on completing a habit and lose it without photo or GPS proof.
- Beeminder. Yellow-brick-road tracking with a financial penalty if your data goes off-track.
- ScreenFine. 25 pushups per 15-minute block you go over your daily phone-use limit (verified by camera or HealthKit).
The last four are explicit commercial commitment devices. The rest are improvised. Both work as long as the three elements are present: pre-decision, real consequence, hard to back out.
Why they work (the research)
Commitment devices work because of three well-replicated findings in behavioural economics:
1. Loss aversion. Kahneman and Tversky's prospect theory (1979) showed humans treat losing $X as roughly twice as motivating as gaining $X. Notifications and apps that frame benefits as gains ("save time," "improve focus") work weakly. A real 25-pushup lock is a loss, and the asymmetry kicks in.
2. Hyperbolic discounting. Humans value present rewards far more than future rewards, but the curve is irrational. We choose $50 today over $100 next week, but $100 in a year over $50 in 51 weeks (Ainslie, 1992). This creates the "preference reversal": calm-you wants to limit phone use; tired-you at 11pm does not. Commitment devices solve this by removing the renegotiation option.
3. Self-command. Schelling (1980, 1992) framed the problem as "the multiple self": calm-you and impulsive-you are effectively different agents with conflicting goals. Commitment devices are how calm-you constrains impulsive-you's choice set. Schelling specifically argued that pre-commitment is more effective than willpower because willpower requires the multiple selves to fight in real time.
Empirical results back the theory. Ariely and Wertenbroch (2002) found students who self-imposed deadlines on three papers performed better than students with one deadline at the end. Gine, Karlan, and Zinman (2010) ran a smoking-cessation trial in the Philippines: smokers offered a money-stake commitment contract were 30 percent more likely to quit than the control group, and the effect was still measurable 12 months later.
The mechanism is robust. The execution is what varies.
A short history of the idea
The original commitment device is in Homer's Odyssey. Ulysses orders his crew to plug their ears with wax and tie him to the ship's mast so he can hear the sirens' song without steering the ship into the rocks. He pre-commits to immobility because he knows that the in-the-moment Ulysses will not be capable of resisting. The deal is signed in advance, and the rope removes the option to back out.
The economic formalisation arrived 2,800 years later. Schelling, in The Strategy of Conflict (1960) and later essays on self-command (1980, 1984), used "Ulysses pact" as the technical term and developed the idea that rational agents will sometimes deliberately limit their own future options.
From there, the idea moved into clinical psychology (cue-exposure therapy with Pavlovian extinction), public policy (mandatory waiting periods for divorce, cooling-off periods for big purchases), and. Most recently. Consumer products. StickK launched in 2008. Beeminder in 2011. Forfeit in 2022. ScreenFine in 2026. Each one is a different shape of the same idea.
Soft vs hard commitment devices
Behavioural economists distinguish between soft and hard commitment devices.
Soft commitment devices raise the cost of breaking the commitment but do not eliminate the option. Examples: a self-imposed deadline, a habit-tracking app, an alarm clock across the room, telling a friend about your goal. They work for people whose self-control is just-barely-not-enough. A small nudge tips the balance.
Hard commitment devices remove the option entirely or attach a consequence severe enough that the choice becomes effectively binary. Examples: prepaying for a year of gym membership, an automatic savings transfer with no easy reversal, a money-stake contract going to charity, verified-exercise locks that hit your card without your active consent in the moment.
The tradeoff is autonomy vs effectiveness. Soft devices preserve flexibility. If circumstances change, you can adapt. Hard devices remove flexibility. Which is the point, but means a misjudged commitment costs you. Most people start with soft, find them not enough, and escalate to hard. ScreenFine is intentionally the smallest possible hard commitment device for screen time: a daily $5 cap puts a floor on the worst case, and pause-anytime preserves the autonomy escape hatch.
When commitment devices fail
Three failure modes are well-documented in the literature:
1. The consequence is too small. If $5 does not hurt, $5 will not change behaviour. You have to pick a stake aversive enough that future-you actually feels the loss. This is why a "lose your streak" gamification is weaker than real money: streaks are gain framing in disguise.
2. The goal is poorly defined. "Spend less time on my phone" is a bad commitment device target. "Spend under 90 minutes a day on my phone, measured by iOS Screen Time, between 7am and 11pm" is a good one. The goal has to be measurable without ambiguity, or you spend mental energy on edge cases instead of on behaviour.
3. The escape hatch is too easy. If you can disable the commitment in the moment, the device collapses to willpower. Apple Screen Time fails this test. The "Ignore for today" button is two taps. ScreenFine deliberately has no in-the-moment override; you can pause from settings, but pause is intentionally one extra step and labelled as a deliberate decision rather than a get-out-of-jail-free button.
A fourth failure mode is rarer but worth naming: commitment devices targeting the wrong problem. If your phone use is not actually willpower-limited but driven by a deeper issue. Anxiety, depression, an unresolved life problem. A commitment device will reduce minutes without improving anything. The minutes go somewhere else. For genuine clinical phone overuse, the answer is therapy + structural change, not a fine jar. ScreenFine is for sub-clinical heavy use; for anything more serious, talk to a clinician.
Commitment devices for phone use
Phone use is unusually well-suited to commitment devices because the behaviour is automatically measurable. iOS Screen Time and Android Digital Wellbeing track minutes per app to the second. You do not have to self-report. You cannot hide a slip-up from yourself. The only piece missing in the standard screen-time apps is a real consequence.
Apple Screen Time, Opal, One Sec, ScreenZen, Brick, Roots, and Jomo are all soft commitment devices: they add friction (a pause, a block, a notification) but the cost of ignoring them is zero. They work for people whose self-control is just-barely-enough.
ScreenFine is the explicit hard commitment device for the same problem. Same automatic tracking, same daily limit you set, but the consequence is a real lock on your tracked apps once you cross the line, with the unlock priced in physical effort (25 pushups, 1,000 steps, a workout, 10 mindful minutes). No in-the-moment override.
For the long version of how commitment devices apply specifically to screen time. Including the redemption mechanic, the daily cap, and how the AI villain delivery layer works. Read the commitment devices pillar guide. For the side-by-side mechanism comparison with another money-stake app, see ScreenFine vs Forfeit.