Working Paper 57. Early Nutrition and Cognitive Achievement in Pre-school Children in Peru

Abstract

The aim of this paper is to examine the link between early stunting and later cognitive achievement. It differs from nutrition–learning studies in other developing countries in that it focuses on pre-school children, and therefore time spent in school plays no role. Data comes from a cohort of children in Peru (Young Lives survey), for which information is available from two points of time: 2002, when they were six to 20 months old; and 2006-7, when they were 4 to 6 years old. For the empirical estimation, I use Ordinary Least Squares (OLS), controlling for lagged child and household characteristics and taking into account community characteristics. To try to identify early nutrition, I use maternal height and exposure to low temperatures during the first months of life as instrumental variables (IVs) for early nutrition. OLS results show a positive and statistically significant impact of early nutrition on cognitive achievement four years later for the full sample. Using maternal height as an instrument, IV estimations produce a higher coefficient for the parameter of interest, implying that OLS might be downward biased. Using average minimum temperature as an IV for a sub-sample of communities located in the Highlands produces results in the same direction. Preliminary results are reported and discussed