Most people trying to change something reach for a habit. Wake earlier. Walk daily. Stop scrolling at eleven. The habit holds for a while, and then a difficult week arrives and it doesn't.
The usual explanation is discipline. A large piece of research published in 2025 suggests something more useful.
Zhu and colleagues pooled nineteen studies covering 13,340 people across physical activity, eating and drinking, and examined the relationship between two things usually studied separately: habit — the automatic, cue-triggered part of a behaviour — and identity — whether a person sees that behaviour as part of who they are.
The two turned out to be strongly related (r = 0.55) but not the same thing. That distinction is the interesting part.
What “not the same thing” means in practice
A habit is a groove. It runs on repetition and context: same time, same place, same trigger. It is efficient, and it is fragile in exactly the way grooves are fragile. Change the context and the groove disappears. This is why routines collapse when people travel, when a child falls ill, when the job changes.
An identity is not a groove. It is an answer to a question — what kind of person am I? It does not need the same cue, the same room or the same hour. It travels.
The research points to identity doing work that habit cannot: maintaining a behaviour when the automatic pull isn't there. When the groove is gone, something still has to decide.
The uncomfortable half of the finding
Here I want to be accurate rather than encouraging.
The same body of work finds that identity is stable — and stability cuts both ways. The thing that makes an identity durable in a hard week is the same thing that makes it slow to shift in the first place. Researchers looking at this are fairly direct: identity may be difficult to change.
So “just change your identity”, which appears across a great deal of the internet and sometimes attached to my own field, is not supported by the evidence in the way people use it. Identity is not a setting you switch. It behaves more like a reputation you hold with yourself — and reputations move slowly, and only on evidence.
Where the leverage actually is
Not in the habit, and not in a declaration. In the gap between deciding and doing.
Almost everyone reading this has intended something they did not do. Psychologists call it the intention–behaviour gap, and it is among the most reliably measured findings in the field. What this newer work adds is that aligning identity with the behaviour at the point the intention is formed strengthens that intention and narrows the gap.
Put plainly: it matters less whether you intend to walk each morning, and more whether the walk is being taken by someone who considers themselves a person who moves. Same behaviour. Different author.
What I take from it
Three things, held loosely.
Habits are worth building, but they are scaffolding rather than structure. They will not survive a genuinely hard month, and that is not a character failure.
Identity change is slow and evidence-based, not declarative. You do not talk yourself into being someone. You accumulate a small pile of proof, and eventually the pile is difficult to argue with.
And when something you were doing has stopped, the useful question is usually not where did my discipline go. It is who was doing this, and are they still around?
If the thing you stopped doing had a person attached to it, who were they?