Dan Ariely speak tonight on the topic of lying. Interesting stuff: he's done studies that show that almost everyone cheats and/or lies a little bit, though few people go big time. The general drift is that most of us only lie to the extent that we don't create conflict with our self-image as good people.
One thing that tends to happen is that when people cheat for a period of time, the "What the heck" response kicks in, and then there's a jump upward. The research team tried to figure out if it was possible to pull that new standard of acceptance down again, and the answer is yes. They found that if you make some sort of confession (e.g., write what you did wrong and feed it into a paper shredder) and then ask for forgiveness (even on paper that gets shredded) you pretty much re-establish your self-image as a good person and revert to your previous norm.
But that doesn't work once a cheat becomes socially acceptable. When almost everyone does it, you partition off that behavior and no longer think of it as bad. Pirating downloads off the internet apparently falls in that category now. It's too late to re-set the norm.
The studies found that long-term consequences have little effect on short-term decisions or behavior. (Those of us with teenage boys could have told you that!) But they also found that if you're reminded of morality before taking a test or undergoing some sort of temptation -- if you read over the Ten Commandments or swear on a Bible -- you're less likely to cheat. This is true even for avowed athiests.
We bought a copy of Ariely's new book, and while getting it autographed I asked if he's considered publishing best practices to prevent cheating in educational magazines so that teachers have access to the research results. He said that what they've found works best in test-taking situations is to have students write out their own personal honor code, by hand, prior to taking a test. It's far more effective than just signing a standard statement.