Recent econometric studies indicate that the effect of government farm subsidies on farmland rental rates may be smaller than once thought. This literature has corrected for bias due to expectation error in measured subsidy payments. We suggest two additional sources of bias—inertia and tenancy arrangements—that may explain the discrepancy between theoretical predictions and empirical estimates of subsidy incidence. We identify a model that accounts for these issues, employ panel data from Kansas to estimate it, and find that an additional dollar per acre of government subsidy increases rental rates by $0.12 per acre in the short run and $0.37 per acre in the long run.