Shrinking in a growing economy is not so puzzling after all

Bodenhorn, Guinnane, and Mroz (2017) argue that the diminution of heights during the Industrial Revolution and in the Antebellum U.S. is an artefact of the biased nature of the samples analyzed. We demonstrate that it would be an unfathomable coincidence if men and women all self-selected into scores of completely independent samples in such a way as to bias them in the identical direction. Instead, wWe show that BGM's periodization is flawed and that their statistical models are misspecified, because they fail to consider the extent to which they introduce severe multicollinearity into their regressions. In addition, they fail to specify how they selected the samples they included in their analysis. In contrast, we argue that the economic transition from a predominantly agricultural to an increasingly industrial society was not a smooth process and lags in adjustment led to nutritional stresses. Height of a typical man in the U.S. decreased by 0.75 inches at a time when incomes were growing at a rate of 1.2% per annum. The developing human body of children and youth was sensitive enough to these nutritional stresses to register their effect better than monetary measures could. While nutritional status did decline during the Industrial Revolution in Europe and at the onset of modern economic in the U.S., by the second half of the 19th century agricultural productivity caught up with the increased demand for foodstuffs and height reversals became a rarity. Thus, although markets adjusted, they did not do so instantaneously. Consequently, physical stature declined during this adjustment process although the wealthy were shielded from the increased price of nutrients. So, the divergence in average incomes and average heights at the threshold of the modern age is not so puzzling after all. (C) 2018 Elsevier B.V. All rights reserved.