Monday, January 29, 2007

Pants of Fire?

I’m starting a new research project with a member of the psychology department here. The idea came to me while watching TV. It’s a very basic question; are people more effective liars when they have something to gain or something to lose. A little bit of looking failed to turn up anything remotely close to what I wanted to find out. It just so happens that one of the psychology faculty does work on lie detection and polygraphs. With a little bit of planning we’re off to the races. Here’s what I’m thinking about.

Subjects are brought into the laboratory and hooked up to a polygraph, and then the polygraph is calibrated. Now here’s the fun part. Half the subjects will be told that they will earn an extra dollar every time they can fool the polygraph, ie lie and get away with it. Later in the experiment the situation will change. Subjects will be told that a dollar will be subtracted from their total earnings if they’re caught in a lie. Subjects are all paid a $5 or $10 show up fee so that they can’t go in the hole, and maybe bonus marks for a class as well. The other half of the subjects will have something to lose first, and something to gain second. This will allow us to figure out if order maters. We can now find out if the subjects are better at lying when they have something to gain or when they have something to lose.

Motivation: Most economic theory treats potential gains and loses of the same magnitude the same way. This is starting to change, as this makes it impossible why somebody would voluntarily buy both insurance and lottery tickets. We’re getting a sense, in part from the kind of neuroeconomics I talked about earlier and from more basic experiments, that people actually treat potential gains and losses very differently, with a potential loss having more impact. If the two possibilities are dealt with in different areas of the brain there is no reason to assume the impact on decision making to be the same.

Here’s the set up. Assume lying effectively (fooling the polygraph) has a relatively constant cost in terms of effort. If potential losses are more powerful than potential gains, subjects will be more effective liars when facing a loss of funds vs when they can gain.

The potential for application is significant. Just think about law enforcement. People who have something to gain by lying should be easier to detect than those who have something to lose. This also has broader implications in terms of personal motivation. It means the stick is more powerful than the carrot. Scary thought. It might be interesting to look at the issue as short term vs long term. It may be that potential losses are more powerful motivators in the short terms, but potential gains are more powerful over the long term. I’m still trying to figure out a way to get at that question.


Vanessa Childs Rolls said...

Interesting. What is your control?

economistatlarge said...

No control - within subjects design.