Why New Year's Resolutions Fail (and What Works)

The standard take on resolutions is that they're a joke. You write one down on January first, you mean it, and by February the gym card is a bookmark and the journal is blank. The cliche has hardened into common sense: resolutions are wishful thinking with a date stamp, so why bother.

The data is more interesting than the cliche, and in two directions at once. Making a resolution actually does something - people who make one are far more likely to change than equally motivated people who just want to change without naming it. And yet most resolutions still fade, exactly as the cynics say. Both of those are true. The useful question isn't whether resolutions work in the abstract. It's what separates the ones that hold from the ones that don't - and at least one of the levers is smaller, and more in your control, than you'd guess.

Naming the Goal Beats Wanting It

Start with the part the cynics get wrong. In 2002, Norcross, Mrykalo, and Blagys ran a clean comparison: 159 people who made a New Year's resolution against 123 people who had a comparable goal but chose not to make one. Same ambitions, different framing. The researchers followed both groups by phone for six months.

The gap was not subtle. At six months, 46% of the resolvers were continuously successful. Among the nonresolvers - people who wanted the same change but never drew the line in the sand - it was 4%.

That contrast deserves a second read, because it inverts the cynical story. The act of making a resolution - choosing it, declaring it, attaching it to a moment - isn't empty theater. It is one of the cleaner demonstrations that intention plus commitment beats vague desire. Wanting to change and deciding to change are not the same thing, and the decision is the part that moves people.

So the cynics are right that most resolutions fail. They're wrong that making one is pointless. It's the best opening move available. The problem is what happens after the opening move.

The Cliff After the First Week

Here's the half the optimists skip. Making a resolution gets you off the line, but staying on the line is a different contest, and most people lose it.

The longitudinal picture is stark. Norcross and Vangarelli tracked 200 resolvers for up to two years and found that 77% were still going after one week - and only 19% were still going at two years. We've walked through that two-year drop in detail in why goals fail and starting over, so the short version is enough here: starting is easy, sustaining is brutal, and almost nobody fails at the start. They fail somewhere in the long middle, when the resolution stops being a decision and becomes a daily grind.

This is the trap that makes resolutions feel like a scam. The first week feels like proof - you did it, you're different now. Then the proof quietly expires, and without something holding the behavior in place, the old defaults win on attrition. The fix isn't wanting it more in January. It's building the resolution so it can survive March.

How You Word It Changes the Odds

Which brings us to the lever most people never touch: the shape of the goal itself.

In 2020, Oscarsson and colleagues ran the largest experiment of its kind - roughly 1,066 people who set a New Year's resolution, followed for a full year. Among those who reported back at twelve months, 54.7% said they had succeeded. (A caveat worth keeping: not everyone responded, and the people who go quiet tend to be the ones it didn't work for, so read that as success among responders, not a clean success rate for everyone who started.)

The headline finding wasn't the overall number. It was a split inside the data, based on how people framed what they were after. Some set approach-oriented goals - resolutions about doing more of something good, like start running three times a week. Others set avoidance-oriented goals - resolutions about quitting or cutting out something bad, like stop eating junk food. Same person, same willpower, different wording. The approach-framed goals did better: 58.9% succeeded, against 47.1% for the avoidance-framed ones.

The gap is real but worth stating honestly: it's about 12 percentage points, not a doubling. Still, that's a meaningful edge from a change that costs nothing - you're not working harder, just aiming at something you can act toward instead of something you can only abstain from.

The why isn't mysterious. An avoidance goal points you at a void. "Stop eating junk food" tells you what not to do and leaves a blank where the new behavior should be. Every craving becomes a test you can only pass by doing nothing, and nothing is hard to celebrate, hard to track, hard to repeat. An approach goal hands you an action. "Cook two dinners a week" is something you can put in the calendar, do, and notice you did. It fills the same gap from the other side.

There's a tempting next claim here that the study does not support: that piling on support makes resolutions stick. It found that some support helped, but extended support did not add more on top. So the takeaway isn't more scaffolding for its own sake. It's the cheaper, sharper move - point the goal at an action you can take, not an action you have to withhold.

Rewrite the Goal Before You Chase It

The practical work happens before you start, in how you write the thing down. Take whatever you've been meaning to do and check which direction it points.

If it's a quit-goal, flip it. "Stop scrolling at night" becomes "read ten pages before bed." "Stop skipping workouts" becomes "walk after lunch on weekdays." You're not abandoning the original intent - cutting the scroll, moving more - you're giving it a positive action to ride on, the kind the data favors and the kind you can actually do something about today. This is the same logic behind the WOOP method, which makes you name the concrete action and the obstacle in advance, and behind treating a goal as an identity you're stepping into rather than a habit you're trying to suppress. "Someone who reads at night" beats "someone trying not to scroll," because one of those is a person doing something.

Then plan for the cliff, not the launch. The first week will feel great and the data says that feeling lies - the test is week six, when novelty is gone and the behavior has to stand on structure instead of enthusiasm. Decide now where this lives in your week, what counts as done, and who's going to notice if it slips. The resolvers who lasted weren't the ones who wanted it most; they were the ones who built something to hold the behavior when wanting it wasn't enough.

And don't wait for January. Nothing in any of this research is about the calendar. The mechanism is naming a goal, framing it as an action, and committing to a start - and that works on a random Tuesday in June exactly as well as on the first of the year. The date was never the active ingredient. The decision was.

The Version That Survives

Strip it down and the picture is oddly encouraging. Making a resolution genuinely beats drifting toward the same goal without one - that's the 46% against 4%. Most resolutions still fade, because starting and sustaining are different skills and the second is the hard one. And the part you control more than you think is the wording: aim at something to start, not something to stop, and you tilt the odds in your favor before you've done a single rep.

So pick the one change you keep circling back to, and do one thing with it tonight: write it as an action, not an absence. Not "stop X" but "start Y." Then put the first instance of Y somewhere real - a time, a day, a place you'll actually be. That's the whole move. A resolution that names something to do, scheduled into a week that exists, with a way to see whether you did it - which, kept honestly over time, is just a goal you're tracking instead of a wish you're hoping for.