April 15th — Cheating and taxes

April 11, 2008 BY danariely

Tax Season is here and cheating is in the air. We are all trying to find receipts, and figure out what we can reasonably claim as deductions. Did we end up talking about work last summer when we invited our co-workers for a BBQ?

Let me tell you about an experiment we did at MIT on cheating with some unsuspecting students. We took a sheet of paper with twenty simple math problems that everybody could solve given enough time, but we didn’t give people enough time to solve them all. We handed the participants the quiz and told them that we would pay them a dollar per correct answer. When the five minutes were over, they handed the sheet back to us. On average, they solved four problems and we paid them each four dollars.

We then gathered another group for the same experiment. But this time, when the five minutes were up, we asked them to shred the piece of paper and simply tell us how many questions they solved correctly. Interestingly, on average participants in this group said they solved about 7 problems correctly.

So what happened? Our students didn’t become smarter all of a sudden. Instead, they cheated. But it’s not as if a few students cheated a lot. What we found was that a lot of students cheated a little bit. Sounds a little bit like the tax issue right?

But here is something else we did. We asked a third group of students to sign an honor code before solving the problems, ” I understand that this survey falls under the MIT honor code.” When the five minutes were over they shredded the piece of paper. But it turns out that these students who signed the honor code didn’t cheat at all.

What does this tell us about the tax code? Well, imagine what would happen if instead of signing your tax form at the bottom, you signed a pledge up front: “I declare that everything I say here will be the truth, the only truth and nothing but the truth so help me God and everybody else,” and then fill out your tax form. Would cheating be lower under that condition? I am not sure but if I were the IRS, I would definitely try it out.


And here is a youtube version of this