The Forever Fallacy: Why Gym Members Really Cancel
Most gym memberships end without a complaint, a conversation, or anything that would show up in a feedback form. What finishes them off is a single thought.
It usually arrives mid-session, somewhere between the second set of lunges and the bit where you pretend to stretch but you’re actually just standing near the water fountain hoping nobody talks to you. The thought is quiet, heavy, and feels completely rational: I can’t do this forever.
The forever fallacy completes its work in a matter of seconds. The member doesn’t assess the experience or weigh up anything resembling a rational case for quitting. They’ve projected how they feel right now across every future visit they can imagine, and concluded the whole thing is unsurvivable. A newly published clinical paper has given this specific cognitive distortion a name, and a field experiment at UC Berkeley measured it happening in real time, at a gym, with real members. The results should concern anyone in the business of keeping people moving.
TL;DR
The forever fallacy is a cognitive distortion where people treat their current emotional state as permanent, projecting present discomfort across an imagined future as if nothing will ever change. A 2026 clinical paper proposes it as a distinct thinking error. A field experiment on gym attendance found that members showed 90% projection bias: they could not predict that their relationship with exercise would change, even when it demonstrably did. Members don’t cancel because the gym is bad. They cancel because they can’t imagine it getting better.
The Distortion Nobody Named
Cognitive behavioural therapy has a taxonomy of thinking errors that hasn’t changed much since the 1980s, a roster that includes catastrophising, fortune telling, and all-or-nothing thinking among its usual suspects. Clinicians have been working from essentially the same list for four decades, which is a bit like a mechanic diagnosing every modern car with the same manual that came with a 1983 Ford Cortina.
Christopher Lomas, a clinician at a rehabilitation facility in Cheshire, noticed a pattern in his patients that didn’t quite fit any of the existing categories. They weren’t predicting disaster, and they weren’t jumping to conclusions about specific negative outcomes. What they were doing was subtler and, in a way, more paralysing than either of those: they were treating their current state as if it would never end.
A 24-year-old professional sat in his office, overwhelmed by the prospect of doing the same job for 40 years. The job itself was fine and the workload was manageable. What had collapsed was the patient’s ability to imagine any version of the future that looked different from right now. Forty years of this feeling, this capacity, this level of exhaustion, stretching out identically ahead of him. The weight pressing down had nothing to do with the work and everything to do with the permanence.
Lomas published his proposal in The Cognitive Behaviour Therapist in 2026, through Cambridge University Press, arguing that this pattern deserves its own classification. He called it the forever fallacy.
The Forever Fallacy
A cognitive distortion in which a person projects their current discomfort, limitations, or emotional state across an imagined future timeline as if it were permanent and unchanging. The distortion involves three components: temporal fusion (inability to distinguish present from future experience), stability projection (assuming current resources and feelings won’t change), and affective permanence (believing the discomfort will last indefinitely). First proposed by Lomas (2026).
Why We’re Bad at Time Travel
The forever fallacy draws on a well-established body of research in affective forecasting, which is the study of how accurately people predict their future emotional states. People are, by virtually every measure, terrible at it.
Timothy Wilson and Daniel Gilbert’s work, spanning decades, established what they called the impact bias. People systematically overestimate how intensely and how long they will feel the effects of future events. A breakup feels like it will destroy you, and a pay rise feels like it will fix everything, but in both cases people adapt far faster than they expect. The emotional reality of six months from now looks almost nothing like the emotional forecast made today.
The reason comes down to what Wilson and Gilbert call “immune neglect.” People forget they have a psychological immune system, a built-in capacity to adapt, find workarounds, and eventually get bored of being miserable. When someone projects their current feeling into the future, whether that’s next week or five years from now, they project a static version of themselves into a world that will keep changing around a version of them that will also keep changing. The forecast is wrong because the forecaster doesn’t believe they’ll be any different by the time the future arrives.
Construal level theory adds another layer. Distant future events get processed abstractly, without detail or context. But when those events are emotionally charged, the abstraction collapses and the future gets experienced as if it were happening now. A vague, shapeless dread about decades of gym sessions doesn’t stay vague. It lands in the body as though every one of those sessions is already underway, simultaneously, right now.
Projection Bias
The tendency to assume that your future preferences, tastes, and emotional states will closely resemble your current ones. People who are hungry overestimate how hungry they’ll be later. People who are tired assume they’ll always be this tired. In a fitness context, a member who finds exercise unpleasant today cannot imagine a future version of themselves who finds it tolerable, let alone enjoyable. First formalised by Loewenstein, O’Donoghue, and Rabin (2003).
The Gym Experiment
In 2015, Dan Acland and Matthew Levy published a field experiment in Management Science that measured projection bias in the place it matters most to this audience: a university gym.
The study paid participants to attend a gym regularly for one month, then tracked their attendance after the incentive was removed. Before and after the intervention, participants were asked to predict how often they’d attend the gym in future weeks.
Two findings matter here.
First, roughly a third of participants formed a genuine short-term habit from the incentivised month. Their post-intervention attendance increased significantly, equivalent to the behavioural effect of receiving a $2.60 subsidy per visit. The habit was real. People who were nudged past the initial resistance found that exercise became easier, more tolerable, and more automatic.
Second, the participants could not see this coming. Their pre-intervention predictions showed 90% projection bias over habit formation. In practical terms, that means people who weren’t yet in the habit of going to the gym could not imagine becoming habitual gym-goers. They looked at their current relationship with exercise, assumed it was permanent, and forecast accordingly. The habit that actually formed was invisible to them before it existed.
Acland and Levy also found that participants substantially overpredicted their future attendance based on current motivation, a pattern they attributed to naivety about present bias. Members felt motivated at the start and assumed that motivation would persist. When it didn’t, they assumed the problem was permanent too.
The Parallel You Already Know
Every gym owner has seen this play out without knowing what to call it.
The January joiner who quits in March is reacting to a projection, not a bad experience. Three months in, the novelty has worn off, the sessions feel harder than they expected, and the thought arrives: this is what it’s going to be like from now on. What follows isn’t an evaluation of the three months that just happened. It’s an evaluation of the next thirty years of months that haven’t happened, conducted entirely with the feelings available right now, which are the only ones the member has access to.
A member who signs a 12-month contract and immediately feels trapped is experiencing exactly the same mechanism Lomas described in his newly-wed patient. The man’s partner had always clicked his neck and hummed. For years of dating, it was background noise, and the moment marriage framed it as permanent, it became intolerable. Nothing about the behaviour had changed, only how long he expected to endure it.
A casual gym-goer who trains three times a week on a no-lock-in membership might be perfectly content. Offer the same person a 12-month agreement at a better weekly rate and they may suddenly feel the weight of 156 sessions they haven’t yet attended, because the economics improved while the psychology of pricing got worse.
Acland and Levy’s research puts numbers to what the industry has always felt intuitively but never quite articulated. Members cannot predict their own capacity to change, and they have no way of foreseeing that the session they’re dreading today will feel routine in six weeks. The forever fallacy locks them into a present tense with no exit, and the only rational response to a feeling that never ends is to make it stop.
Research Summary
| Concept | Mechanism | Fitness Parallel |
|---|---|---|
| Forever Fallacy (Lomas, 2026) | Current emotional state projected across future timeline as permanent | Member assumes today’s discomfort represents every future session |
| Impact Bias (Wilson & Gilbert, 2003) | Overestimation of intensity and duration of future emotional reactions | New member believes the difficulty of early sessions will never diminish |
| Projection Bias (Acland & Levy, 2015) | Current preferences assumed to persist; 90% projection bias measured in gym context | Non-habitual exercisers cannot imagine becoming habitual, even when habit formation is demonstrated |
| Immune Neglect (Gilbert & Wilson, 2007) | Failure to account for psychological adaptation | Members forget they adapted to previous challenges (new jobs, relationships, routines) |
| Temporal Fusion (Lomas, 2026) | Collapse of distinction between present experience and imagined future | 12-month contract feels like all 12 months are happening simultaneously at sign-up |
The Nuance
An uncomfortable complication sits inside all of this, and it cuts against the grain of most retention advice.
If members can’t predict their own capacity to change, then asking them to commit to long-term goals might be reinforcing the very distortion that drives them away. A 12-week transformation challenge sounds like a solution. For a member already in the grip of temporal fusion, it may sound like a sentence.
Lomas is careful to note that the forever fallacy is not simply pessimism with a new label. A pessimist expects bad outcomes, while someone experiencing the forever fallacy isn’t necessarily expecting things to get worse. They’re expecting things to stay exactly the same, which is a different kind of trap entirely. A member might enjoy a Tuesday evening class and still cancel their membership because they cannot imagine themselves enjoying it on an indefinite number of future Tuesdays. Both the enjoyment and the inability to project it forward are completely real.
Acland and Levy’s data adds a second complication. The habits that formed were short-lived. A semester break was enough to erode most of the gains. The projection bias was wrong about habit formation, but it was accidentally right about habit fragility. Members couldn’t see the habit coming, and the habit, once formed, didn’t survive its first interruption. The forever fallacy is a distortion, but the future it fears is sometimes not entirely fiction.
The Question
If your members are cancelling a future that hasn’t happened yet, based on a feeling that won’t last, using a version of themselves that is already changing, at what point does a retention strategy need to address how people think about time rather than how they think about your gym?
People Also Ask
What is the forever fallacy in fitness?
The forever fallacy is a cognitive distortion where a person treats their current emotional or physical state as permanent. In a fitness context, it occurs when a gym member assumes that how they feel about exercise right now is how they will feel about exercise indefinitely. This can drive cancellations even when the member’s actual experience has been positive, because the imagined future feels unbearable regardless of what the present evidence suggests. The term was proposed by Lomas (2026) in The Cognitive Behaviour Therapist.
Why do gym members overpredict how often they’ll attend?
Research by Acland and Levy (2015) found that gym members show significant naivety about their own present bias. They feel motivated at the point of sign-up and assume that motivation will persist. When it doesn’t, they can’t separate the temporary dip from a permanent state. This is compounded by projection bias: members project their current preferences onto the future, unable to account for how those preferences will shift as habits form or circumstances change.
How does affective forecasting affect gym member retention?
Affective forecasting is how people predict their future feelings. Research consistently shows that people overestimate the duration and intensity of emotional reactions, a pattern called the impact bias. For gym retention, this means a member who has a difficult week of training may forecast that difficulty continuing indefinitely, leading them to cancel even though their actual experience would likely improve. Wilson and Gilbert’s research suggests people underestimate their own ability to adapt, a phenomenon called immune neglect.
Nobody ever cancelled a gym membership because of today. They cancelled because of every tomorrow they could see from here, and none of them looked any different.
References
7. Burns, D. D. (1980). Feeling Good: The New Mood Therapy. William Morrow and Company.
If any of this rings a bell with your own members, I’d be keen to hear about it. ray.smith@fitnessisbs.com
Discover more from Fitness is BS.
Subscribe to get the latest posts sent to your email.
