The instruction to “do what makes you happy” sounds harmless until happiness becomes another performance target. Then an ordinary flat or difficult mood can feel like evidence that something is wrong, and changing the feeling becomes work of its own.
A 2025 paper offers one explanation for why that effort can backfire. Across four studies, people who pursued happiness more intensely reported worse self-regulation, persisted for less time on mundane or difficult tasks and, after a happiness cue, ate more freely available chocolate. The authors argue that active emotional management draws on resources also used for self-control.
This is one paper containing four studies, not settled consensus. Its experiments measured short-term behavior and persistence. They did not measure a literal supply of mental energy, follow participants’ happiness over months or show that every deliberate attempt to feel better carries the same cost.
The Artful Age
A weekly letter on aging well, family across generations, and the creative life after the kids leave home.
The study defined happiness-seeking quite narrowly
Aekyoung Kim and University of Toronto professor Sam Maglio published Happiness depletes me: Seeking happiness impairs limited resources and self-regulation in the 2025 volume of Applied Psychology: Health and Well-Being. The paper appeared online on January 30, 2025.
For this research, seeking happiness meant trying to experience as much positive emotion, and as little negative emotion, as possible. That is not identical to valuing a good life, choosing work one finds meaningful, enjoying a friend or making time for an absorbing hobby. The proposed cost lies in monitoring how one feels and deliberately trying to move the feeling toward a preferred state.
This distinction is central to the apparent paradox. Happiness as an experience does not require the same mental stance as happiness as a goal under constant evaluation. One can be absorbed in dinner with a friend; it is another task to keep asking whether the dinner is producing enough happiness.
The first two studies established a correlation
Study 1 surveyed 532 adults about how strongly they sought happiness and how well they believed they regulated their behavior. Greater happiness-seeking was associated with worse self-reported self-control in daily life.
Study 2 moved from self-description to behavior. A sample of 369 participants ranked consumer products in order of preference. People who scored higher on habitual happiness-seeking spent less time persisting with the ranking task.
These two studies fit the theory, but neither can establish direction. People who already struggle with self-control may monitor their mood more closely. Stress, perfectionism, anxiety or another unmeasured characteristic could contribute to both. The researchers therefore followed the correlations with experiments designed to manipulate the pursuit of happiness.
The experiments used chocolate and anagrams as proxies
Study 3 was small. Thirty-six people recruited at a university library saw either an advertisement prominently using the word “happiness” or a neutral advertisement. They were then given a bowl of chocolates, invited to eat as many as they wanted and asked to rate the taste.
The happiness-cued participants ate more chocolates than the control group. The researchers interpreted this as greater vulnerability to temptation after happiness had been made salient.
Chocolate eaten in a library is not a pure meter of self-control. Hunger, liking chocolate, curiosity and assumptions about the study can affect consumption. The random assignment helps isolate the advertisement under those conditions, but the sample of 36 makes the result less stable than a large experiment would be.
Study 4 addressed a different question: was happiness special, or would making any series of goal-directed choices reduce persistence? The researchers randomly assigned 188 online participants to evaluate 25 pairs of everyday products. One group chose the item that would improve their happiness, while the comparison group pursued accurate judgment. Everyone then attempted a difficult anagram task and could quit at any time.
The happiness group stopped sooner. The researchers treated persistence time as a behavioral measure of the self-regulatory capacity remaining after the choices. The University of Toronto’s account of the research presents the chocolate and anagram findings as evidence that seeking happiness makes people more susceptible to temptation and less persistent.
What happened next was not measured. The study did not test whether eating the chocolates, stopping the anagram or any later decision actually made participants less happy. That longer cycle is the authors’ proposed interpretation, not a directly observed outcome of these experiments.
“Limited mental resources” should not be read as a literal battery
Kim and Maglio explain the pattern through resource depletion. Monitoring and changing emotion requires self-regulation; a person may then have less capacity or willingness available for the next act of persistence or restraint.
The wider claim that self-control operates like a finite fuel supply remains disputed. In 2016, a 23-laboratory preregistered replication involving 2,141 people tested a standardized ego-depletion sequence. The average effect was extremely small, and its 95 percent confidence interval included zero.
That replication did not test happiness pursuit and therefore does not invalidate the 2025 results. It does limit how literally the resource metaphor should be taken. The experiments did not observe or quantify mental fuel. They observed less persistence and more consumption after particular cues and choices.
Other mechanisms could produce the same behavior. Happiness-focused choices may increase self-conscious monitoring, make an unpleasant follow-up task feel less relevant or shift motivation toward immediate pleasure. Fatigue, attention and willingness are not interchangeable, even when they lead someone to quit at the same moment.
Valuing happiness is not automatically harmful
The study belongs to a broader literature on what happens when happiness becomes a goal. Kim and Maglio’s earlier 2018 paper, Vanishing time in the pursuit of happiness, reported across four studies that seeking happiness made time feel scarcer, particularly while the goal still seemed unfinished. That perceived shortage partly explained lower happiness.
Other evidence prevents a simple “stop wanting happiness” conclusion. A 2024 longitudinal study of 8,331 Dutch adults found that people who valued happiness generally reported higher life satisfaction, more positive emotion and less negative emotion. Within the same individuals, however, increases in valuing happiness did not predict greater life satisfaction one year later and had mixed emotional consequences, increasing both positive and negative affect.
Those results distinguish endorsing happiness as worthwhile from judging every present emotion against an ideal. A person can want a satisfying life without treating sadness, boredom or frustration as immediate failures to correct.
Culture and the content of happiness may also matter. Happiness imagined as social connection or meaningful activity may guide different behavior from happiness imagined as an internal feeling that must be maximized. The 2025 paper does not establish one rule across countries, ages or kinds of happiness goal.
The responsible takeaway is about monitoring, not surrender
The experiments do not show that exercise, therapy, social contact, behavioral activation or any other structured attempt to improve well-being is self-defeating. They did not compare those practices. Nor do they suggest that a person should simply accept persistent distress without support.
The Artful Age has previously covered a randomized study in which keeping a weekly gratitude journal was followed by better reported well-being. That finding need not conflict with this one. Attending to something already present may create a different psychological task from repeatedly trying to manufacture a preferred emotion.
The title’s reference to “constantly regulating” is the proposed chronic version of what the paper studied through trait measures and brief manipulations. The experiments did not place people under months of continuous emotional regulation. That longer-term account remains a hypothesis supported indirectly by the pattern across studies.
The narrow conclusion is useful enough. Making happiness an object of continual monitoring can interfere with the next act of persistence or restraint. That is different from saying happiness should never be pursued. It is the difference between living through a good moment and standing outside it, checking whether it is working hard enough.