The Half-Life of a New Year's Resolution by Category

This paper presents findings from a four-year longitudinal study tracking New Year's resolution adherence across 2,400 subjects, establishing category-specific decay curves and identifying factors that predict resolution persistence. Subjects reported resolutions across eight categories and provided weekly adherence updates through a mobile application for twelve months following resolution formation.

Our primary finding is that resolution decay follows predictable patterns that vary dramatically by category. Exercise-related resolutions exhibit the shortest half-life (12 days), meaning that 50% of subjects have abandoned the resolution by day 12, with 90% abandonment by day 36. Financial resolutions show greater persistence (half-life: 23 days), possibly due to the presence of concrete, measurable outcomes.

The category "learn a language" exhibits a statistically unusual bimodal distribution that defies the exponential decay model fitting other categories. Subjects either abandon language learning within 8 days or persist beyond 90 days, with relatively few subjects falling in the intermediate range. We hypothesize that this reflects a threshold effect: those who survive the initial difficulty plateau and experience meaningful communication success become intrinsically motivated, while those who do not reach this threshold quickly disengage.

Secondary analyses reveal that resolution specificity (measured by a 5-point concreteness scale) is the strongest predictor of persistence, exceeding the predictive power of self-reported motivation, prior resolution history, or demographic factors. Resolutions framed as approach goals ("I will exercise three times weekly") show significantly greater persistence than avoidance goals ("I will stop being sedentary"), even when the behavioral target is equivalent.

We conclude with practical recommendations for resolution formation, including optimal specificity levels and the importance of early success experiences.