Safety behaviours are actions taken during a feared situation to prevent a feared outcome or to manage anxiety. They are subtler than avoidance and cause many of the same problems, and they are the main reason people can face their fears repeatedly without improving.
The core problem
Someone with social anxiety attends a party, rehearses every sentence, holds a drink to hide shaking hands, and leaves after an hour.
Nothing terrible happens, and no learning occurs, because the person attributes the outcome to the precautions rather than to the situation being safe.
The conclusion is that they got away with it, not that the fear was unfounded. The belief survives intact.
Misattribution of safety
This is the mechanism, described extensively in the work of Paul Salkovskis and others.
Because the safety behaviour was performed, the absence of catastrophe is credited to it.
Someone who breathes carefully during a panic attack and does not collapse concludes that the breathing prevented collapse, rather than that collapse was never going to happen.
Examples in panic
Sitting down immediately, holding onto something, carrying water or medication without using it.
Checking your pulse, controlling your breathing deliberately, staying near exits, having someone accompany you.
Examples in social anxiety
Rehearsing conversations in advance. Avoiding eye contact. Speaking quietly or briefly to reduce the chance of saying something wrong.
Holding a drink or a bag as a prop. Wearing particular clothes to conceal blushing or sweating.
Self-monitoring during conversation, which is particularly damaging since attention directed inwards makes performance genuinely worse.
Examples in health anxiety and OCD
Checking the body repeatedly. Seeking reassurance from others or from searching online. Repeated medical consultations for something already investigated.
In OCD, compulsions function as safety behaviours, whether visible rituals or mental acts such as reviewing and neutralising.
They make things worse directly
Beyond preventing learning, some safety behaviours produce the outcome they are meant to prevent.
Monitoring your own speech makes you less fluent. Trying not to blush increases arousal and makes blushing more likely.
Gripping something to steady shaking hands increases muscle tension and therefore the tremor.
Overbreathing to control panic produces the very sensations being feared.
Why they are hard to identify
They feel like sensible precautions rather than symptoms, and many are invisible to others.
Mental safety behaviours are the hardest to spot: reciting facts to reassure yourself, mentally rehearsing an escape plan, or repeatedly checking how anxious you feel.
People frequently complete a course of exposure without improvement because something like this was running throughout.
How to find your own
Ask what you do to get through the situation, and what you would feel unable to face it without.
Ask what you carry, who you bring, what you check and what you rehearse.
A useful test is to plan doing the thing without a particular precaution and notice the resistance. Strong resistance identifies a safety behaviour.
Dropping them
Usually gradually, one at a time, since removing everything at once is rarely sustainable.
The standard approach is to enter the situation, drop one safety behaviour, and observe what actually happens.
Anxiety typically rises initially, which is expected and is the point. The learning depends on discovering that the outcome does not follow from the precaution.
Why exposure fails without this
People frequently complete a course of exposure work with little benefit, and retained safety behaviours are the most common explanation.
Someone can face a feared situation twenty times and learn nothing, because each time they attribute survival to the precaution rather than to the situation being safe.
This is why a good therapist asks in detail what you did during the exposure rather than only whether you did it.
Where genuine precautions differ
Not everything protective is a safety behaviour. Carrying an inhaler for asthma or an adrenaline pen for anaphylaxis is medically necessary.
The distinction is whether the behaviour addresses a real risk or an anticipated catastrophe that has never occurred.
Where you are unsure, that is worth discussing with a clinician rather than deciding alone.
This is general information rather than medical advice.