Selection Mechanism and Variation of Years of Schooling across Birth Months in Malawi

The years of schooling in Malawi varies across birth months substantially and consistently at least over thirty years. Those who were born in the second half of each year have 1.6 years longer of schooling than those who were born in the first half of each year. The difference is substantial given that the average years of schooling in Malawi is about six years. The availability of food across months and the variation of birth weight across birth months do not match the variation of years of schooling across birth months. Compulsory education law does not explain this pattern either. To explain the pattern of years of schooling across birth months, we propose a selection mechanism hypothesis that among individuals who was born in the second half of each year, only those who have high innate ability could survive the malnutrition during pregnancy and the most vulnerable periods after the birth. This implies that those who were born in the second half of each year and those who are alive now have higher innate ability on average. Because of higher innate ability, such individuals had longer years of schooling than other individuals. To prove the validity of our hypothesis, we first show that the number of individual who were born in the second half of each year and who are alive now is 50 percent lower than the number of individuals who were born in the first half and are alive now. Second, using a novel approach used by G{\o }rgens, Meng and Vaithianathan (2012), we regress each person's years of schooling on his or her parents' birth months controlling each person's birth month and parents' education. We show that the years of schooling of children whose parent were born in the second half of each year is longer than those of children whose parents were born in the first half of each year. This result shows that individuals who were born in the last half of each year survived sever malnutrition and have innate ability.