the entire subsidy incidence should be on the farmland owners. Using a complementary set
of policy quasi experiments, I find that farmers who rent the land they cultivate capture 75
percent of the subsidy, leaving just 25 percent for landowners. This finding contradicts the
prediction from neoclassical models. The standard prediction may not hold because of less
than perfect competition in the farmland rental market; the share captured by landowners …