For most of the twentieth century, the dominant cultural model of behavior change was willpower. You wanted to stop drinking, you exercised the willpower muscle, and if you were strong enough you succeeded. If you failed, you had insufficient willpower.
This model has not survived scrutiny. Roy Baumeister's original ego-depletion experiments, which were taken as evidence that willpower was a finite resource, have largely failed to replicate. The broader cultural framework that treats willpower as the primary mechanism of behavior change has been quietly retired by most working researchers in the field.
What follows is the current best understanding of why willpower-based approaches tend to fail, and what predicts cessation success better.
What the original studies found, and didn't
Baumeister's famous "radish experiment" in 1998 had participants either eat radishes or resist eating chocolate, then attempt a difficult puzzle. The chocolate-resisters gave up on the puzzle faster, which was interpreted as evidence that exerting self-control depleted a finite resource available for later self-control.
For two decades this framework was taken as established science. It produced popular books, productivity advice, recommendations about glucose levels and decision fatigue, and a generation of self-improvement methodology built on the idea that you should spend your willpower carefully because it ran out.
The framework began to crack in the twenty-tens when large multi-lab replications failed to find the effect, or found effects so small they were within noise. The 2016 "Many Labs" replication of ego depletion found essentially zero effect. The current consensus among most behavioral researchers is that willpower is not a depleted resource in the way the original studies suggested, and that the model itself was probably wrong.
This does not mean self-control does not exist. It means that "exert willpower until the urge passes" is not a viable long-term strategy because the mechanism on which it was based does not work the way the popular model implied.
What actually predicts cessation success
The research on what predicts successful cessation across various behaviors (smoking, drinking, drug use, weight loss) consistently identifies several factors that matter more than self-reported willpower or self-control.
Environmental design. People who remove the cue from their environment do better than people who keep the cue and resist it. Not having alcohol in the house is more effective than having it in the house and not drinking it. This is not about weakness; it is about not requiring repeated decisions where each repetition is an opportunity for failure. Decision-once-design-out beats decision-many-times-resist.
Substitute behaviors. People who replace the drinking with another behavior, even an arbitrary one, do better than people who simply remove the drinking. The brain expects something at the time the drinking would have happened. Giving it a different something (a walk, a tea, a phone call, a small ritual) satisfies the slot. Leaving the slot empty produces more pressure on the empty slot.
Identity reframe. "I don't drink" outperforms "I'm trying not to drink" in nearly every comparison. The identity statement is closing the question. The trying statement is keeping it open. People who think of themselves as non-drinkers, even very early in cessation, do better than people who think of themselves as drinkers who are abstaining.
Social context. People who tell at least a few trusted others about their cessation do better than people who keep it private. The accountability mechanism does some real work, but the larger effect is just that you cannot drink secretly when other people know you are not drinking. The option closes.
Time, simple time. The single best predictor of remaining sober at year two is being sober at year one. The single best predictor of being sober at year one is being sober at ninety days. The single best predictor of being sober at ninety days is being sober at thirty days. Each milestone passed makes the next milestone more likely. The mechanism is partly habit formation, partly identity consolidation, partly the simple fact that cravings are largest in the early period and become smaller over time.
Why willpower frameworks fail
If willpower-based approaches sometimes work but often fail, what is going on?
The most likely explanation is that what willpower-based approaches actually do is the same thing as environmental design and substitute behaviors, but inefficiently. Someone who "exerts willpower" by not buying wine is doing environmental design. Someone who "exerts willpower" by drinking sparkling water instead is doing substitution. The willpower is the description; the mechanism is the practical change.
Pure willpower without environmental change or substitution typically fails because it requires a stable supply of effortful self-control across many situations. The cue is still there. The drink is still on the table. The expectation that you will drink is still in the room. You have to repeatedly refuse, and each refusal has some failure probability. Over months, the failure probability accumulates.
The reason "I just decided to stop drinking" works for some people is not that they have more willpower than the people for whom it fails. It is that the decision was accompanied by other changes (environment, identity, social context) that are doing the actual work. The decision is the framing; the practical changes are the mechanism.
What this implies practically
Stop framing cessation as a willpower contest. You are not stronger or weaker than the people who succeed or fail. The framing itself is wrong.
Frame cessation as a design problem instead. What environment will you live in? What substitute behaviors will fill the slot? How will your identity describe itself? Who in your life knows? What is your default answer when offered a drink?
These questions are answerable. The answers compound. Six months of well-designed substitution becomes a year, becomes two years. The willpower framing makes every day a small fight. The design framing makes most days easy and the hard moments rare.
The shift in framing is not a self-help trick. It is what the current research actually says.