Lesson 7: Safety Behaviors — Why Playing It Safe Can Backfire

In Lesson 6, we designed experiments to test anxious predictions. This lesson looks at something that can quietly undermine those tests: safety behaviors. These are the subtle things you do to prevent a feared outcome. They feel protective, but they often keep the fear exactly where it is.

Module: 4, Testing things out

Learning objective

By the end of this lesson, you should be able to describe what safety behaviors are, explain how they maintain anxiety, and identify your own.

The concept: what safety behaviors are

A safety behavior is an action you take during a feared situation to prevent a feared outcome. Paul Salkovskis described the idea in 1991, arguing that the way people seek safety during anxiety is not just a symptom but part of what keeps the problem going [1].

Some examples:

Feared situation	Feared outcome	Safety behavior
Giving a talk	“I’ll freeze and people will judge me”	Reading from a script word for word
Socializing	“People will see I’m nervous”	Holding a drink to hide shaking hands
Driving on a motorway	“I’ll lose control of the car”	Gripping the wheel, driving very slowly, only using the left lane
Eating in front of others	“I’ll spill food and embarrass myself”	Only ordering food that’s easy to eat

Each behavior is designed to prevent the feared outcome. The problem is what happens next.

How they maintain anxiety

The mechanism is simple and frustrating. If you grip the steering wheel tightly and arrive safely, your mind attributes the safe arrival to the gripping, not to the fact that driving is generally safe. You never update the belief “I’ll lose control,” because you never tested it without the safety net.

Clark and Wells described this process in their cognitive model of social anxiety. A person who believes they will say something foolish may rehearse every sentence mentally, speak very little, and avoid eye contact. When the conversation ends without disaster, they conclude it went well because of the safety behaviors, not despite them [2]. The belief stays intact, and so does the anxiety.

This connects directly to the behavioral experiments in Lesson 6. If Deepa had eaten in the canteen while staring at her phone the whole time, she might have concluded “Nobody looked at me because I didn’t look up.” The experiment would have been invalid.

What the research shows

A study randomly assigned 49 socially anxious participants to two conditions: both included exposure to a feared social situation, but one group was asked to drop their safety behaviors while the other was allowed to keep them. The group that dropped safety behaviors showed greater reductions in anxiety, in negative beliefs and in self-focused attention [3].

The picture is not entirely simple, however. A review of the debate noted that some researchers have argued for judicious use of safety behaviors in the early stages of exposure, particularly when the alternative is that the person avoids the situation entirely [5]. The key principle from an inhibitory learning perspective is that what matters most is whether the feared outcome is violated: the expectation must be tested, not just endured [6].

A broader review of the role of safety behaviors in exposure concluded that dropping them tends to produce better outcomes, but the evidence base is still relatively small and the clinical picture depends on the specific anxiety disorder and the specific safety behavior [4].

A worked example

The following is a fictional example used for teaching.

Leo has a job interview next week. His main fear is that he’ll go blank when asked a difficult question. His safety behaviors include:

• Preparing scripted answers for every possible question.
• Bringing notes to glance at during the interview.
• Steering the conversation toward topics he’s rehearsed.

These make him feel more prepared, but they have costs:

• He sounds rehearsed and stilted, which ironically makes a worse impression.
• He spends so much energy managing the script that he can’t listen properly to what the interviewer is saying.
• Afterwards, he thinks: “It only went OK because I had my notes. Without them I’d have fallen apart.” The belief that he can’t handle an unexpected question remains.

A therapist might suggest a behavioral experiment: at a low-stakes meeting (not the big interview itself), Leo tries answering a question without referring to any prepared material, and notes what actually happens.

Identifying your own safety behaviors

Safety behaviors can be hard to spot because they feel like sensible precautions. A few questions can help:

1. When I’m in the feared situation, what do I do to feel safer? Look for physical actions (gripping, holding, sitting near the exit), mental actions (rehearsing, monitoring your body, counting), and interpersonal ones (only speaking to people you know, asking questions to avoid being asked).
2. What would I be afraid would happen if I didn’t do this? The answer points to the belief underneath.
3. Does this thing I do prevent me from finding out whether the feared outcome would actually happen? If yes, it’s likely a safety behavior.

The gray area

Not everything that looks like a safety behavior is one. Taking medication prescribed by your doctor is not a safety behavior. Wearing a seatbelt is not a safety behavior. The test is whether the action prevents you from updating a belief that is keeping your anxiety going.

A common misunderstanding

“I should stop all safety behaviors immediately.” Dropping safety behaviors works best when it’s planned, graded and connected to a specific prediction. Just as exposure is more effective when it violates a specific expectation, dropping a safety behavior is more useful when you write down what you fear will happen without it, and then observe what actually does.

Optional reflection

Think about a situation where you regularly feel anxious. Can you identify one or two things you do to manage the anxiety while in the situation? For each, ask: “If I didn’t do this, what do I think would happen?” You don’t need to drop the behavior yet. For now, just noticing it is the step.

When to get support. If your safety behaviors are extensive, for example, if you have a long routine of checking or preparing before you can leave the house, or if dropping them feels impossible, a therapist trained in CBT can help you work through them gradually. If anxiety is severely limiting your life, speak to a GP. If you are in crisis, contact emergency services or a crisis line: in the UK and Ireland, Samaritans on 116 123; in the US, call or text 988.

Recap

• Safety behaviors are actions taken during a feared situation to prevent a feared outcome.
• They maintain anxiety by preventing disconfirmation of the feared belief.
• Dropping them during exposure tends to produce better outcomes, though the evidence base is still growing.
• Identifying safety behaviors starts with asking what you do to feel safer, and what you fear would happen without it.
• Dropping them works best when planned and connected to a specific, written prediction.

Previous lesson: Lesson 6: Behavioral Experiments | Next lesson: Lesson 8: The Anxiety Maintenance Cycle