Pay for Performance?

Does higher pay ensure better performance? Columbia University sociology professor Duncan Watts ran an experiment that mimics Amazon.com's Mechanical Turk service. He determined that higher pay does indeed motivate people to do more work, but not necessarily better work. The addition of a bonus tied to accuracy had no measurable impact. Indeed, the bonus seemed to encourage those already doing poor work to crank out yet more substandard output. The July/August Pennsylvania Gazette carries a full story on this experiment.

I suspect that Watts failed to incorporate adequate feedback mechanisms into his experiment, however. Like a real manager, did he issue warnings to those doing subpar work? Did he tell them that they were in danger of losing their jobs, let alone their chances for earning bonuses? If he did not, as the article seems to imply, the experiment is seriously flawed.

Meanwhile, Citigroup is joining the ranks of those major financial firms raising base pay in response to the political heat over bonuses.