The (Honest) Truth About Dishonesty

With those hot commodities in hand, Francesca Gino, Mike Norton (a professor at Harvard University), and I set about testing whether participants who wore fake products would feel and behave differently from those wearing authentic ones. If our participants felt that wearing fakes would broadcast (even to themselves) a less honorable self-image, we wondered whether they might start thinking of themselves as somewhat less honest. And with this tainted self-concept in mind, would they be more likely to continue down the road of dishonesty?

Using the lure of Chloé accessories, we enlisted many female MBA students for our experiment. (We focused on women not because we thought that they were different from men in any moral way—in fact, in all of our previous experiments we did not find any sex-related differences—but because the accessories we had were clearly designed for women.) We wondered whether to use the sunglasses or the handbags in our first experiments, but when we realized that it would have been a bit more difficult to explain why we wanted our participants to walk around the building with handbags, we settled on the sunglasses.

AT THE START of the experiment, we assigned each woman to one of three conditions: authentic, fake, or no information. In the authentic condition, we told participants that they would be donning real Chloé designer sunglasses. In the fake condition, we told them that they would be wearing counterfeit sunglasses that looked identical to those made by Chloé (in actuality all the products we used were the real McCoy). Finally, in the no-information condition, we didn’t say anything about the authenticity of the sunglasses.

Once the women donned their sunglasses, we directed them to the hallway, where we asked them to look at different posters and out the windows so that they could later evaluate the quality and experience of looking through their sunglasses. Soon after, we called them into another room for another task. What was the task? You guessed it: while the women were still wearing their sunglasses we gave them our old friend, the matrix task.

Now imagine yourself as a participant in this study. You show up to the lab, and you’re randomly assigned to the fake condition. The experimenter informs you that your glasses are counterfeit and instructs you to test them out to see what you think. You’re handed a rather real-looking case (the logo is spot-on!), and you pull out the sunglasses, examine them, and slip them on. Once you’ve put on the specs, you start walking around the hallway, examining different posters and looking out the windows. But while you are doing so, what is going through your head? Do you compare the sunglasses to the pair in your car or the ones you broke the other day? Do you think, “Yeah, these are very convincing. No one would be able to tell they’re fake.” Maybe you think that the weight doesn’t feel right or that the plastic seems cheap. And if you do think about the fakeness of what you are wearing, would it cause you to cheat more on the matrix test? Less? The same amount?

Here’s what we found. As usual, lots of people cheated by a few questions. But while “only” 30 percent of the participants in the authentic condition reported solving more matrices than they actually had, 74 percent of those in the fake condition reported solving more matrices than they actually had.

These results gave rise to another interesting question. Did the presumed fakeness of the product make the women cheat more than they naturally would? Or did the genuine Chloé label make them behave more honestly than they would otherwise? In other words, which was more powerful: the negative self-signaling in the fake condition or the positive self-signaling in the authentic condition?

This is why we also had the no-information (control) condition, in which we didn’t mention anything about whether the sunglasses were real or fake. How would the no-information condition help us? Let’s say that women wearing the fake glasses cheated at the same level as those in the no-information condition. If that were the case, we could conclude that the counterfeit label did not make the women any more dishonest than they were naturally and that the genuine label was causing higher honesty. On the other hand, if we saw that the women wearing the real Chloé sunglasses cheated at the same level as those in the no-information condition (and much less than those in the fake-label condition), we would conclude that the authentic label did not make the women any more honest than they were naturally and that the fake label was causing women to behave less honestly.