Gratitude: what the research shows and where it ends

“Be grateful you have a job at all.” The sentence lands over Sunday lunch, just as you were trying to say that work has been suffocating you for months. The person who said it meant no harm. You go quiet, nod and finish your plate. On the way home you carry the original weight plus a new one: the feeling that you are ungrateful.

A few days later you read on your phone that keeping a gratitude journal “rewires the brain” and that three good things a day are all that stands between you and a better life. The same word, two entirely different uses. In the first, gratitude is a stopper for a conversation; in the second, a miracle cure. Neither is what researchers actually studied.

In this post I would like to look calmly at what the research on gratitude shows, how large the effect is when it is measured fairly, and where gratitude ends — where it turns into something that harms a person instead of helping.

What gratitude actually is

The psychologist Robert Emmons, who has worked in this field longer than anyone, describes gratitude in two steps. First you notice that there is something good in your life. Then you recognise that at least part of that good did not come from you alone: it was given by another person, by circumstances, by nature, and for a person of faith, by God. Gratitude is therefore an emotion with an addressee, or at least with a sense of having received something.

That second step is why gratitude has been part of religious traditions for thousands of years, in the form of prayer and thanksgiving. People who do not believe take it just as seriously: you can be grateful to a grandmother, a teacher, a stranger who helped you change a tyre, or to the sheer chance of being born in the age of antibiotics. The studies I describe below were carried out with believers and non-believers alike; the exercises require neither conviction. Someone who thanks God and someone who thanks people or chance are making a very similar psychological movement: for a moment they stop looking only at what is missing.

It also matters what gratitude is not. It is not a debt, it is not politeness, and it is not downward comparison. “Others have it worse” is a comparison, not gratitude. As you will see, researchers kept the two apart from the very first experiments.

The study that started it all

In 2003 Robert Emmons and Michael McCullough published a paper with three experiments that almost every text on gratitude cites today. In the first, they divided students into three groups. The first group spent ten weeks writing down, once a week, up to five things they were grateful for. The second wrote down up to five hassles, and the third simply listed events of the past week, with no instruction about what kind.

At the end, the group that wrote about gratitude rated their lives somewhat more favourably, looked at the coming week with more optimism, and reported fewer physical complaints and more exercise than the other two groups. In the second experiment participants wrote every day for two weeks, and one of the comparison groups reflected on ways in which they were better off than others. Gratitude and downward comparison were thus separate conditions by design: the researchers did not consider them the same thing.

I find the third experiment the most interesting. It involved adults with neuromuscular diseases who wrote down what they were grateful for every evening for 21 days; the comparison group only completed questionnaires. People in the first group reported more pleasant emotions, better sleep and a stronger sense of connection with others. Their partners noticed the change in their own ratings too.

Two years later Martin Seligman and colleagues tested several brief exercises online. The “three good things” exercise, in which you write down each evening three things that went well and why, was associated with greater happiness and fewer depressive symptoms six months later. The gratitude visit, in which you write a letter to someone you never properly thanked and read it to them, produced a large immediate effect that faded after about a month.

What later meta-analyses found

A single study is like a single photograph. Only when you lay many side by side can you see whether the smile in the first one was a fluke. Meta-analyses, which pool the results of dozens of experiments, have made the picture of gratitude considerably calmer.

In 2016 Don Davis and colleagues reviewed experiments with gratitude exercises and found that the answer depends on what you compare the exercise with. Compared with groups that only filled in questionnaires, well-being after gratitude exercises was better by roughly 0.3 of a standard deviation, which psychologists call a small effect. When gratitude was compared with another exercise that also keeps a person occupied and raises an expectation of improvement, the advantage shrank further, and against other psychologically active exercises it could no longer be reliably shown.

In 2021 David Cregg and Jennifer Cheavens looked at a narrower question: do gratitude exercises reduce symptoms of depression and anxiety? The effects were small, and smaller for anxiety than for depression. The authors wrote something you rarely read in popular articles: anyone seeking help with depression or anxiety would do better to choose approaches with stronger evidence behind them.

Why the number is smaller than the headlines

There are several reasons, all of them quite down to earth. As early as 2010 Alex Wood and colleagues pointed to a problem with comparison groups. If you ask one group to write about gratitude and another to write about hassles, and the first feels better at the end, you do not know whether gratitude lifted their mood or listing hassles lowered it. Some of the early, most exciting results rested on exactly such comparisons.

The second reason is expectation. Someone who signs up for a study on happiness and is given an exercise that sounds sensible expects to feel better. The third is time: most experiments last a few weeks, and we know little about what happens after a year. The fourth is habituation. People get used to good things quickly; an exercise that felt fresh in the first week can be just one more chore on the list by the sixth.

A small average effect, however, is not the same as “it does not work”. The average hides people for whom the exercise helps noticeably and people who get nothing from it or even find it unpleasant. So the question is not whether gratitude “works” but for whom, when, and in what form.

Where gratitude genuinely helps

I find it most convincing where it is directed not inwards but towards another person. In 2018 Amit Kumar and Nicholas Epley asked people to write someone a letter of thanks and to predict how the recipient would feel. The writers consistently underestimated how pleased and surprised the recipient would be, and overestimated how awkward it would feel. Because of that mistaken forecast we say thank you less often than would be good for both sides.

In relationships, spoken thanks has another role: it tells the other person they were seen. “Thank you for taking the children today, I could see you were tired yourself” is not a courtesy phrase. It is a message that the effort did not fall into a void. Couples and families in which such sentences are rare often suffer not from a lack of love but from the feeling that everything they give is taken for granted.

Gratitude is also useful as a counterweight to an attention that evolution has tilted towards danger. The brain notices mistakes, insults and threats faster, and remembers them better, than whatever went smoothly. A few minutes of deliberately looking for the good does not erase that, but it can make the picture of the day more accurate. It works in a similar way with the constant measuring against others that I wrote about in Comparing yourself on social media: what constant measuring does to the brain: for a moment it turns your gaze from what others have to what is already here.

Where gratitude ends

The trouble starts when gratitude is no longer an addition to what you feel but a substitute for it. In recent years this has acquired the name toxic positivity: the expectation that you will answer every hard thing with something bright, and at once. The term is not a clinical diagnosis, but it describes something emotion research knows well.

In 2018 Brett Ford and colleagues showed in a series of studies that people who accept their unpleasant emotions and thoughts without judging them are psychologically better off in the long run than those who fight them. An unpleasant emotion you are allowed to feel usually passes sooner than one you first have to defeat. Gratitude used as a tool for silencing therefore achieves the opposite of what was intended.

Three limits seem especially important to me.

• Gratitude as a debt. Someone who grew up with “after everything we did for you” often experiences gratitude not as warmth but as a bill that is never settled. Out of this grows a guilt that has nothing to do with real responsibility; I describe the difference in Healthy guilt and toxic guilt: how to tell them apart.
• Gratitude as a reason to stay. “At least he doesn’t hit me, I should be grateful” or “other bosses are worse” are sentences people use to explain to themselves why they must not want more. Gratitude for what is good in a relationship does not outweigh the harm the relationship does.
• Gratitude as proof that you may not be sad. You can have healthy children, a roof and a job and still be exhausted, lonely or sad. One does not cancel the other.

When gratitude becomes a stopper

A conversation at home might sound something like this (composed, not a real person):

“Lately I can barely get up. At work I just wait for the day to end.”
“Come on, do you know how many people would love a salary like yours? Be a bit grateful.”
“I am grateful. I’m just tired.”
“Then I don’t know what else you want.”
“… Nothing. Forget it.”

Nobody in this conversation is being unkind. The person recommending gratitude often cannot bear their own helplessness in the face of someone else’s distress, and reaches for the first sentence that promises calm. The effect, though, is that the conversation ends before it has begun. The one who tried to say something learns that it is safer to stay silent.

It would be different if the reply were: “That sounds exhausting. What wears you out the most?” Gratitude could still enter that conversation, but later, and from the mouth of the person who is struggling. When it comes too early and from outside, it works as an order. Something very similar happens to the people I write about in You have everything, yet you are unhappy: the depression others don’t understand.

We are just as capable of putting the stopper in ourselves. “I have no right to be in a bad mood, others don’t even have this much.” That inner voice is usually inherited and sounds like gratitude, but it is really self-criticism in its Sunday best. A kinder route is described in Self-compassion instead of self-criticism: what Kristin Neff’s research shows.

An exercise: one thank-you a week

If you would like to try gratitude the way it has been studied, and without pressure, I suggest a form that combines writing with thanking a specific person. Once a week is enough; with daily writing the exercise quickly turns into a chore.

1. Choose a day and ten minutes. Sunday evening, Wednesday after lunch, whenever, as long as it is the same each week. Take a sheet of paper or the notes on your phone.
2. First write down what was hard this week. One or two sentences, no editing. This tells you the exercise is not there to cover anything up.
3. Write down one to three specific things that were good. Not “health” or “family” but “my neighbour brought me soup when I had a cold”. For each, add why it happened and who had a hand in it.
4. Choose one and pass it on. A message, a call or a sentence in person: what the person did and what it meant to you. Two sentences are enough.
5. Notice your body. If there is a little more room after the exercise, carry on. If you wrote it through gritted teeth or with a sense of “I ought to”, skip it this week. That is information too.

If you live your faith, the third step can be a prayer of thanks. If you do not, it can be a thought for the people and coincidences that brought you to this week. The exercise is the same. If you like to follow how your mood moves over the weeks, a mood diary can help as well.

When gratitude is not the right answer

This whole post argues for a moderate, honest use of gratitude, so I also need to say where that argument does not hold.

In depression. A person who is depressed often knows they “should be grateful” and cannot feel it. That is not a flaw of character but a sign of a state in which pleasant emotions are muted; there is more on this in Anhedonia: when nothing brings joy any more. A gratitude list can then become one more piece of evidence that something is wrong with you. If you recognise yourself here, the right step is a conversation with your GP or a psychiatrist, not a new notebook.

When you are in danger or in a relationship that degrades you. Then you do not need more gratitude; you need safety and support. If you are experiencing violence, in Slovenia you can call the SOS helpline for women and children who are victims of violence on 080 11 55, and in an emergency 112 or the police on 113.

In fresh grief. For someone who has just lost a person, “be grateful for the years you had” as a rule does not help. Gratitude for the time shared often comes by itself, but later and on its own timetable.

When the question is bigger than the exercise. Sometimes what lies behind “nothing brings me joy” is not a lack of gratitude but the question of what all this is for. I write about that in Existential crisis: when you ask yourself what it is all for. For some people gratitude is closely woven into faith and into doubts about it; how to bring that into the therapy room is the subject of Faith and therapy: can I talk to my therapist about God?.

And finally: for some people this exercise simply does not fit. Research reports averages. If writing lists gives you nothing, that does not make you an ungrateful person. Perhaps you notice the good on a walk, in conversation, or while working with your hands.

What a conversation can do

When someone tenses up at the word gratitude, it is worth asking where they learned that. In many families gratitude was a currency: it paid for food, schooling, the parents’ sacrifices. A child who had to be grateful instead of being allowed to be angry or disappointed as well finds it hard, as an adult, to tell warmth from debt. In therapy the two can slowly be untangled: seeing what you really received, what you missed, and that you are allowed to feel both at once.

As a therapist in training under supervision in marriage and family therapy, I work in person in Ljubljana, and also with English-speaking clients. If you are curious how the attitude to giving, receiving and owing has been passed down the generations in your family, the longer genogram session A Look into the Past is designed for exactly that.

What to remember

• Gratitude exercises have a small but real effect on well-being in research; compared with other active exercises the advantage is smaller still.
• In depression and anxiety the effects are modest. Gratitude is an addition, not a treatment.
• It works best when it is specific and spoken to a person. The recipient is usually more pleased by thanks than you expect.
• Gratitude stands beside difficult emotions, not instead of them. When it becomes an order, a debt or a reason to stay where it hurts, it is no longer gratitude.
• The exercise neither requires faith nor excludes it: you can thank God, people or chance.