The relationship between farm size and productivity is a recurrent topic in development economics, almost as old as the discipline itself. This paper emphasizes the importance of choice of productivity measures in the inverse relationship literature. First, we seek to clarify the common measures, their relationships, and their advantages and limitations in empirical work. Second, we argue that much of the existing literature inappropriately uses partial measures such as land productivity. Third, we discuss the dynamic nature of the farm size -productivity relationship and show that the identification of these dynamics will in part depend upon the choice of productivity measure. Lastly, using a pseudo-panel of Brazilian farms that are aggregated at the municipality and farm size levels over the period 1985-2006, we provide new empirical evidence on the inverse relationship between farm size and both land productivity and total factor productivity. The empirical exercise highlights the importance of choice of measure when testing the inverse relationship. The inverse relationship between size and land productivity is alive and well. The relationship between total factor productivity and size, in contrast, has evolved with modernization during this period, becoming increasingly U-shaped or even positive.

Lokshin and Radyakin (2012) present evidence that month of birth affects child physical growth in India. We replicate these correlations using the same data and demonstrate that they are the result of spurious correlations between month of birth, age-at-measurement and child growth patterns in developing countries. We repeat the analysis on 39 additional countries and show that there is no evidence of seasonal birth effects in child height-for-age z-score in any country. Furthermore, we demonstrate that the Demographic and Health Survey data used to estimate the correlation is not suitable for the task due to a previously unrecognized source of measurement error in child month of birth. We document results from several papers that should be re-interpreted in light of this issue.

No abstract