Compassionate deception: Do children tell lies to be kind?

© 2018 Gwen Dewar, Ph.D., all rights reserved

There are times when we approve of lying, when we believe it’s the better option.

For instance, suppose a would-be murderer comes to your door. He asks you to tell him where his intended victim is hiding. Should you tell him the truth?

Throughout the world, adults have the same intuition: Truth-telling isn’t an absolute moral imperative. Sometimes, the good of preventing harm to others outweighs the good of telling the truth.

But how do children wrestle with these considerations?

The youngest kids will have trouble because they lack the developmental skills to tell lies. As I explain elsewhere, an effective liar needs to have solid “theory of mind" skills, and show advanced levels of self-control.

But as children approach the age of 4, they become more adept with lying, and eventually they may begin to tell prosocial lies: fibs designed to spare other people's feelings, or otherwise protect other people from harm. 

The tricky part is verifying when, during development, children take this step. We know that even young toddlers show sympathy and kindness toward others (Tousignant et al 2017; Warneken 2013). But using deception to benefit others is a complicated, cognitive demanding business. 

In addition to the normal workload associated with lying -- figuring out how to fool someone else, and keeping the details of your story straight -- you also have to understand the perspective of your intended beneficiary, and anticipate the effects of your actions. 

What will happen to him if I tell the truth? What will happen to her if I lie? And what sort of lie should I tell?

It's a lot to juggle, and the problem doesn't only arise when a murderer comes to your door. We face daily decisions about telling the truth -- social interactions where our blunt honesty would hurt another person's feelings, or otherwise cause harm. And this is true for children as well as adults.

When does it all come together?

Studies suggest that some kids might begin telling prosocial lies during the preschool years. But the experiments can be hard to interpret, because they don't always pinpoint a child's motive for lying.

Is a child lying to protect another person? Or is the child motivated by something else -- like the desire to avoid conflict, or court social approval?

On closer examination, it appears true, selfless, prosocial lying hasn't been well-established in preschool children.

Some kids may tell prosocial lies at an early age -- especially when they pay special attention to the feelings and needs of others. But overall, studies don't show clear evidence of widespread, spontaneous, prosocial lying until children are 6 or 7 years old. 

Here's a look at several key experiments.

Your nose looks weird...but I won't tell you so.

We've all experienced the dilemma: Somebody asks us to evaluate his or her physical appearance, and we're conflicted. We'd like to avoid lying, but if we tell the truth, our listener will be offended or hurt. How do kids handle this situation?

Victoria Talwar and Kang Lee addressed this question by putting approximately 100 children, aged 3-7, to the test. 

Each child began by meeting a friendly adult with a camera. The adult showed the child how to use the camera, and then asked to be photographed. But the adult also asked for feedback:

"Before you take a picture of me, do I look okay for the picture?"

Most children answered "yes," but it wasn't true. Throughout the interaction, the adult's nose had been marked with a bright splotch of red lipstick. And the kids had definitely noticed it. At the end of the experiment, they confided to another person that the camera-wielding adult had not looked "okay."

Did these kids lie in order to be kindMaybe, but other interpretations are possible. They might, for instance, have felt too intimidated to say anything negative to the adult. In a follow-up discussion, only five children -- all of whom were older than 68 months -- provided clear evidence that their motives were prosocial. They explained that they hadn’t wanted to embarrass the person (Talwar and Lee 2002).

So while many children lied, we can't tell how of them were chiefly concerned with protecting the feelings of another person. Where else can we look for evidence? Let's consider how children respond when they receive gifts.

I don't like this gift, but I'll pretend to be thankful.

In another experiment, researchers gave more than 225 children an undesirable present (a gift-wrapped bar of soap). The children ranged in age from 3 to 11 years. How would they react?

When the gift-giver asked the children if they liked the undesirable gift, most kids said yes, and this response was especially likely if the kids had received parental coaching immediately before receiving the gift: When their parents urged them to act appreciative no matter how they felt, kids were more likely to tell a polite lie (Talwar et al 2007).

But once again, we must consider the question of motive. Even before the children participated in the gift-giving experiment, they had been exposed to parental and cultural messages about etiquette. Like children all around the world, they had probably been trained to show polite gratitude in response to receiving a gift. 

We therefore can’t assume that these children really met our criteria – that they were motivated by a desire to spare the feelings of the gift-giver.

What’s needed is an experiment that puts the spotlight on feelings. And that’s just what Felix Warneken and Emily Orlins came up with.

She's sad -- should I cheer her up?

Suppose you drew a picture, and asked your child for feedback. Would your child take your feelings into account?

That’s what Warneken and Orlins wanted to know, so they recruited 80 kids (ranging in age from 5 to 11 years), and administered the following test to each child.


1. An adult supervisor gives the child a set of hand-drawn sketches to evaluate. The child is instructed to sort the sketches into two piles: one for good drawings, the other for bad ones.

2. After the child has finished the sorting task, a second adult enters the room. She is holding a sketch in her hand, which she identifies as her own work. It is an objectively bad effort -- something the experimenters cooked up and tested on a focus group in a previous study. So we can assume the child will agree.

3. Depending on the child’s group assignment, one of two things happens next:

  • A child who has been randomly assigned to the “sad" condition will hear the artist explain that she feels sad and disappointed. She says she has worked very hard on the drawing, and is upset at the poor results.
  • A child who has been randomly assigned to the “neutral" condition will hear the artist say that she doesn't care that her sketch has turned out badly. She feels fine.

4. Finally, after listening to the artist talk, the child is asked to place the artist’s sketch in one of the two piles. The artist watches. What will the child do?


If a child's lies are motivated by empathic concern, his or her response should depend on the artist’s display of feeling. The child should be more likely to rate a bad sketch as “good" after hearing that the artist is sad.

And that’s what Warneken and Orlins found – at least among the older kids in their study.

Children aged 7 and up were more likely to place the sketch in the “good" pile if they had heard the artist express sadness. And the effect was especially striking for 10- and 11-year-olds. In the neutral condition, when the artist said she didn't care, children this age almost never bothered to lie. But when the artist said was sad, 10- and 11-year-olds lied about 70% of the time.

By contrast, 5-year-olds didn’t reliably distinguish between the “sad" and “neutral" conditions, and overall, they tended to place the sketch in the “bad" pile.

So in this experiment, many children showed evidence of lying to spare the feelings of someone else. They lied to make the artist feel better, just as an adult might.

It wasn’t true for the youngest children, but Warneken and Orlins found a way to change that. In another trial of the experiment, they provided kids with a role model – someone who behaved charitably.

This role model told the kids she wanted to make the artist “feel good," and then placed one of the artist’s (poor) sketches in the “good" pile.

That was enough for the 5-year-olds to get the message. When it was their turn to judge, they, too, distinguished between the “sad" and “neutral" conditions. They were more likely to place a sketch on the “good" pile when the artist seemed sad (Warneken and Orlins 2015).

What about the murderer at the door?

In the experiments we’ve discussed so far, the question was whether or not kids tell lies in order to spare the feelings of the recipient. We’ve found that they do, at least by the age of 7 years.

But what about the murderer at the door? In that scenario, the liar isn’t trying to make the recipient of the lie – the would-be murderer – feel better. The goal is to mislead the recipient into doing something against his own interests, and thereby protect a third party.

It’s a more complex set of issues to keep track of. You have to do all the usual mental work required for deception, but you also have to recognize the needs of a third party. And you need to figure out what sort of lie would be the most helpful to the person you're protecting.

When the murderer asks, should you simply claim that you don't know where his intended victim is? Or should you actively throw him off the scent, send him looking in the wrong direction? If so, where should you tell him to look?

In essence, the murderer-at-the-door scenario requires us to understand the perspectives of two different people simultaneously, and to help one by manipulating the other.

When do children show signs of mastering this sort of problem?

Answers come from fascinating experiments by Teresa Harvey and her colleagues. The researchers didn't present children with an actual murderer at the door. Obviously! But they came up with an analogous scenario that children would find much less distressing:

Should you lie to a thief to prevent him from stealing another child's toy?

A total of 270 children, ranging in age from 5 to 8 years old, participated. And the central procedure in these experiments involved telling kids a story. The story was presented as a true, and reflective of current events: Two children were hanging out at a local park, but one child couldn't find the other. 

The researchers randomly assigned each child participant to hear a different version of this story:

  • Kids randomly assigned to the stealing condition were told that a child named Alex was playing with a "cool toy." Another child, Jamie, was also at the park, and he wanted to steal the toy. But Alex was hiding, and Jamie (the seeker) didn’t know where to look.
  • Kids randomly assigned to the sharing condition were told that a child named Riley was looking for his friend, Dylan, at the local park. Riley had two cookies, and wanted to share one with Dylan. But Dylan was hidden from view, and Riley (the seeker) didn’t know where to look.

After learning those details, kids in both conditions were shown pictures of the park – pictures that revealed the precise location of the hidden child.

In addition, an adult provided kids with a map of the park, explaining that she was going to deliver this map to the seeker (Jamie or Riley). 

The adult gave kids pens, and asked them to mark the map. Could you please circle the place where you want the seeker (Jamie or Riley) to look?

Here, then, was the crux of it: The kids could either give away the true location of the hidden child, or deliberately mislead the seeker with a lie. What did they do?

Interestingly, the 5- and 6-year-olds seemed to lie at random. They were no more likely to help or hinder the Seeker, whether it was Jamie the Thief, or Riley the Cookie-Sharing Friend.

But the 7- and 8-year-olds showed a pronounced tendency to tell lies along prosocial lines: They had more than 8 times the odds of lying in the stealing condition than lying in the sharing condition.

Were the younger kids simply clueless about maps?

It appears not, because the researchers checked the children’s comprehension of the map procedure.

And researchers found they could improve the youngest children's performance by helping them think about the consequences of theft. In one version of the experiment, the adult telling the story spells out how Alex would feel if Jamie took his toy:

"If Jamie takes Alex’s toy, Alex will be very sad."

Adding that single sentence to the story was enough to change how the 5- and 6-year-olds behaved on the map test. Now they distinguished between the stealing scenario and the sharing scenario. They were much more likely to circle the wrong location on the map when it meant misleading the would-be thief (Harvey et al 2018).

And what about self-interest? Are kids sometimes dissuaded from telling "noble lies" because they don't want to pay the cost?

There's no doubt that telling prosocial lies can carry a cost, and experiments show that children take these costs into account.

For example, experiments reveal that kids are less likely to say they like an undesirable gift when they believe this will prevent them from obtaining a better one. And the effect is bigger for younger children: Preschoolers are less likely than older kids to tell prosocial lies when the personal cost is high (Popliger et al 2011).

So the development of prosocial deception depends on many things -- empathy, parental coaching, cultural pressure, the emergence of cognitive skills, and a child's tolerance for self-sacrifice. 

For more information about the development of lying, see this article. For information about fostering empathy and kindness in children, see these evidence-based tips.


References: Prosocial lies

Harvey T, Davoodi T, Blake PR. 2018. Young children will lie to prevent a moral transgression. J Exp Child Psychol. 2018 Jan;165:51-65.

Talwar V and Lee K. 2002. Emergence of white lie-telling in children between 3 and 7 years of age. Merrill-Palmer Quarterly. 48:160–181.

Talwar V, Murphy SM, Lee K. 2007. White lie-telling in children for politeness purposes. Int J Behav Dev. 2007 Jan;31(1):1-11.

Tousignant B, Eugène F, Jackson PL. 2017. A developmental perspective on the neural bases of human empathy. Infant Behav Dev. 48(Pt A):5-12.

Warneken F. 2013. The development of altruistic behavior: helping in children and chimpanzees. Social Research 80 (2):431-442.

Warneken F and Orlins E. 2015. Children tell white lies to make others feel better. Br J Dev Psychol. 2015 Sep;33(3):259-70.

Title image of father and son joking by Eugene Kim / flickr

image of wolf at door from an illustration by Walter Crane, shared by emmeffe6 / flickr

image of boy hiding in tree by Victor Ramos/flickr