Punishment can be compelling because it often appears to work. A child becomes quiet after losing a privilege. A student stops disrupting a lesson after receiving a consequence. An adult changes what they are doing after being reprimanded. From the outside, the sequence seems straightforward: an unwanted behaviour occurred, a consequence followed, and the behaviour stopped.
But stopping a behaviour in the moment is not necessarily the same as changing it. This distinction has existed within psychology for decades. In Science and Human Behaviour, B. F. Skinner (1953) acknowledged that punishment can produce an immediate reduction in behaviour, while questioning its effectiveness as a way of creating lasting behavioural change. the important question, then, is not simply “Did the behaviour stop?” But “What has the person actually learned?”
What do psychologists mean by punishment?
In everyday language, punishment tends to mean being told off, receiving a sanction or experiencing a consequence for doing something wrong. In behavioural psychology, however, the term is more specific. Punishment refers to a consequence following a behaviour that reduces the likelihood of that behaviour occurring again. Contemporary behaviour analysis distinguishes between positive punishment, where something aversive is added following a behaviour, and negative punishment, where something valued is removed (Cooper, Heron & Howard,2020).
Importantly, positive and negative do not mean good and bad. They refer simply to whether something is added or taken away. A reprimand, for example, may function as positive punishment if it reduces the behaviour tat preceded it, while losing access to a privilege may function as negative punishment if that behaviour subsequently becomes less frequent.
This definition also exposes an important limitation. A consequence cannot simply be assumed to be a punishment because an adult intended it to be a punishment because an adult intended it to be one. Its effect has to be considered. If the same behaviour continues – or even increases – then repeatedly applying a consequence and calling it a punishment tells us very little about whether it is actually changing behaviour.
Stopping behaviour is not the same as teaching behaviour
One difficulty with relying heavily on punishment is that it tells a person what not do to without necessarily teaching them what they can do instead. Removing an unwanted behaviour can leave an important gap: if that behaviour was serving a purpose – gaining attention, escaping a difficult situation, communicating frustration or meeting another need – the person may still need another way of achieving the same outcome.
This is one reason reinforcement plays such an important role in behavioural interventions. Rather than focusing only on reducing an unwanted behaviour, approaches such as differential reinforcement of alternative behaviour aim to strengthen a more appropriate behaviour that can take its place. A review by Petscher, Rey and Bailey (2009), covering 116 empirical studies, found substantial evidence supporting differential reinforcement across a range of behavioural difficulties.
The same principle can be seen in Alan Kazdin’s work on parent management training. Kazdin (1997) reviewed a large body of controlled research showing that programmes which teach caregivers to identify, prompt and reinforce desired behaviours can reduce oppositional, aggressive and antisocial behaviour in children. Crucially, he emphasis is not simply on making unwanted behaviour costly.It is also on actively teaching and strengthening the behaviour that we want to see instead.
This changes the questions from “What consequence should follow this behaviour?” To “What does this person need to learn or be able to do differently next time?” Punishment may interrupt a behaviour; teaching an alternative gives the person somewhere else to go.
When stopping looks like changing
One reason punishment can feel so effective is that its impact may be immediate. A reprimand can stop a conversation. A sanction can end disruption. The person administering the consequence sees the behaviour disappear and, understandably, concludes that the intervention has worked.
Psychological research gives us a more complicated picture. Reviews of punishment research show that punitive consequences can produce substantial reductions in behaviour, but their effects depend on factors such as the nature, timing and consistency of consequence, the circumstances in which it occurs and what continues to reinforce the behaviour (Lerman and Vorndran, 2002). In other words, the question is not whether punishment is capable of changing behaviour, but whether the change is durable, generalises beyond the situation in which it occurs and teaches anything useful in its place.
Kazdin has similarly distinguished between suppressing an unwanted behaviour in the moment and developing a lasting behavioural pattern. A person may learn, “I should not do this when this person is present”, or “I need to avoid being caught”, rather than learning how to manage frustration, communicate a need or respond differently when the same situation arises again.
This matters because compliance can easily be mistaken for learning. If behaviour disappears while the threat of a consequence is present but returns when the consequence is removed, we may have successfully controlled the immediate situation without necessarily changing the skills, motivations or circumstances that contributed to the behaviour in the first place.
What is the behaviour doing?
A more useful way of understanding behaviour is sometimes to ask about its function. Rather than looking only at what a behaviour looks like, functional approaches consider what tends to happen before it, what follows it, and what outcome the behaviour may be helping the person achieve or avoid.
This idea has a long history within applied behaviour analysis. Carr and Durand (1985), for example, examined situations in which behaviours such as aggression and tantrums were more likely to occur. They found that factor including difficult tasks and reduced adult attention could be important, and then taught children alternative ways of requesting help or attention. When those alternative communication behaviours served the same function, the problem behaviours decreased.
This is important because the same outward behaviour can serve very different purposes. Refusing a task might help one person escape something they find overwhelming. For someone else, similar behaviour might reliably bring interaction or support. Looking only at the behaviour itself risks treating these situations as though they are identical.
A consequence may therefore fail, not because it was insufficiently severe, but because it never addressed what was maintaining the behaviour. If someone has no workable alternative for communicating “I need help”, or “this is too difficult”, or “I need a break”, making the original behaviour more costly does not automatically give them another way to meet that need.
Understanding function does not mean approving of every behaviour. It means recognising that effective behaviour change usually requires us to understand why this behaviour, in this situation, for this person – rather than responding only to what is easiest to see.
When someone knows better but cannot do better
There is another reason that consequences alone may fail to change behaviour: sometimes the problem is not that a person does not understand what s expected of them. In moments of heightened stress, the cognitive abilities needed to meet those expectations can become harder to access.
Research on the prefrontal cortex is particularly relevant here. The prefrontal cortex supports higher-order processes including attention, working memory, behavioural inhibition and the regulation of emotion. Arnsten (2009) reviewed evidence showing that even relatively mild, uncontrollable acute stress can impair prefrontal functioning. This means that the very abilities we often expect someone to use when they are distressed – pausing, thinking ahead, remembering instructions and inhibiting an immediate response – may become less effective at precisely the moment they are most needed.
Research on childhood self-regulation points in a similar direction. Blair (2010) described how stress can affect the neural systems underlying executive function and self-regulation, while Blair and Raver (2015) emphasised the importance of self-regulation for children’s ability to engage with learning and adapt successfully within school environments.
This creates an important distinction between knowing what to do and being able to do it in the moment. Repeating a rule or increasing a consequence may add little if the difficulty lies not in understanding the expectation, but in accessing he regulation and executive function skills required to meet it.
None of this means that distress removes responsibility or that expectations should disappear. It means that behaviour change may sometimes require more than motivation. If the difficulty involves regulation, overwhelm or reduced executive control, then supporting the person to regulate, practice alternative responses and develop the necessary skills may be more useful than assuming that a stronger consequence will produce better behaviour.
When punishment creates unintended learning
Consequences do not only affect whether a behaviour happens again; they can also shape what a person learns about the situation, the person delivering the consequence and how best to avoid experiencing it again. Sometimes the lesson learned may be the one we intended. At other times, it may be “do not do this when someone is watching,” “do not get caught,” or simply “avoid the person or situation associated with the consequence”.
This possibility fits with both behavioural and social-learning accounts of behaviour. Skinner’s work on punishment emphasised that suppressing behaviour can be accompanied by escape and avoidance, while Bandura’s social learning theory highlighted that people also learn through observing others behave and how conflict, authority and aggression are modelled around then (Bandura, 1977). Behaviour change therefore takes place within a wider social context, not simply through sequence of rule-breaking and consequence.
Physical punishment provides one particularly well-researched example of why this distinction matters. Gershoff and Grogan-Kaylor’s (2016) meta-analysis examined 111 effect sizes involving more than 160,000 children. Spanking was associated with increased risk of a range of detrimental outcomes, including aggression, anti-social behaviour and poorer parent-child relationships, while the review found no evidence that spanking was associated with improved child behaviour over time.
It is important not to generalise findings about physical punishment to every kind of consequence. Losing screen time is not psychologically equivalent to being hit, and a calm, proportionate consequence is not the same as an intimidating or humiliating response. The broader point is that what a person learns from a consequence cannot be assumed from what the person administering it intended.
If our aim is lasting change, it is therefore worth asking not only whether the behaviour stopped, but what the experience may have taught. Did the person learn a more useful way to respond? Did they understand the impact of what happened? Or did they simply learn how to avoid the consequence next time?
So what does effective behaviour change look like?
If punishment alone is not always enough, the alternative is not simply to remove boundaries or consequences. Effective behaviour change is more likely to involve understanding what is maintaining the behaviour, teaching an alternative response, reinforcing that response when it occurs, and ensuring that the person has the skills needed to use it.
This is reflected across several strands of behavioural research. Kazdin’s work on parent management training, for example, emphasises the active teaching and reinforcement of desired behaviour rather than than relying primarily on punishment (Kazdin, 1997). Carr and Durand’s (1985) research demonstrated that when children were taught a more appropriate way to communicate the same need that has previously been expressed through challenging behaviour, that behaviour could decrease. Reviews of punishment research likewise suggest that consequences are most useful when considered as part of a broader behavioural intervention rather than as an isolated response (Lerman and Vorndran, 2002).
In practice, this might mean helping someone recognise the early signs of frustration before they reach the point of losing control; teaching them how to request help, space or break; changing an environment that repeatedly overwhelms them; reinforcing successful attempts to use an alternative response; and, where appropriate, using clear and proportionate consequences when boundaries are crossed.
The balance here matters. A psychologically informed response does not require us to choose between understanding and accountability. We can acknowledge that a behaviour caused harm while remaining curious about why it occurred. We can maintain a boundary while asking whether the person had the skills, regulation and support required to meet it. And we can use consequences while recognising that consequences alone may not teach the behaviour we hope to see next time.
Ultimately, lasting change is les about making an unwanted behaviour sufficiently unpleasant and more about making a different response possible, meaningful and increasingly available.
References & Further Reading
Arnsten, A. F. T. (2009). Stress signalling pathways that impair prefrontal cortex structure and function. Nature Reviews Neuroscience, 10, 410–422.
Bandura, A. (1977). Social Learning Theory. Prentice Hall.
Blair, C. (2010). Stress and the development of self-regulation in context. Child Development Perspectives, 4(3), 181–188.
Blair, C., & Raver, C. C. (2015). School readiness and self-regulation: A developmental psychobiological approach. Annual Review of Psychology, 66, 711–731.
Carr, E. G., & Durand, V. M. (1985). Reducing behavior problems through functional communication training. Journal of Applied Behavior Analysis, 18(2), 111–126.
Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied Behavior Analysis (3rd ed.). Pearson.
Gershoff, E. T., & Grogan-Kaylor, A. (2016). Spanking and child outcomes: Old controversies and new meta-analyses. Journal of Family Psychology, 30(4), 453–469.
Kazdin, A. E. (1997). Parent management training: Evidence, outcomes, and issues. Journal of the American Academy of Child & Adolescent Psychiatry, 36(10), 1349–1356.
Lerman, D. C., & Vorndran, C. M. (2002). On the status of knowledge for using punishment: Implications for treating behavior disorders. Journal of Applied Behavior Analysis, 35(4), 431–464.
Petscher, E. S., Rey, C., & Bailey, J. S. (2009). A review of empirical support for differential reinforcement of alternative behavior. Research in Developmental Disabilities, 30(3), 409–425.
Skinner, B. F. (1953). Science and Human Behavior. Macmillan.
Leave a Reply