"Incentives matter" is a cliche. But is it true? One new paper suggests not. German researchers got subjects to find words from a set of letters - a task requiring some creativity and inventiveness - and found that their ability to do so was barely affected at all by whether they were paid a flat fee, by results or by giving a prize to the better solvers. They concluded:

Neither on the aggregate nor on the individual level do we find effects of incentives on performance.

This is consistent with a claim made by Daniel Pink - that financial incentives can actually be bad for creativity:

For more right-brained undertakings - those that demand flexible problem-solving, inventiveness or conceptual understanding - contingent rewards can be dangerous. Rewarded subjects often have a harder time seeing the periphery and crafting original solutions. (Drive, p 46)

Research by Theresa Amabile, summarised here (pdf), corroborates this.

All this, though, runs into a question. Tim Harford and Robert Allen say that the industrial revolution - the burst of creativity that most transformed human life - occured in Britain because of incentives; high wages gave inventors an incentive to look for labour-saving machines. How can we reconcile this with the laboratory evidence that incentives don't spur creativity? I suspect there are two ways.

One is that attention (pdf) is a scarce resource, and incentives act to draw our attention to problems. So, for example, England's high wages in the 18th century drew attention towards the question of how to save labour costs, in a way that didn't happen in lower-wage nations.

Secondly, there's a selection effect. Incentives might not motivate a given group of people to be creative, but they might change the composition of that group, by - at the margin - attracting some able people to the task*.

My point here is that there's a big difference between believing that incentives matter and actually designing an effective incentive scheme. Instead, incentives matter through elusive, long-term and hard-to-manage channels, and are often unintended consequences. After all, nobody actually consciously designed the structure of incentives that gave us the industrial revolution.