Good thing to be thinking about as school starts. I'm a bit skeptical about Dan Pink, but several people I know and respect have a high opinion of him. It feels sometimes like he's selling creativity.
But on his recent TED talk, he reports (link to his talk) some social psychology experiments that help me understand a big divide in teaching: incentives in classroom management. A simplification of his point would be that incentives work best on simple, direct tasks. On more challenging problems rewards don't work and, in fact, impede. What does work? Choice! Autonomy! (Conditions of Learning chalk up another point.)
Plus he's pretty entertaining.