Why Habits Are Hard to Change: What Behavioural Science Reveals About Automatic Behaviour

Habits connect familiar situations with learned responses. That can make everyday life easier, but it also helps explain why a sincere decision to change does not always change what happens next. Repetition, context and opportunity matter alongside motivation.

You unlock your phone to check one message and find yourself opening another app. You take the familiar turning despite intending a different route. These small discrepancies between intention and action raise a useful psychological question: what selects a response when we are not deciding everything afresh?

Habit research offers part of the answer. It examines how experience makes particular responses readily available in recurring situations. Understanding that process shifts attention from judging someone's commitment towards examining what has been learned and what their environment keeps inviting them to do.

What psychologists actually mean by habit

A habit is a learned tendency for a cue in a recurring context to generate a relatively automatic response. The cue might be arriving home, finishing lunch or encountering a familiar object. In Gardner's account, the cue generates an impulse to act; that impulse need not determine the eventual behaviour (Gardner, 2015).

A routine is a repeated sequence, which may still require considerable deliberation. Goal-directed action is selected with reference to anticipated consequences and their current value. Someone can prepare the same meal daily because they repeatedly judge it suitable, without every component being habitual.

These influences interact. Choosing to visit a friend may be deliberate while parts of driving there are familiar and automatic. Goals can organise the situations in which habits operate, and habits can support goals rather than obstruct them (Wood & Neal, 2007). Automaticity therefore does not mean an entirely unconscious person or two independent parts of the brain battling for control.

How repetition changes what an action requires

Repeating an action in similar circumstances gives opportunities to learn associations between the situation and the response. Eventually, encountering the cue can bring that response to mind with less fresh deliberation. Frequency matters, but so do contextual consistency, competing responses and what exactly is being repeated (Wood & Rünger, 2016).

In a longitudinal study, 96 volunteers selected an eating, drinking or activity behaviour to repeat in a consistent context for 12 weeks. Self-reported automaticity often increased rapidly at first before gains slowed, although participants differed substantially and not everyone's data followed the model well (Lally et al., 2010).

Complex activities also have separable stages. Starting an exercise session can become habitual while choosing exercises remains deliberate. A systematic review of longitudinal physical-activity studies found this distinction between instigation and execution useful; associations did not settle whether habit caused greater activity or repeated activity strengthened habit (Feil et al., 2021).

Why there is no universal habit-formation deadline

The claim that habits form in exactly 21 days is not supported as a general psychological rule. Nor should it simply be replaced with a promise of 66 days. In Lally and colleagues' study, that median described the modelled time to approach an individual's automaticity plateau among the 39 participants whose data fitted well. Estimates ranged from 18 to 254 days, with longer estimates extending beyond the observation period (Lally et al., 2010).

A 2024 systematic review included 20 studies, but only four directly reported formation times. Their reported medians or means differed considerably; most included studies had high risk of bias. The evidence challenges a fixed deadline while remaining too limited to establish a reliable personal countdown (Singh et al., 2024).

Behavioural complexity, opportunities to repeat an action and individual circumstances can all affect learning. Missing one opportunity did not materially derail formation in the Lally study, but that does not establish that every interruption is inconsequential. Automaticity develops by degrees, rather than appearing on a designated morning.

Why the environment matters

A cue can involve a place, time, preceding action, social setting or object. The relevant context may be a combination: sitting on a particular sofa after dinner with a phone within reach. Its familiarity can make a response available before someone has extensively considered alternatives.

In two field experiments, strong cinema popcorn-eating habits predicted eating that was relatively insensitive to whether popcorn was fresh or stale. This pattern depended on the cinema context and on being able to execute the familiar response with the dominant hand (Neal et al., 2011). These specific findings illustrate context dependence, rather than establishing that people always ignore enjoyment.

Research following students transferring universities likewise found that continuity in aspects of context mattered for whether established behaviours persisted (Wood et al., 2005). Moving or changing jobs can interrupt cues, but some cues travel with the person. A new address is therefore neither necessary nor sufficient to break a habit.

Reinforcement helps explain repetition

Consequences can encourage the repetitions through which habits develop. Positive reinforcement increases behaviour through an added consequence, such as enjoyable social contact following an activity. Negative reinforcement increases behaviour through removing or preventing something unpleasant, such as relief following escape. Negative reinforcement is not punishment: punishment reduces behaviour.

A reward is an outcome someone values; calling it a reinforcer requires evidence that it increases the response. Immediate consequences may sustain repetition while delayed consequences receive less influence in the moment. This does not mean every repeated action was consciously chosen for a reward.

Learning and later performance also need distinguishing. A behaviour initially repeated for an outcome can become readily cued even when that outcome is less valued. Reviews of habit learning examine how reinforcement and context associations contribute to this relative persistence (Wood & Rünger, 2016).

Our discussion of avoidance and anxiety explores how immediate relief can maintain repeated responses. That is one possible learning process, not grounds for labelling every unwanted habit an addiction.

Why sincere intentions can leave behaviour unchanged

A meta-analysis of 47 experimental tests found that interventions producing substantial changes in intentions produced smaller changes in behaviour (Webb & Sheeran, 2006). Motivation can influence action, but changing what someone intends does not automatically rearrange what happens when a familiar cue appears.

Competing goals, interruptions and limited time also matter. Someone may want to walk after work while needing to collect a child, manage pain or complete another shift. Continuing the familiar pattern cannot, by itself, reveal whether commitment was weak.

Established cues can also keep generating the old response after intentions change. Holding a new goal in mind and repeatedly interrupting a familiar action may require attention that is already occupied. This makes changing circumstances and planning relevant, without implying that every intention–behaviour gap is caused by habit.

Does stress make behaviour more habitual?

An influential laboratory experiment found that acute stress increased responding for an outcome that had been made less desirable by eating to satiety. The authors interpreted this as a shift towards habitual control (Schwabe & Wolf, 2009). Such experiments test responses under specified conditions; they do not directly reproduce years of everyday behaviour.

The wider evidence is mixed. In two experiments involving 129 participants, Zwosta and colleagues found no evidence that acute stress increased habitual control in tasks involving gaining rewards or avoiding losses (Zwosta et al., 2025). Task design, stress procedures and individual responses may help explain differing results.

Stress and cognitive demands can make flexible responding harder in some circumstances, but stress does not invariably remove conscious control. Useful routines may also provide continuity during demanding periods. Our article on work stress and burnout considers why demands and recovery opportunities require attention alongside individual responses.

Why an unwanted pattern can keep returning

A new intention may coexist with an old cue–response association. If the environment repeatedly presents that cue and makes the old action easy, it supplies repeated opportunities for the familiar response. Immediate convenience or relief may further support the pattern, while longer-term benefits of changing remain distant.

This helps explain why information and motivational messages can be insufficient. Wood and Neal's review proposes combining support for new responses with disruption of existing cues and patterns (Wood & Neal, 2016). Its intervention framework is useful, but does not establish that one combination works for every behaviour.

What behaviour-change research suggests may help

Changing cues and practical arrangements can reduce opportunities for the old response and make an alternative easier. A field experiment involving 800 participants found an intervention promoting sustainable behaviours was more effective among recent movers. Outcomes were self-reported, and relocation itself was not randomly assigned (Verplanken & Roy, 2016). Context disruption can create an opening without guaranteeing change.

Implementation intentions specify a response to a situation: for example, “After lunch, I will put on my walking shoes.” Gollwitzer and Sheeran's meta-analysis found benefits for goal attainment across varied tasks (Gollwitzer & Sheeran, 2006). This supports planning, not the claim that writing an if–then sentence instantly creates a habit. The cue must occur and the action must be feasible.

A randomised trial comparing routine-based and time-based cues for a nutrition behaviour found increasing automaticity in both groups, without a clear difference between them (Keller et al., 2021). One cue type should not be marketed as universally superior.

Self-monitoring can make patterns and progress more visible. A meta-analysis of experimental studies found that interventions encouraging progress monitoring improved goal attainment on average (Harkin et al., 2016). However, monitoring a behaviour is different from demonstrating automaticity, and detailed tracking can itself demand time.

Alternative responses and manageable repetitions are plausible components of change. Reinforcement may support their repetition; removing support may expose whether they endure. Intervention packages often combine several techniques, making their individual contributions difficult to isolate. Success needs assessment through actual behaviour and sustained outcomes, not adherence to a formula.

Habits in health and psychological care

Habit research is relevant to physical activity, sleep routines and carrying out recommended healthcare behaviours. A familiar cue may help initiate an action without repeated reminders. However, a stable bedtime routine is different from treating insomnia, and frequent checking alone cannot establish obsessive–compulsive disorder.

Clinically significant compulsions, addictions and avoidance difficulties require understanding their function, severity and wider context. Habit replacement is not a general treatment for complex psychological conditions. Research on ordinary habits can inform questions in clinical and health psychology while leaving assessment and treatment claims to evidence concerning those particular problems.

People do not have equal opportunities to change

A stable daily cue is easier to arrange when work, housing and caring responsibilities are predictable. Rotating shifts, insecure employment, disability, pain, inaccessible facilities and limited money can restrict both repetition and alternatives. An apparently simple plan may assume resources a person does not have.

The COM-B framework explicitly considers capability and opportunity alongside motivation when analysing behaviour (Michie et al., 2011). It is a framework for designing and examining interventions, rather than proof that a particular programme will work.

Individual differences in experience, enjoyment, attention and existing routines also matter. Supportive relationships may make a change practical; competing obligations may make the same change costly. A fair account of self-regulation asks what control is available and what support would expand it. Difficulty maintaining an intervention can reveal a mismatch with circumstances rather than a deficit in character.

What habit research still cannot establish

Researchers do not always use “habit” in the same way. The Self-Report Habit Index includes repetition, automaticity and identity-related content (Verplanken & Orbell, 2003). A later four-item automaticity measure narrows that focus (Gardner et al., 2012). Both provide useful reports, but neither directly observes an automatic process occurring.

Frequent goal-directed behaviour can resemble habit, and people may infer automaticity from how often they act. When both habit and behaviour are self-reported, shared reporting tendencies may strengthen associations. Longitudinal designs clarify timing but cannot exclude all confounding or reciprocal effects.

Laboratory tests often infer habit when responding persists after an outcome is devalued. Yet de Wit and colleagues reported five failures to show that extended training produced greater habitual control in humans (de Wit et al., 2018). This challenges straightforward assumptions about induction and measurement; it does not show that everyday habits are imaginary.

Experiments can test causal effects of particular manipulations within their tasks. Transferring those findings to everyday routines, different populations or clinical conditions requires additional evidence. Small samples, selective publication, varied measures and limited follow-up further constrain confidence.

The practical question is therefore broader than whether someone wants change enough: which cues support the old response, what alternative is possible, and can the surrounding conditions sustain it? Habits are learned patterns operating within lives. Understanding both makes behavioural change easier to investigate without turning difficulty into a moral verdict.

About this article: This educational overview does not provide a diagnosis or an individual treatment plan. Everyday examples are illustrative.

References

de Wit, S., Kindt, M., Knot, S. L., Verhoeven, A. A. C., Robbins, T. W., Gasull-Camos, J., Evans, M., Mirza, H., & Gillan, C. M. (2018). Shifting the balance between goals and habits: Five failures in experimental habit induction. Journal of Experimental Psychology: General, 147(7), 1043–1065. https://doi.org/10.1037/xge0000402

Feil, K., Allion, S., Weyland, S., & Jekauc, D. (2021). A systematic review examining the relationship between habit and physical activity behavior in longitudinal studies. Frontiers in Psychology, 12, Article 626750. https://doi.org/10.3389/fpsyg.2021.626750

Gardner, B. (2015). A review and analysis of the use of ‘habit’ in understanding, predicting and influencing health-related behaviour. Health Psychology Review, 9(3), 277–295. https://doi.org/10.1080/17437199.2013.876238

Gardner, B., Abraham, C., Lally, P., & de Bruijn, G.-J. (2012). Towards parsimony in habit measurement: Testing the convergent and predictive validity of an automaticity subscale of the Self-Report Habit Index. International Journal of Behavioral Nutrition and Physical Activity, 9, Article 102. https://doi.org/10.1186/1479-5868-9-102

Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69–119. https://doi.org/10.1016/S0065-2601(06)38002-1

Harkin, B., Webb, T. L., Chang, B. P. I., Prestwich, A., Conner, M., Kellar, I., Benn, Y., & Sheeran, P. (2016). Does monitoring goal progress promote goal attainment? A meta-analysis of the experimental evidence. Psychological Bulletin, 142(2), 198–229. https://doi.org/10.1037/bul0000025

Keller, J., Kwasnicka, D., Klaiber, P., Sichert, L., Lally, P., & Fleig, L. (2021). Habit formation following routine-based versus time-based cue planning: A randomized controlled trial. British Journal of Health Psychology, 26(3), 807–824. https://doi.org/10.1111/bjhp.12504

Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W., & Wardle, J. (2010). How are habits formed: Modelling habit formation in the real world. European Journal of Social Psychology, 40(6), 998–1009. https://doi.org/10.1002/ejsp.674

Michie, S., van Stralen, M. M., & West, R. (2011). The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implementation Science, 6, Article 42. https://doi.org/10.1186/1748-5908-6-42

Neal, D. T., Wood, W., Wu, M., & Kurlander, D. (2011). The pull of the past: When do habits persist despite conflict with motives? Personality and Social Psychology Bulletin, 37(11), 1428–1437. https://doi.org/10.1177/0146167211419863

Schwabe, L., & Wolf, O. T. (2009). Stress prompts habit behavior in humans. The Journal of Neuroscience, 29(22), 7191–7198. https://doi.org/10.1523/JNEUROSCI.0979-09.2009

Singh, B., Murphy, A., Maher, C., & Smith, A. E. (2024). Time to form a habit: A systematic review and meta-analysis of health behaviour habit formation and its determinants. Healthcare, 12(23), Article 2488. https://doi.org/10.3390/healthcare12232488

Verplanken, B., & Orbell, S. (2003). Reflections on past behavior: A self-report index of habit strength. Journal of Applied Social Psychology, 33(6), 1313–1330. https://doi.org/10.1111/j.1559-1816.2003.tb01951.x

Verplanken, B., & Roy, D. (2016). Empowering interventions to promote sustainable lifestyles: Testing the habit discontinuity hypothesis in a field experiment. Journal of Environmental Psychology, 45, 127–134. https://doi.org/10.1016/j.jenvp.2015.11.008

Webb, T. L., & Sheeran, P. (2006). Does changing behavioral intentions engender behavior change? A meta-analysis of the experimental evidence. Psychological Bulletin, 132(2), 249–268. https://doi.org/10.1037/0033-2909.132.2.249

Wood, W., & Neal, D. T. (2007). A new look at habits and the habit-goal interface. Psychological Review, 114(4), 843–863. https://doi.org/10.1037/0033-295X.114.4.843

Wood, W., & Neal, D. T. (2016). Healthy through habit: Interventions for initiating & maintaining health behavior change. Behavioral Science & Policy, 2(1), 71–83. https://doi.org/10.1177/237946151600200109

Wood, W., & Rünger, D. (2016). Psychology of habit. Annual Review of Psychology, 67, 289–314. https://doi.org/10.1146/annurev-psych-122414-033417

Wood, W., Tam, L., & Witt, M. G. (2005). Changing circumstances, disrupting habits. Journal of Personality and Social Psychology, 88(6), 918–933. https://doi.org/10.1037/0022-3514.88.6.918

Zwosta, K., Karcz, T., Möschl, M., Walser, M., Goschke, T., Ruge, H., & Wolfensteller, U. (2025). No evidence for increased habitual or decreased goal-directed action control after acute stress. PLOS ONE, 20(9), Article e0327807. https://doi.org/10.1371/journal.pone.0327807