Abstinence violation effect
After one slip, an all-or-nothing mindset turns a single lapse into a full relapse.
What it means
The abstinence violation effect is the tendency for a single lapse from a self-imposed rule of total abstinence to trigger a cascade of negative reactions that converts a minor slip into a full-blown relapse. When someone who has committed to 'never again' breaks the rule once, they often experience guilt, self-blame, and a sense of lost control, and they may attribute the lapse to stable personal failings rather than to the situation. Because the violation shatters a black-and-white self-image, the person reasons that the goal is already blown, so further indulgence no longer seems to cost anything. The effect explains why a dieter who eats one forbidden cookie may finish the box, or why an abstinent drinker treats one drink as license to binge. It matters for behavior change because rigid all-or-nothing goals are fragile, and reframing a lapse as a recoverable, situational event rather than proof of failure is central to relapse prevention.
Why it happens
Two things break at once. The first is a self-image. Total abstinence fuses the rule to the identity, so 'I don't smoke' is falsified by one cigarette, and the cheaper repair is to revise the identity rather than the rule. The second is the accounting. A rule of never counts streaks, not quantities, so it has no unit of partial failure: one is the same category of event as ten, and the ledger has no way to price the difference. Guilt is supposed to do its own work, because using is the fastest way to stop feeling bad about having used, though the evidence below is less kind to this limb than to the others. Rigidity builds the cliff. A goal stated as a limit rather than a prohibition has no moment at which everything is lost.
What the evidence shows
Curry, Marlatt and Gordon validated the construct in smokers in 1987, showing that attributing a lapse to characterological causes predicted worse outcomes. The sharpest modern test complicates it. Kirchner, Shiffman and Wileyto tracked 203 smokers through 1,001 lapse episodes using electronic diaries and pulled the effect into its parts. The self-efficacy limb held: confidence dropping after a lapse accelerated the next one. Guilt had no consistent effect at all — its direction reversed depending on whether the smoker was on an active patch or a placebo. Self-blame showed no accelerating effect, and after a longer stretch of abstinence it protected against progression, as though owning the lapse prompted corrective action. Responses to the first lapse predicted almost nothing — only the confidence drop carried any signal; the responses to later lapses were what mattered. The effect is real, but it is not one thing.
Limits and caveats
The eating evidence is worse for the strong version. Laboratory work had long shown that restrained eaters overeat after a forced preload, but Tomiyama and colleagues followed dieters through ordinary days and found no such effect. After a violation they ate 1,410 calories against 1,497 on normal days, a gap indistinguishable from noise: total intake did not rise even with the shake's calories counted in, which looks more like compensation than collapse. A paradigm that measures the violation with a bowl you cannot leave may manufacture the effect it measures; when the same preload is followed by ordinary days, the effect disappears. The same gap appears clinically: relapse prevention works overall, yet components aimed specifically at reducing the abstinence violation effect have generally failed to move outcomes, which suggests the effect describes relapse better than it causes it.
Related but distinct
Self-licensing runs the other way. There a person earns permission through a workout or a donation and indulges without guilt, looking forward. This effect is retrospective and painful: the licence comes from damage already done, and shame is the engine rather than its absence. The what-the-hell effect is the wider family, the response to violating any goal, with or without an abstinence rule. What makes this version distinctive is the vow. Only a promise of never creates a state that one act can destroy completely, and only a self-image built on that promise gives the lapse something larger than the behaviour to break.
Using it in practice
Design the goal so a slip is priced honestly. Quantity limits, weekly allowances and floors such as 'never two in a row' leave the goal standing after one deviation, where a streak leaves nothing. Write the lapse plan while the rule still feels easy: what counts as a slip, what happens next, who hears about it. Given the evidence, protect confidence before guilt: restoring the belief that the next situation is survivable predicts more than reducing remorse does. Marlatt's framing of a lapse as a mistake in learning rather than a verdict on the person is the vehicle for that, not an end in itself. A streak counter that resets to zero teaches the cliff.
Examples
A smoker who has quit takes one cigarette at a party, concludes 'I've failed anyway,' and is smoking a pack a day by the next week.
A no-spending month collapses after one takeaway on day nine. The month is ruined anyway, she reasons, and by day twelve she has ordered three more.
Someone on a thirty-day phone detox opens an app once on the train, decides the streak is dead and that he has no willpower, and scrolls for the rest of the evening.
A vegetarian of nine years eats a slice of ham at a party, concludes she was never really one, and orders steak for every meal of the rest of the trip.
A team with a zero-defect release rule ships one bug, decides the quarter's record is already gone, and waves through three more fixes that would previously have blocked the build.
First described in G. Alan Marlatt (relapse prevention model, 1980s).
Key references
- Kirchner, T. R., Shiffman, S., & Wileyto, E. P. (2012). Relapse dynamics during smoking cessation: Recurrent abstinence violation effects and lapse-relapse progression. Journal of Abnormal Psychology, 121(1), 187-197. doi.org/10.1037/a0024451
- Tomiyama, A. J., Moskovich, A., Haltom, K. B., Ju, T., & Mann, T. (2009). Consumption after a diet violation: Disinhibition or compensation? Psychological Science, 20(10), 1275-1281. doi.org/10.1111/j.1467-9280.2009.02436.x
- Witkiewitz, K., & Marlatt, G. A. (2004). Relapse prevention for alcohol and drug problems: That was Zen, this is Tao. American Psychologist, 59(4), 224-235. doi.org/10.1037/0003-066X.59.4.224
- Irvin, J. E., Bowers, C. A., Dunn, M. E., & Wang, M. C. (1999). Efficacy of relapse prevention: A meta-analytic review. Journal of Consulting and Clinical Psychology, 67(4), 563-570. doi.org/10.1037/0022-006X.67.4.563
- Larimer, M. E., Palmer, R. S., & Marlatt, G. A. (1999). Relapse prevention: An overview of Marlatt's cognitive-behavioral model. Alcohol Research & Health, 23(2), 151-160. pmc.ncbi.nlm.nih.gov/articles/PMC6760427/
- Curry, S., Marlatt, G. A., & Gordon, J. R. (1987). Abstinence violation effect: Validation of an attributional construct with smoking cessation. Journal of Consulting and Clinical Psychology, 55(2), 145-149. doi.org/10.1037/0022-006X.55.2.145