Adolescence as a Critical Period? A Multi-country Study Using Heights


Speaker:Alessandro Tarozzi, Associate Professor at Universitat Pompeu Fabra
Date: Thursday 9 May 2019
Time: 14.00
Location: Constantine Leventis

Further details

Economic historians have often used adult height as an indicator of economic development (Fogel 1994). Indeed many countries have experienced significant increases in average height over time, during periods when GDP was also growing rapidly, see e.g. Hatton and Bray (2010). However, more recently the cross-sectional association between average height and economic development has been shown to be complex, with data from several developing countries showing little or no relationship between average adult height and GDP at birth, Bozzoli et al. (2009), Deaton (2007). A key proposed explanation is that if early life mortality is sufficiently high (as it is in very poor countries), the negative impact of health insults on height may be counter-balanced by a "harvesting" effect that only leaves stronger and taller individuals alive. Using updated data on the height of more than 2.5 million women from all Demographic and Health Surveys (DHS) available at the time of writing we confirm that these associations are indeed weak. In contrast, using data on child height measured in the same surveys, we show that the associations between child height and GDP at birth is close to zero at birth but is positive and increasing with age. We hypothesize that these contrasting patterns are explained by a "reversal" that likely takes place during the adolescent growth spurt. A simple model can be calibrated to produce elasticities of height with respect to conditions at birth that are first growing and then declining with age. Using panel data from four different developing countries from the Young Lives Project we show that such U-shape patterns also exist when we link height to measures of socio-economic status at birth, with the inflection point actually taking place during adolescence.