Are small farms really more productive than large farms?

This paper shows that the study of the farm size-productivity relationship hinges on the choice of productivity measure. Our main insight is that using yields, a partial measure of productivity commonly used in the literature, may not be informative. This occurs because, in addition to total factor productivity, yields pick up input markets distortions and deviations from constant returns to scale. We examine the empirical relevance of this insight using detailed microdata from Uganda. We find an inverse relationship between yields and farm size. We show the relationship turns positive when accounting for market distortions and returns to scale; or when using a farm- specific component of total factor productivity.