Negative Punishment

How negative punishment works

Negative punishment requires that a reinforcer be present or available before the behavior, so that it can be taken away after. A teenager who has the car keys can lose them; a child who is playing can be removed from play; a driver with a clean license can lose points. The behavior costs the person something they had, and its future rate drops.

• A child hits a sibling → is removed from the game for two minutes → hitting decreases.
• A player fouls → sits out for two minutes → fouling decreases.
• A driver is caught speeding → loses three points → speeding decreases.

The usual rules apply. The loss must follow the behavior closely (a privilege revoked on Friday for something done on Monday teaches little), be contingent on the behavior and nothing else, and be consistent. The A-B-C analysis is the same as for reinforcement, except that the consequence is a subtraction.

Check any scenario in two questions

Not sure whether a consequence is this quadrant or a neighbor? Answer the two questions and the checker names it.

Time-out from positive reinforcement

What time-out actually is

The full name matters: time-out from positive reinforcement. Time-out is a period during which, contingent on a behavior, the person loses access to the reinforcers that were available a moment earlier. The chair, the corner, and the bedroom are incidental; the procedure is the removal of access to reinforcement. It is not "go and think about what you did," and it works only when the environment the child leaves — the time-in — is actually rewarding.

Time-out entered clinical practice in the early 1960s. In one of the first applied case studies, a three-and-a-half-year-old boy whose tantrums and self-injury had made him impossible to treat was placed briefly in his room contingent on tantrums while appropriate behavior was reinforced; the tantrums declined until he could wear his glasses, eat at the table, and eventually attend school.^[2] Time-out has since become one of the most researched discipline techniques in existence and a standard component of evidence-based parent training.^[3]

Exclusionary vs. non-exclusionary time-out

The least restrictive version that works is the right one. For most children in most homes, that is a brief non-exclusionary or corner time-out, delivered calmly.

Why time-out fails when time-in isn't reinforcing

The most common reason time-out does not work is that the child was not losing anything. A classic pair of experiments made the point. In one, time-out reduced a teenager's tantrums when the time-in environment was enriched with activities and attention, but not when it was impoverished. In the other, time-out actually increased a young child's problem behavior, because the time-out period gave her uninterrupted opportunity for self-stimulatory behavior she found reinforcing.^[5] Time-out is only punishing relative to what it interrupts; for a child in a dull or demanding situation, being removed is an escape or an opportunity — hence the child happy to be sent to a bedroom full of toys, and the student who acts out to be sent from a hard lesson to a comfortable office. Before using time-out, ask what time-in offers. If the answer is "not much," fix that first.

What the evidence says

Decades of research, mostly with young children and their parents, support time-out for reducing aggression, noncompliance, and tantrums when it is brief, consistent, and embedded in a home rich in positive attention.^[3][6] It is a component of every major evidence-based parent-management program, and the American Academy of Pediatrics recommends it as an alternative to physical punishment.^[7] A 2019 review examined the claim that time-out harms attachment and concluded that, implemented as designed, the evidence does not support that concern — and that discouraging time-out risks steering parents away from one of the few discipline strategies with a strong evidence base.^[8] One caution: the time-out parameters recommended in popular books and websites often diverge from what the research supports, so the details below matter.^[9]

Response cost

Response cost is the removal of a specified amount of a reinforcer, contingent on a behavior — a fine, in other words. The term comes from laboratory work in the early 1960s in which people working for points lost some of them for responding under certain conditions, which reduced responding.^[10] In applied settings it usually means losing tokens, points, money, minutes of a privilege, or the privilege itself.^[11]

• A student in a token economy loses two tokens for leaving her seat without permission.
• A driver loses points from his license — and eventually the license — for moving violations.
• A teenager loses thirty minutes of screen time for each unfinished chore.
• A gym member forfeits a deposit for each missed class (a commitment contract).

Response cost fits naturally inside a token economy, where reinforcers are earned for desired behavior and lost for undesired behavior. The practical rules: the person must have something to lose (a child at zero tokens has no reason to behave); the cost should matter without wiping out the day's earnings; and earnings should outpace losses. A system in which people lose more than they earn becomes an aversive environment everyone tries to escape.

Examples of negative punishment

The dog-training row is negative punishment done well: the reinforcer removed (play) is exactly the one maintaining the nipping, the removal is immediate, and play resumes as soon as the dog is calm, so the contrast is unmistakable. It teaches without adding anything aversive, which is why it is the recommended way to handle puppy biting. Reward-based dog training › · Examples for every quadrant ›

Negative punishment vs. extinction

This is the distinction students most often get wrong, because both procedures involve a reinforcer failing to arrive and both reduce behavior. The difference is which reinforcer.

The test question: was the removed reinforcer the one paying for the behavior? If whining simply stops working, that is extinction — you declined to give something. If the whine costs a token, that is response cost — the token was never the reason for whining, and you removed it. The two behave differently: extinction produces an extinction burst and requires you to control the maintaining reinforcer, while response cost can be applied even when you cannot control what maintains the behavior (peer laughter, for instance).

Negative punishment vs. negative reinforcement

Both begin with "negative," so both involve removing something. That is where the similarity ends.

A single event can be both, for different people. When a parent carries a screaming toddler out of a restaurant, the toddler may be negatively punished (loses the crayons and the attention) while the parent is negatively reinforced (the embarrassment ends). And if a "punishment" removes something the person wanted to escape anyway, it has flipped into negative reinforcement.

Effectiveness and side effects

Negative punishment is the preferred form of punishment when punishment is used at all. It adds no aversive stimulus, so it produces less of the fear, reflexive aggression, and conditioned aversiveness that follow positive punishment; it is easier to apply consistently without escalation; and it fits naturally with reinforcement-based systems as the debit side of a ledger.^[12][13] But it is still punishment, and it shares punishment's limits:

• It does not teach a replacement. A time-out says what not to do. The child still needs a reinforced way to get what hitting was getting.
• Emotional responding and escape. Children resist going to time-out, argue about fines, and sometimes escalate; the struggle to enforce the procedure can become more aversive than the procedure.
• Response cost can provoke aggression and collapse. Losing a great deal at once elicits anger, and once a person is at zero there is nothing left to lose.
• Overuse turns a rich environment into a poor one. A classroom where tokens are constantly taken away is one children want to leave, which brings escape-maintained behavior with it.
• Delay kills it. "You're grounded next weekend" is a weak consequence for a Tuesday-night behavior.

How to do time-out correctly

Most failed time-outs fail on the details. The research-supported procedure looks like this:^[3][9]

1. Make time-in rich first. Time-out only removes what time-in provides. Catch the child being good many times a day.
2. Decide in advance which behaviors earn it. A short list — hitting, throwing, deliberate destruction. Minor behavior is better handled by planned ignoring and reinforcing the alternative.
3. Deliver it immediately and calmly. One brief statement ("No hitting. Time-out."), no lecture, no negotiation. Anger and explanation are attention, and attention is a reinforcer.
4. Keep it brief. A few minutes is enough; about one minute per year of age is the common rule of thumb for young children. Longer is not more effective and is harder to enforce.
5. End it on calm, not the clock alone. Release when the interval has passed and the child has been quiet for a few seconds, so release does not reinforce protesting.
6. Return to time-in without a debrief. Back to the activity, and reinforce the first good behavior you see. The point is the contrast: good behavior gets warmth and access; hitting gets a brief, boring pause.
7. Be consistent. Every listed behavior, every time, from every caregiver. Intermittent time-out teaches that the behavior sometimes pays.
8. Track it. If the behavior is not decreasing within a couple of weeks, the procedure is not functioning as punishment. Check whether time-in is rewarding and whether time-out offers an escape.

The same principles transfer to response cost: small, immediate, consistent, and embedded in a system where the person earns far more than they lose.

Common mistakes

• Time-out from nothing. Removing a child from a situation they wanted to leave is negative reinforcement, and it will increase the behavior.
• Time-out with a lecture. Explaining, scolding, and checking in during the interval deliver attention and turn the procedure into a conversation.
• Making it long. A thirty-minute time-out is less effective than a two-minute one and more likely to end in a fight.
• Fining people into the red. Once tokens hit zero the system has no leverage.
• Confusing it with extinction. Withholding the maintaining reinforcer is extinction; removing a different reinforcer is negative punishment. They need different planning.
• Using it instead of teaching. Punishment of any kind supplements reinforcement of the behavior you want; it never substitutes for it. Positive parenting strategies ›

Key takeaways

• Negative punishment needs a reinforcer that is present or available before the behavior, so that it can be taken away after. You cannot remove what the person does not value, and whether a removal is punishing is decided by its effect on the behavior.
• Time-out means time-out from positive reinforcement. The chair and the corner are incidental; the procedure works only when time-in is rewarding, and for a child in a dull or demanding situation, being removed is an escape or an opportunity.
• Response cost is a fine: a set amount of a reinforcer is removed each time the behavior occurs. It works only when the person has something to lose and earns more than they lose.
• Extinction withholds the reinforcer that has been maintaining the behavior; negative punishment removes a reinforcer the person already has, usually a different one. Ask whether the removed reinforcer was the one paying for the behavior.
• Negative punishment is the preferred form of punishment because it adds no aversive stimulus, but it still teaches no replacement, weakens with delay, and turns a rich environment poor when overused. Punishment supplements reinforcement of the behavior you want; it never substitutes for it.

Check yourself

A bored student acts out during a hard lesson and is sent to sit in the office. Acting out increases. Was this negative punishment that failed?

No. It was negative reinforcement. The student lost nothing they wanted and escaped something they wanted to leave, so the removal strengthened the behavior. A consequence is classified by its effect, not by its label.

A child whines for candy. Parent A never gives candy for whining again. Parent B takes a token off the sticker chart for each whine. Which one is negative punishment?

Parent B. The token was never the reason for whining, and it was removed because the behavior occurred: response cost. Parent A is using extinction, in which the maintaining reinforcer is withheld and nothing is taken away, and should expect an extinction burst.

A parent puts a child in a two-minute time-out for hitting, explains during the interval why hitting is wrong, and checks in every thirty seconds. Hitting does not decrease. What went wrong?

The time-out was delivering attention. Explaining, scolding, and checking in are attention, and attention is a reinforcer, so the procedure became a conversation rather than a removal of reinforcement. One brief statement, no lecture, then back to time-in and reinforce the first good behavior.

A classroom token system fines students for every minor infraction, and several students are at zero tokens by mid-morning. Why does the system stop working?

Once a person is at zero there is nothing left to lose, so response cost has no leverage. A room where tokens are constantly taken away also becomes an aversive environment children want to leave, which brings escape-maintained behavior with it. Earnings must outpace losses.

Explain it to a friend. Explain why a time-out only works when the situation it interrupts is rewarding, using an example from a sport or a game rather than from parenting.