App blockers work until the day you really need in. That day the block is not a helpful structure, it is an obstacle between you and something you decided matters. You disable it, and once you have disabled it once the cost of disabling it again is zero. Two weeks later the app is gone.
This is not a willpower failure and it is not a bug in any particular app. It is the predictable result of removing a choice from someone who had already chosen. Psychology has a name for it.
Reactance is the mechanism
Psychological reactance, described by Brehm in 1966, is the motivational response to a perceived loss of freedom. When a behavior is restricted, the restricted behavior becomes more attractive, and people act to restore the freedom.
A hard block is a small, repeated freedom restriction, applied by software you installed yourself. It produces exactly the response the theory predicts: first irritation, then circumvention, then removal.
Self-determination theory points the same way from the other side. Deci and Ryan's account of durable motivation is that it has to be internalized rather than imposed. A behavior you maintain because a piece of software will not let you do otherwise is by definition external. Take away the software and the behavior goes with it, which is what the deletion is.
The practical version of this: the goal of a screen time app is not to stop you today. It is to still be installed in March.
The evidence points at friction, not walls
The interesting part is that the softer intervention also measures better.
The InteractOut study at CHI 2024 tested subtle input friction against the conventional timed lockout across a five week field experiment with 42 participants. Against the lockout, friction cut usage time by a further 15.6 percent and opens by 16.5 percent. It also scored 25.3 percent higher on acceptance, with less reported frustration.
The PNAS field study of one sec, with 280 participants over six weeks, found participants abandoned the app they had just opened 36 percent of the time the intervention appeared, and by week six were attempting to open target apps 37 percent less than in week one.
Both results share a shape. The intervention asks a question, makes it genuinely easy to answer either way, and the answer is often no. Nothing is locked.
What Munen does instead of a door
Munen keeps the app reachable. Opening a flagged app costs a pause of up to ten seconds and then gives you a limited access window of a few minutes. Both numbers come from a zone recalculated each night from how you used those apps the day before.
| Blocker pattern | Munen pattern |
|---|---|
| The app is closed during a window | The app is always reachable, at a cost |
| You set the strictness by hand | Your last 24 hours set it |
| Breaking the rule needs you to disable the app | Getting in is a normal, expected path with a price |
| Nothing carries from yesterday | Yesterday is the only input |
| The worst day gets the hardest wall | The worst day gets a longer pause and a small window |
That last row is the important one. In a blocker, your worst day and its hardest restriction arrive together, which is the exact moment the app gets deleted. Munen's default on the lowest zone is a ten second pause and a three minute window, not a locked screen.
A wall does exist, and you can turn it on. The app calls it the hard day contract, and you choose it ahead of time, in a calm moment, for what should happen on your worst days. It is a commitment device you opted into, not a surprise. That distinction is the whole difference between structure and a door slamming.
The counterintuitive part: you are allowed in
The access window is not a loophole. It is the mechanism.
If getting in is possible but costs something, then getting in is a decision, and a decision can be counted. Munen counts it: minutes spent inside access windows are a signal in the zone score. Spend 45 minutes there and you lose a zone. Spend under 15 by 8 pm from a low zone and you gain one back the same evening.
A blocker cannot do any of this, because it has no data about what you do when the block is off. It only knows whether you complied.
Common questions
So blockers never work? They work well for a specific case: a defined window you want protected, like a work block or after 10 pm, where you have no legitimate reason to be in the app. That is a scheduling problem and a blocker solves it. The reflex problem is a different problem.
Is a pause not just a slower block? No, because the outcome differs. A block ends with the app closed. A pause ends with you choosing, and in the field data a third of the time the choice is to close it.
What stops me from just pushing through every time? Nothing stops you, and that is intentional. But it is counted, and doing it repeatedly lowers your zone, which makes tomorrow's pause longer and tomorrow's window shorter.
What if I want to be locked out? Turn on firm mode in the hard day contract. It gives zone 1 no entry at all. It is never the default, because for an audience dealing with burnout or low mood a zero-recourse wall on the worst day is the single most likely thing to get the app deleted.
The short version
Blockers fail on the day you need in, and reactance explains why. Friction measures better than lockouts and people accept it more. Munen makes getting in possible, priced, and counted, and moves the price with your last 24 hours. For the mechanics side by side against a scheduled blocker, see Munen vs Opal, and for the reasoning behind the ten second cap see why the pause is never longer than ten seconds.