Is “every lie a sin,”as St. Augustine held some 1,600 years ago? People do sometimes tell “white lies” to spare feelings. But what if deception, in the right circumstances, doesn’t simply tread lightly on sensitivities, but actually breeds trust and promotes other forms of good? Many will judge those kinds of deceptions to be ethical, moral and even helpful, according Maurice Schweitzer, a Wharton professor of operations and information management, and co-author Emma E. Levine, a Wharton doctoral student. In their recent research paper, “Are Liars Ethical?: On the Tension between Benevolence and Honesty,” they look at “deception that can sometimes be helpful to other people.”

In this interview with Knowledge at Wharton, Schweitzer and Levine also consider the implications for employers and employees when it comes to such things as codes of conduct, performance reviews and doctor-patient relationships.

An edited version of the transcript appears below.

On looking at deception ‘a little differently’: 

Maurice Schweitzer: In our work, “Are Liars Ethical?: On the Tension between Benevolence and Honesty,” we look at deception a little differently than people have in the past. We look at deception that sometimes can be helpful to other people. We typically think about deception as selfish deception: I lie to gain some advantage at the expense of somebody else. And we typically think of honesty as something that might be costly to me, but helpful to others. 

In our research, we actually disentangle those two things. We think about deception that can help other people, and honesty that might be helpful to myself and maybe costly to somebody else. When we separate honesty and deception from pro-social and pro-self-interests, we find that people actually don’t care that much about deception. We find that the aversion to lying, when people say, “Don’t lie to me,” what they really mean is, “Don’t be really selfish.”

On understanding when is the right time to lie: 

Emma Levine: Broadly, the advice on the table until now has been, “Always be honest. Honesty is the best policy.”We’re suggesting that maybe we revise that advice, and make the statement, “Lie sometimes.” The key takeaway is to understand when we should lie. When does honesty actually harm trust and seem immoral? And when can deception actually breed trust, and be seen as moral?

We could think about how that might be used in delivering feedback, because managers often have to balance this tension between benevolence – giving kind, supportive feedback – and honesty – giving critical, harsh feedback. We might suggest that managers actually err on the side of benevolence, if their goal is really to promote trust and relationships.

On separating dishonesty from selfishness: 

Schweitzer: What surprised us was that when we really separated dishonesty from selfishness, we found that what’s driving all of this behavior – what people judge to be moral or immoral, when people trust somebody or they don’t trust somebody – we find that what really matters is how kind, how benevolent somebody is, and that deception plays a very small role….

“When we tell people, ‘Never lie to me,’what we really mean is, ‘Don’t be selfish.’” –Maurice Schweitzer

On using economic gains to study the topic: 

Levine: We studied this tension between benevolence and honesty, using economic gains. We had people come to a lab and watch someone, or interact with someone, who had an opportunity to lie about the outcome of a coin flip or a die roll, some objective fact, in order to make money for their partner. So, they could lie and be dishonest in a way that would help a counterpart. And then we had participants judge this dishonesty versus truth-telling, which would harm the counterpart.

We found that when participants are lied to in a way that helps them, they actually really appreciate that dishonesty. It boosts trust, and it causes them to think that their counterpart is moral. Also, when they simply observe somebody who is lying to help someone else, to earn someone else more money, they also trust this person more, and think this person is more moral than someone who is always honest.

We ran three different experiments in our paper …and each experiment had about 200 participants. So, each experiment builds on one another. The first one just looks at lying to help someone else, versus being honest, which harms others. And then the second experiment, which had another 200 participants, looks at the effect of lying to help someone else, even when it actually causes harm. So, we disentangle intentions from outcomes. And then in our third experiment, which again, had another 200 or so participants, we start to disentangle the effect of lying on the self, versus the effect of lying on others.

On balancing the truth with benevolence: 

Schweitzer: We think about interaction with other people — it could be our customers, it could be subordinates, it could be that we’re giving people feedback, or we might be explaining why we’re terminating a contract. The interactions we have with other people are often influenced not just by the brutal honesty that we want to communicate, but also by important concerns about demonstrating the care we have for other people.

When we, for example, tell a client that we’re going to go with a different vendor, sometimes we want to phrase things in a way that is different from the complete unvarnished truth. We find that we often really care about the benevolence that people signal through their communication and that’s often more important than the harsh and complete truth.

So, the advice here is to think about ways in which we might balance these two things. Our findings suggest we want to tilt toward benevolence, and we often do this intuitively. One key takeaway here is: Some of the codes of conduct, some of the exhortations we tell people when we say that honesty is a supreme concern, or honesty is always the best policy – what we’re finding is that that’s not really what we mean.

On new rules and procedures that related to the research: 

Schweitzer: There are really two key ideas that we suggest for implementing some of these findings. One is to change codes of conduct. That is, when we talk about codes of conduct, and we have honesty in there, for the sake of hypocrisy, I would take that out because it’s not what we mean, and it’s not what we do. And second, we think about training, we think about the rules that we implement when we interact with our colleagues at work and our external clients. We don’t always strive to be completely honest. It’s not a core value, really. And our argument is to be consistent, both in the codes of conduct, and the way we train and teach other people. We want to think about demonstrating concern, and other elements of morality that are going to be really important in guiding institutions.

“We found that when participants are lied to in a way that helps them, they actually really appreciate that dishonesty. It boosts trust….” –Emma E. Levine

I think, ultimately, we can become more consistent and less hypocritical when we think about the role and this tension between honesty and benevolence, and how we don’t always mean that the complete truth is what we’re after. I think instead of saying we should always be honest with others, we should think about always treating others the way we’d want to be treated. We should think about creating a caring community. And we should think about this tension. That is, as managers, as executives, we want to think about balancing this in our training programs and the way we deal with other people, and explicitly acknowledge that we’re making trade-offs between honesty and care for others. That’s the key idea.

On other practical implications of the research: 

Levine: There are a lot of domains in which individuals face this conflict between honesty and benevolence very, very intensely. One example is in health care. Doctors frequently have to deliver very negative news to patients. And actually, prior research has found that oftentimes, doctors do lie. They inflate the positivity of these prognoses. We seem to think this is bad, and I think doctors feel a lot of tension and conflict around how to handle the situation.

But our research suggests that perhaps doctors, and maybe teachers and parents, should be explicitly acknowledging this tension and this trade-off, and thinking about and talking about how to navigate it. When might benevolence and kindness, and maybe a little dishonesty, be right or be appreciated, and how could that enhance the delivery of medicine? 

On how the research applies outside a business setting:  

Schweitzer: Another case where we manage this balance between being completely honest and managing being benevolent is when we interact with children. This is true as parents, and it’s also true as educators. So, teachers will need to give feedback to students. And they have to balance this tension between being completely candid, and being benevolent and kind — demonstrating kindness and concern for the child.

As parents, we frequently tell our kids, “Never lie.” But that’s not at all what we mean. For example, before we go over to Grandma’s house, we might tell them, say, “Remember, thank Grandma for that sweater, and tell her how much you like it, even though we both know you never wear it.”

On misperceptions dispelled by the research: 

Levine: For centuries we’ve been talking about honesty as a virtue and dishonesty as a vice. That’s an idea that’s been around in philosophy …starting with Immanuel Kant, and actually before that. But also, empirically, there’s a wealth of research on deception, documenting that deception harms trust, that deception makes us angry, that deception is immoral. But these studies have really just looked at selfish deception — the lies we tell that hurt others, and help ourselves. 

We’re trying to overturn this idea, because we find that when we actually disentangle deception from selfishness, people don’t care all that much about dishonesty, per se. They really care about helping others. So, I think this sheds new light on this idea that deception is bad.

“Instead of saying we should always be honest with others, we should think about always treating others the way we’d want to be treated.” –Maurice Schweitzer

On what sets the research apart: 

Schweitzer: Prior work that studied deception has really confounded deception with selfishness. What we know about deception is really what we know about selfish deception. Our work separates this out, so we can actually learn about deception, per se. What we find is that the way we feel about deception itself is actually not all that bad. And it challenges this enormous body of research that has made these broad claims about what deception does or doesn’t do. Really, what we know is what deception that is selfish does or doesn’t do.

On what’s next: 

Levine: I’m really excited to look at the effects of pro-social lying, and how this tension between benevolence and honesty plays out in the feedback process. How does it affect performance, and how does it affect relationships when we give feedback? A lot of people might say that honesty is the way we improve. Honesty is important to get accurate feedback, to become better employees, better students. But it’s very possible that we need benevolence. We need kindness. Sometimes we need dishonesty to be given the confidence to improve.

Benevolence and honesty don’t only have effects on trust and character perceptions, but might actually affect performance and organizations. I’m excited to take a look at that next.