Where Resolutions Collapse
On January 1, we promise: 'This year I'll exercise.' 'Quit smoking.' 'Finish the thesis.' By March half are gone; by June, 80%. The usual diagnosis is 'I'm weak-willed.' But NYU social psychologist Peter Gollwitzer's 1999 American Psychologist paper Implementation Intentions: Strong Effects of Simple Plans offers a different one. The problem isn't willpower; it's the form of the plan.
Gollwitzer separated two kinds of intention.
- Goal intention: 'I want to exercise more.' 'I want to lose weight.' — a commitment to an outcome.
- Implementation intention: 'WHEN I leave the office at 6pm on Mon/Wed/Fri, THEN I go straight to the basement gym.' — an if-then link from situation to action.
The first stays at 'I want to'; the second moves to 'already decided.'
How If-Then Works
Gollwitzer argues if-then plans activate two mechanisms. First, attention to the situational cue (if) becomes automatically heightened. 6pm leaving the office stops being 'just clocking out' and becomes 'workout signal.' Second, once the cue is detected, the response (then) fires near-automatically. No fresh deliberation each time.
Shifting behavior from willpower territory to stimulus-response territory — converting conscious intention into an 'automatic-like' response — is Gollwitzer's central claim. Habit researchers Wood & Neal (2007) call implementation intentions a 'bridge between conscious goals and automatic habits.' Repeat the cue-action chain enough times and the cue itself retrieves the behavior.
What the Meta-Analysis Shows — d=0.65
In 2006 Gollwitzer & Sheeran published a meta-analysis in Advances in Experimental Social Psychology covering 94 studies, n=8,461. Result: groups forming implementation intentions outperformed goal-only groups by d=0.65 — a medium-to-large effect, unusually robust for a psychological intervention.
Applications span:
- Medication adherence (Sheeran 2005): linking pill-taking to time and place raises compliance.
- Studying (Oettingen 2000): 'After dinner, I sit at my desk and solve problems for 30 minutes.'
- Diet and exercise initiation: Bélanger-Gravel 2013 meta (~26 studies) shows small-to-medium effects in physical activity.
- Voting: Nickerson & Rogers 2010 Psychological Science field experiments — asking voters when, where, and how they would vote significantly raised turnout vs. simply asking if they would vote.
- Hand hygiene: a 2018 Korean Society of Infectious Diseases report on hospital workers found if-then hand-hygiene plans improved compliance.
Hagger & Luszczynska 2014 (Appl Psychol Health Well Being) concluded implementation intentions enhance a wide range of health behaviors.
Comparing Three 'Plans'
| Form | Definition | Example | Evidence |
|---|---|---|---|
| Goal intention | Statement of desired outcome | 'I want to exercise more' | Weak — wanting doesn't guarantee acting |
| Implementation intention (if-then) | Cue–response linkage | 'When I leave at 6pm Mon/Wed/Fri, I go straight to the gym' | Gollwitzer & Sheeran 2006 meta d=0.65 |
| WOOP / MCII | Wish–Outcome–Obstacle–Plan (mental contrasting + if-then) | Wish: lose weight / Outcome: health / Obstacle: dinner parties / Plan: 'When a dinner is scheduled, I eat first and arrive after drinks start' | Oettingen 2014 — outperforms simple positive thinking |
WOOP comes from NYU psychologist Gabriele Oettingen, Gollwitzer's spouse. In Rethinking Positive Thinking (2014) she reports a counterintuitive finding: mere positive thinking can backfire. Oettingen & Wadden 1991 found dieters with the most positive fantasies lost less weight after a year — the fantasy consumes the reward in advance. WOOP therefore forces 'mental contrasting' between wish and obstacle, then closes with an if-then plan.
In Korean Workplaces, Schools, Clinics
Korean research has accumulated:
- Jung Hye-won (2010) Korean Journal of Psychology — implementation intentions narrowed the intention-behavior gap for exercise.
- Kim Min-ji (2013) — Korean office workers with if-then plans showed greater 12-week behavior change in exercise and diet.
- Cho Yu-mi (2015) — among Korean undergraduates, plans like 'When I sit down at the library at 7pm, I put my phone in my bag and focus for one hour' increased study time.
- Korean Society of Infectious Diseases (2018) — if-then hand-hygiene plans improved compliance among healthcare workers.
Korean phrasing examples:
- 'When the alarm rings, I throw off the covers immediately.'
- 'When I sit back down after lunch, I drink a glass of water first.'
- 'When I drop my kid at daycare, I walk the 30-minute route home.'
Three rules: cue must be concrete (time, place, prior action); response must be immediately doable; ideally no deliberation between cue and action.
Where the Effect Weakens — Limits and Critique
Implementation intentions are not magic. Boundary conditions:
- Effects shrink sharply when goal commitment is weak. If you don't really want to exercise, if-then is an empty sentence.
- Easy goals show small added benefit; effects are largest for difficult goals.
- High-temptation environments can swamp cues. Drinking parties and addictive contexts need additional safeguards.
- Some smoking-cessation trials report modest or inconsistent effects — addiction is more than an intention-behavior gap. Effect sizes are heterogeneous across goals and individuals.
Distinguishing from Affirmations and Manifestation
If-then plans are often confused with two popular self-help genres, but the scientific standing differs sharply:
- Positive affirmations: repeating 'I can do it.' Wood (2009) found affirmations make people with low self-esteem feel worse. Affirmations contain no cue-action link.
- Law of attraction / manifestation: 'Want it hard enough and the universe delivers.' No peer-reviewed evidence. Per Oettingen, fantasy can actually deplete reward circuits and weaken action.
Implementation intentions answer not 'what do I believe I can do' but 'when X happens, what comes next.' Not a creed — an algorithm.
Conclusion: Less Willpower, More Structure
The consolation Gollwitzer offers: you didn't lack willpower — you lacked a plan. Convert vague resolutions into one-line 'When X, then Y' sentences. Write them on paper, in a note app, on a sticky.
The meta-analytic d=0.65 isn't a statistic; it's the gap between someone who doesn't have to decide what to do tonight and someone who collapses in the same spot again. Don't increase the amount of willpower — reduce its frequency of use. Well-designed cues will decide for you, day after day.
