We conducted a Öeld experiment with real-life tenants in Ethiopia
to test the incentive e§ects of Öxed-wage, sharecropping, Öxed-rent, and
ownership contracts. The experimental task resembles a common process
in agricultural production. The sharecropping contract is essentially a
piece-rate scheme framed as a proÖt sharing agreement. The sharecropping
output was about 11 percent smaller than the Öxed-rent output. Surprisingly,
it is statistically indistinguishable from the Öxed-wage output,
despite substantial piece rates. This e§ect is driven by real-life sharecroppers.
Their sharecropping output was signiÖcantly smaller than that of
non-sharecroppers, and in one region, it was even 10 percent lower than
sharecroppersí Öxed-wage output. Based on qualitative interviews and
historical accounts, we argue that our subjects dislike sharecropping contracts
because of the unfair proÖt sharing and the controversial allocation
of the land. The contractual performance may therefore depend on the
perceived fairness of the incentive scheme.