We contribute to the inverse farm size-productivity puzzle (IR) literature by examining the relationship using a unique data set from southern Ghana that covers farms between 5 and 70 ha. The study uses an instrumental variable (IV) for land size to mitigate some effects of measurement error in land size. The inverse relationship between farm size and farm productivity is upheld when ordinary least squares estimators (OLS) are applied but becomes insignificant although still negative in the IV estimation. The results show that measurement error in land size attenuates the IR. While some studies found the IR to flatten and then become positive, this study finds that in Ghana, the IR only flattens.
