Bonus and Malus in Principal-Agent Relations with Fixed Pay and Real Effort

Drawing on the proposer-responder game examined by Andreoni, Harbaugh, and Vesterlund (2003), I experimentally test four variations of a principal-agent relationship with fixed pay and real effort. Depending on the treatment, the principal can voluntarily, but at her own expense, (1) only reward her agent, (2) only punish her agent, (3) either reward or punish her agent, or (4) neither reward nor punish her agent. Compared to Andreoni et al., I find substantially higher investments in reward, about the same expenses for punishment, and no significant difference among the four treatments regarding output levels. I attribute these findings to the real effort aspect of the experimental design.