Extrinsic rewards work for mechanical tasks and backfire on creative ones

Pink opens with the candle problem, a classic experiment by psychologist Karl Duncker in which participants must attach a lit candle to a wall using only a candle, a box of tacks, and matches, without letting wax drip on the floor. Solving it requires the insight to empty the tack box and use it as a candle-holder—a creative leap. When researchers offered cash rewards for faster solutions, performance got worse, not better, because the reward narrowed attention and squeezed out the peripheral thinking needed for the insight.

The same reward structure, however, reliably improves performance on mechanical tasks with a clear, algorithmic path to the answer—stack these boxes, sort these files faster for more pay, and people do exactly that. Pink argues this split explains why so much modern management, built on carrots and sticks originally designed for industrial-era, rule-based labor, actively undermines performance on the nonroutine, conceptual work that increasingly makes up the knowledge economy.

Takeaway: Match your incentive structure to the task—cash bonuses for creative work often narrow the very thinking you need.