Do affirmations work? What the research says, and when they backfire

Do affirmations work? The short answer is that it depends on who is repeating what. In 2009 Joanne Wood and colleagues took people with low and high self-esteem and asked some of them to repeat the sentence "I'm a lovable person" while others did something else. Then they measured mood. A survey in the same paper showed that people trust affirmations: many use them and believe they help.

What came next was not what most people expect. Among people with low self-esteem, those who repeated the sentence felt worse than those who did not. Among people with high self-esteem, repeating it made them feel better, but only a little. The authors summed it up this way: repeating positive statements may help some people and backfire for the very people who "need" them most.

This is one paper, and a series of small experiments. But it explains why affirmations feel great to some people and grating to others, and why the second reaction does not mean something is wrong with you.

Why affirmations don't work, and sometimes make things worse

Say something that sounds true and it lands. Say something you do not believe and an objection shows up at once: "sure, right." The objection is not neutral. It brings evidence: yesterday, last year. You are not practising confidence, you are practising an argument with yourself, and losing it out loud.

Wood's team tested this too. In one experiment people were asked not just to repeat the sentence but to think about how it was true. For people with low self-esteem that did not help either: they felt worse. When they were asked to consider how the sentence was both true and not true, their mood did not drop. The honest version turned out to be safer than the pretty one.

Among people with low self-esteem, mood got worse after repeating "I'm a lovable person".

How self-affirmation differs from affirmations

It is easy to mix these up because the names are close. Self-affirmation theory is not about repeating sentences. The typical procedure in those studies is that a person writes about what matters most to them, their values, and why. Not "I am great" but "here is what I care about and here is how it showed up".

In a 2014 review, Geoffrey Cohen and David Sherman describe the mechanism like this: people need to feel broadly adequate and capable, and when something threatens that picture, a defensive reaction kicks in that gets in the way of learning and of hearing unwelcome truths. Writing about values widens the picture. One failure stops being a verdict on everything, because something else that is also about you sits next to it. According to their review, interventions of this kind improved outcomes in education, health, and relationships, and the benefits sometimes lasted months or years. The word "sometimes" matters: this is what happened in a number of studies, not a guarantee.

A 2015 meta-analysis by Epton and colleagues gathered 144 experimental tests of self-affirmation in health contexts. The effects are small but reliable: on acceptance of health-risk information (d+ = 0.17), on intentions to change (0.14), and on behavior itself (0.32). The authors point out that this is comparable to other interventions aimed at changing health behavior. So not a miracle and not "reprogramming", just a small dependable shift.

The practical upshot: if you look for what in this area rests on data, it is writing about values, not repeating bold claims about yourself.

How to phrase an affirmation so you don't argue with it

The format below is mine, not a protocol from a specific study. It takes the general idea (values, writing, specifics) and turns it into something you can do in one evening.

Pick one value. Not a list of ten. Honesty, care, curiosity, independence, craft. Any one that resonates without effort.

Write for ten minutes about why it matters and where it showed up in what you did. Specific scenes, not slogans. "Calling back someone I did not want to talk to, because I had promised" works better than "I am a responsible person".

Do not grade what you wrote. It is not an essay, and nobody is marking it.

If you do want a short line, change the verb. "I handle everything" invites an objection. "I am learning to say what I think" does not: it describes a process, and it is hard to argue with a process. Workable starters: "I am learning to...", "what matters to me is...", "today I can...", "I choose to try...".

Run the lie test. Read the line and listen: did a voice say "that's a lie"? Then it is too far from what you feel. Move it one step: from "I am calm" to "I can slow down for a minute", from "I deserve love" to "the people near me matter to me, and I want to be honest with them".

What the research does not show

First: none of the studies above shows that affirmations "reprogram the subconscious" or change your life. There is no such evidence. Epton's meta-analysis concerns health and small behavior shifts, Cohen and Sherman's review concerns writing about values, and Wood's paper is about short-term mood.

Second: Wood's result does not mean affirmations hurt everyone. For people with high self-esteem repeating was slightly helpful, and for everyone else remember that this is one experiment with one sentence. The real conclusion is more modest: a line that triggers an inner "that's a lie" is unlikely to be worth repeating on a schedule.

Third: if the feeling that something is wrong with you lasts for weeks and gets in the way of living, writing about values or any other phrase will not replace it. Then a conversation with a professional matters more than any journaling practice.

What to do tonight

Choose a value. Set a timer for ten minutes and write about where it showed up in what you actually did over the past year. If a line turns up along the way that you want to keep, run it through the lie test and adjust the verb. And if no line comes, that is better: the scenes work without slogans.

If you also want to hear these lines rather than only read them, there is a separate piece on that: recording affirmations in your own voice.

Frequently asked questions

Do affirmations work or not?

It depends on the form. In a 2009 experiment, repeating bold statements about yourself made mood worse for people with low self-esteem. Writing about your own values, studied as self-affirmation, produces small but reliable effects.

Why don't affirmations work?

When a sentence is far from what a person believes about themselves, an objection appears and repeating turns into an argument. In Wood's study, mood dropped for exactly those people who found the statement implausible.

How do I write affirmations that help rather than hurt?

Describe a process or something that matters, not a finished result: "I am learning to...", "what matters to me is...", "today I can...". Run the line through the lie test and move it a step if the objection shows up.

How is self-affirmation different from affirmations?

An affirmation is a short line you repeat. In research, self-affirmation is writing about your values and how they showed up in practice.