Taxed Avoidance: American Participation in Unsanctioned Internation

Summary of working paper 6313
Featured in print Digest

In 1910-31, 25 percent of babies died by day ten if they weighed between 1500 and 1999 grams; in 1988, only 3 percent of babies in the same weight range died.

In Unequal At Birth: A Long-Term Comparison of Income and Birth Weight(NBER Working Paper No. 6313), NBER Faculty Research Fellow Dora Costa examines the relationship among income, birth weight, and infant health. Using data from the period 1910 to 1988, she finds that differences in birth weight by income class were only slightly larger in the past than they are today. Rather, it was differences in maternal height that accounted for most of the historical difference in birth weight by income class. That is, the taller the mother, the more likely she was to have a larger birth weight baby. Since maternal height is partly a function of adequate nutrition early in a girl's life, past inequalities in terms of health were transmitted across generations, Costa shows. Further, as income status grows, so does the level of nutrition, providing increased opportunity for a woman to reach her full biological potential in terms of height.

While the average birth weights of babies born in 1910-31 and in 1988 are about the same, there are huge differences in mortality during those two periods for low birth weight babies. In 1910-31, 25 percent of babies died by day ten if they weighed between 1500 and 1999 grams; in 1988, only 3 percent of babies in the same weight range died. Further, in the past low birth weight babies weighed less than they now do on the tenth day of life -- the critical neonatal period -- which has significant ongoing health and mortality consequences. Costa attributes this disparity to poor obstetrical and medical practices during that time, including best medical practices which we now know produced insufficient feeding of the newborn. Costa notes that the data suggesting that average birth weights from 1910-31 and 1988 were similar may be misleading because many low birth weight babies in the earlier period did not survive the stresses of childbirth.

Costa used data from the New York Lying-In Hospital in Manhattan, which served the working class, primarily immigrants, during 1910-31, and compared it with data from the 1988 National Maternal and Infant Health Survey. To make the data compatible, she restricted the study to white births to married mothers and examined subsamples of the data.

Many factors affect birth weight, including infectious diseases, alcohol or other drug use, smoking, and heavy physical work. Nutrition is also a key determinant of birth weight, but the relationship is not linear. In other words, nutritional deprivation must pass a threshold level before fetal development is affected significantly, since the fetus will initially draw nutrients from the mother's body resources.

Costa's research controlled for all observable variables including mother's height and age, number of children in the family, infant's time in the uterus, and sex of the newborn child. Her data suggests that modern obstetrical and neonatal technological advances, coupled with better nutrition for both mother and newborn, have improved both birth and adult health outcomes.