What the data actually shows
The mainstream position in affective science is the functional account of emotion: emotions evolved because they coordinate attention, body, and behaviour in ways that were, on average, useful. Within that view there is no category of inherently bad emotions, only emotions that are pleasant or unpleasant to experience and well or poorly matched to the actual situation.
Work on emotion regulation, especially by James Gross and colleagues, distinguishes between strategies that help and strategies that tend to cost you. Expressive suppression — clamping down on the outward signs of a feeling — has been associated in his research with the feeling persisting or rebounding, reduced memory for what happened, and more physiological strain, without much reduction in the emotion itself. Reappraisal, by contrast — rethinking what a situation means — tends to look more effective and less costly in the same body of work.
There is also evidence that trying hard not to feel something can amplify it. Research on thought suppression (often associated with Daniel Wegner) found that instructing people not to think about something can make it more intrusive, an 'ironic' rebound effect. The general pattern across these literatures is consistent rather than proven beyond doubt: fighting an unpleasant emotion head-on is usually less helpful than understanding what it is telling you. Effect sizes vary and not every study agrees.
Why this feels different from how it actually is
Unpleasant emotions feel bad in the moment, and it is a short step from 'this feels bad' to 'this is bad and I should not be feeling it.' The discomfort is real and immediate, while the usefulness of the signal is abstract and delayed, so the feeling reads as a malfunction rather than a message.
Culture reinforces this. A lot of everyday language and wellness messaging sorts emotions into 'positive' and 'negative,' and treats the negative ones as states to be eliminated or fixed. Being told to 'stay positive' or 'not let it get to you' implies that anger, fear, or sadness are failures of self-control rather than ordinary responses with a job to do.
We also tend to see other people's calm exteriors and assume they are not feeling what we are feeling. Because emotions are mostly private, it is easy to conclude that having strong difficult feelings is unusual or a sign something is wrong with you, when in fact they are a near-universal part of being a person.
It is more accurate to call these emotions unpleasant than bad. An unpleasant emotion can be doing exactly what it evolved to do.
What the research says to do about it
The most consistent recommendation from the regulation literature is to work with an emotion rather than against it: notice it, name it, and ask what it is responding to before deciding what to do. Simply labelling a feeling accurately — 'this is anger because a plan got blocked' — is associated in some studies with a small reduction in its intensity, sometimes called affect labelling.
Reappraisal — looking for a different, accurate way to frame the situation — has the strongest support as a deliberate strategy in Gross's work, particularly for situations you cannot change. It is not the same as forcing a positive spin; it is asking whether your first interpretation is the only true one. Choosing or adjusting situations in advance, so you encounter fewer needless triggers, is another well-supported lever.
Because this is general information and not treatment, the honest guidance is also to watch duration and impact. Difficult emotions that pass and that you can act on reasonably are part of normal life. Emotions that are persistent, distressing, or interfering with sleep, work, or relationships are worth discussing with a qualified clinician, who can offer support that no article can.
What the research says does not help
Trying to suppress or 'push down' an unpleasant emotion is one of the least effective strategies in the research. Gross's work links expressive suppression to the feeling persisting, more bodily strain, and worse recall of events, with little of the relief people hope for. Bottling up rarely makes the emotion go away; it more often delays or amplifies it.
Treating difficult emotions as enemies to be defeated tends to add a second layer of distress on top of the first — being anxious about being anxious, or angry at yourself for being angry. This meta-distress is frequently more wearing than the original feeling and does not make it pass faster.
Forced positivity — insisting you should only feel good, or papering over a real feeling with affirmations you do not believe — has weak support and can backfire. The evidence favours accurate framing over cheerful framing: acknowledging the real feeling and what prompted it tends to do more than denying it.
What tends to cause trouble is not the presence of a difficult feeling but what we do with it.
What this looks like in real life
Anger as a blocked-goal signal
A plan you cared about falls through and you feel a hot flash of anger. On the functional view, that anger is information — it flags a blocked goal or a sense of unfairness — not a defect. You can feel it intensely without lashing out; the leverage is in the response, not the feeling. Naming it accurately ('this is anger because a plan got blocked') is associated in some studies with a small drop in its intensity.
Being anxious about being anxious
You notice worry, decide you shouldn't be feeling it, and start worrying about the worry. This second layer — meta-distress — is frequently more wearing than the original feeling and does nothing to make it pass faster. Treating the difficult emotion as an enemy to defeat tends to add distress on top of distress rather than resolving it.
Real numbers in context
This is a topic where honest context means resisting hard numbers. There is no credible statistic that ranks emotions as good or bad, and any percentage claiming to do so should be treated with suspicion. What the research offers instead is a direction of effect, not a precise figure: across studies, suppression tends to look costly and reappraisal tends to look helpful, with effect sizes that are real but modest and that vary by person and situation.
The most useful thing to internalise is qualitative. Difficult emotions are ordinary and near-universal, they usually carry information worth reading, and the leverage is in how you respond rather than in whether you feel them at all. Where feelings are persistent or overwhelming, that is a reason to seek a clinician, not a sign that the emotion itself was the failure.