Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We study the impact of information manipulation by a principal on the agent's effort. In a context of asymmetric information at the principal's advantage, we test experimentally the principal's willingness to bias (overestimate or underestimate) the information she/he gives to her/his agent on her/his ability in order to motivate her/him to exert more effort. We find that (1) principals do bias information, and (2) agents trust the cheap‐talk messages they receive and adjust their effort accordingly. Therefore, biased messages improve both the agent's performance and thus the principal's profit. This, however, does not increase efficiency. We also find that overestimation occurs much more often than underestimation. Making the signal costly in an additional treatment reduces this effect. (JEL D83, C92, M12)