THOMPSON: Predictably Irrational
Dan Ariely's experiments in Predictably Irrational offer another explanation of why it was inevitable that data-driven accountability would have a corrupting effect on schools, and the theorists who embraced it. Ariely showed that cheating increases dramatically when the stakes are not money but a coupon, even if it can immediately be transformed into money. This obviously explains the additional lure of corruption with stocks and complex financial packages, as well as NCLB-type accountability. Dishonesty also increases when people see others cheating in an open manner, thus creating a social acceptance of the transgression. Cheating grows further when it is obvious that violators come from the same social milieu. Dishonesty is discouraged, however, when people see cheating by people from another social set.
Neither does Behavioral Economics does not lend support for Pay for Performance. People who would change a flat tire for free are less likely to help if offered five dollars, because the transaction is thus switched from the "social domain" to the "market domain."
So, why not focus on "win win" solutions, rather than seeking fool-proof accountability and incentive schemes? Although Ariely demonstrated that tax evasion could be reduced by requiring filers to sign on the top of the form rather than the bottom, the IRS refused to consider the proposal. So, Ariely convinced an insurance company to participate in his experiment. After showing the savings that would be produced by such a minor change, the company still refused to alter its forms?!?!
By the way, Ariely’s experiments showed more cheating in bars frequented by Wall Street analysts than Washington establishments serving congressional aides. - John Thompson