Essays and worked examples on changing unwanted habits

The Psychology of Quitting

Understand the argument that keeps bringing you back

"The Damage Is Already Done": When Sunk Costs Keep the Habit Going

The past cannot be changed. The next hour can.

The thought has a grim logic to it. You have smoked for twenty years. You have drunk heavily for a decade. You have been sedentary since your twenties. The damage, you tell yourself, is baked in. Stopping now is like closing the stable door after the horse has bolted.

The reasoning feels realistic. It is actually a form of accounting error.

The sunk cost fallacy

In behavioral economics, the sunk cost fallacy is the tendency to continue an activity because of what has already been invested, even when the future costs exceed the future benefits [1] [5]. You have already spent the money on the theater ticket, so you sit through a terrible play. You have already invested three years in the career, so you stay despite being miserable.

Applied to habits: “I’ve already done twenty years of damage” treats past harm as a reason to accept future harm. But the past damage is a sunk cost. It is gone regardless of what you do next. The only question that remains is: what happens from here?

What the evidence says about recovery

For smoking, the data on health recovery after cessation is among the strongest in all of medicine. The US Surgeon General’s report documented the timeline: heart rate and blood pressure begin to drop within hours, lung function improves within weeks to months, and the excess risk of coronary heart disease halves within a year [2].

A large cohort study in the United States found that smokers who quit at any age had substantially lower mortality than those who continued [3]. The British Doctors Study, a fifty-year prospective study, found that those who quit at 30 gained nearly a full decade of life expectancy compared with those who continued, while those who quit at 60 still gained about three years [4].

The damage from smoking is cumulative, and some of it is irreversible. But the risk of additional damage drops continuously from the moment of cessation. “The damage is already done” is true of the past and false of the future.

For alcohol, the picture is similar in direction if less dramatic in magnitude. Reducing heavy drinking reduces the risk of liver disease, several cancers and cardiovascular events [6]. The body’s capacity to recover is not unlimited, but it is greater than “the damage is done” implies.

A fictional example

The following is an illustrative example, not a real person’s story.

Ruth smoked from age sixteen to fifty-two. She has a chronic cough and shortness of breath. When her doctor raises quitting, her response is weary: “What’s the point now? The damage is done.”

  • What she thinks: “Thirty-six years of smoking. My lungs are already ruined.”
  • What she feels: A dull hopelessness, mixed with something that might be relief at not having to try.
  • What the data says: If Ruth stops now, her risk of a heart attack begins to drop within a day. Within five years, her stroke risk approaches that of a never-smoker. Her lung function decline slows. Her risk of lung cancer, while elevated, begins to decrease year on year.

Ruth’s past damage is real. Her future damage is optional.

The accounting error

“The damage is already done” conflates two questions: “Can I undo what happened?” and “Can I prevent what has not happened yet?” The answer to the first is often no, or only partly. The answer to the second is almost always yes.

Every day of continued use adds to the ledger. Every day of cessation stops the additions. The earlier you stop, the more future damage you prevent, but stopping later is always better than not stopping.

The thought also makes an implicit assumption about thresholds: that below a certain level of health, improvement is not worth pursuing. But wellbeing is not binary. Better breathing, better sleep, more energy, fewer sick days, these are gains regardless of what came before. A person who quits at sixty and gains three years of life has gained three years.

A reframe

Instead of “the damage is already done,” try: “the damage so far is done. The damage from tomorrow is not.”

This is not optimism. It is arithmetic. You cannot change the past inputs, but you have complete control over whether you add to them today.

For the thought that sees the habit as permanent because of past failures, see “I’ve tried before and it didn’t work”. For the thought that delays action to a future date, see “I’ll deal with it later”.

When to get more support. If you have been using the thought “the damage is already done” to avoid making a change, your GP can give you a realistic assessment of your current health and what stopping would achieve at your specific stage. You may be surprised by how much recovery is still possible. In the US, SAMHSA’s National Helpline (1-800-662-4357) offers free, confidential referral. In the UK, your GP can refer you to a stop-smoking service or other specialist support. If you are in distress, call or text 988 in the US, or contact Samaritans on 116 123 in the UK and Ireland.

Frequently asked questions

Is the damage from smoking really reversible?

Partly. The risk of heart attack drops within 24 hours of stopping. Lung function begins to improve within weeks. The risk of coronary heart disease halves within a year. Some damage, particularly severe emphysema or established cancer, is not reversible, but the risk of new damage decreases continuously with abstinence.

What is the sunk cost fallacy?

It is the tendency to continue investing in something because of what has already been spent, even when the future costs outweigh the future benefits. In habit terms: 'I've already damaged my health, so there's no point stopping now' treats past damage as a reason to accept future damage.

Does stopping at any age help?

Yes. Research on smoking cessation shows benefits at every age, though the earlier the better. People who quit at 30 gain nearly a full decade of life expectancy. People who quit at 60 still gain years. The benefit is always positive and never zero.

Sources and further reading

  1. Arkes and Blumer — The psychology of sunk cost

    Organizational Behavior and Human Decision Processes, 1985;35(1):124–140. Foundational paper demonstrating the sunk cost effect across multiple experiments.

  2. US Surgeon General — The Health Benefits of Smoking Cessation

    US Department of Health and Human Services, 1990. Report documenting the timeline of health recovery after smoking cessation.

  3. Jha et al. — 21st-century hazards of smoking and benefits of cessation in the United States

    New England Journal of Medicine, 2013;368(4):341–350. Large cohort study showing that cessation at any age reduced mortality. DOI: 10.1056/NEJMsa1211128.

  4. Doll et al. — Mortality in relation to smoking: 50 years' observations on male British doctors

    BMJ, 2004;328(7455):1519. Fifty-year prospective study of British doctors. Quitting at 30 gained nearly 10 years; at 50, about 6 years; at 60, about 3 years. DOI: 10.1136/bmj.38142.554479.AE.

  5. Thaler — Misbehaving: The Making of Behavioral Economics

    W.W. Norton, 2015. Accessible account of behavioral economics concepts including sunk costs, mental accounting and loss aversion.

  6. Rehm et al. — The relationship between different dimensions of alcohol use and the burden of disease: an update

    Addiction, 2017;112(6):968–1001. Updated review of the alcohol-disease relationship. DOI: 10.1111/add.13757.

Sources checked October 16, 2026. This is an editorial summary, not a systematic review or individual advice.

Related articles

"I'll Quit After This Event": The Moving Finish Line

Why 'I'll quit after this event' sounds like a plan but usually functions as a delay, what research on temporal landmarks and implementation intentions shows, and how to stop the finish line from moving.

"I'll Quit When It Gets Serious": The Moving Threshold

Why waiting for the problem to become serious enough before acting is a strategy that protects the habit, what research on normalization and shifting baselines shows, and how to set a fixed threshold before the drift begins.