To what extent is children’s health determined by their mothers’ health? This column analyses three decades’ worth of data on over two million children across 38 developing countries to explore how health is transmitted across generations – and how public policy can respond.
Using representative data for thirty-eight developing countries, we show that investments in the nutrition and health of women have an intergenerational payoff, improving the survival and health of future generations. Girls under-nourished as children tend to grow up to have less successful pregnancies, producing children who are of lower birth weight, more likely to die young and, if they survive, to experience growth retardation. This suggests that son-preferring behaviour is short-sighted (see Bhalotra and Rawlings 2011, 2013).
Intergenerational transmission in developing countries
The analysis uses comparable micro data on over 2 million children born to about 600,000 mothers in 38 developing countries during 1970-2000. The research makes two main contributions. It presents the first estimates of the intergenerational transmission of health across countries and over time; and it investigates the extent to which policy-amenable conditions around the birth of a child alter the intergenerational transmission of health.
Our study uses height, body mass index (BMI), and anaemia status to indicate the health of the mother. There is considerable previous evidence that height is an indicator of health and that adult height is sensitive to nutritional conditions in early childhood so, by using the height of the mother, we are able to study the relevance of conditions in her childhood for her offspring. Child health is indicated by mortality risk (risk of death) and anthropometric failure (low birth weight and stunted growth).
We find that mothers who are short, overweight or anaemic are more likely to produce children of low birth weight, who suffer early life mortality, and have stunted growth. A one standard deviation in mother´s height for instance raises the risk of poor child health by between 5% and 10%, depending on the outcome. A positive intergenerational correlation of health is widespread, evident in between 21 and 29 of the 38 countries in the sample, depending on the measure.
We find that children of less healthy mothers are more sensitive to changes in the socio-economic environment, which suggests that the intergenerational transmission of health involves not only genomic but also non-genomic mechanisms.
Getting better or worse?
Given that we find that intergenerational transmission is strongest among the least healthy mothers, we might expect intergenerational health mobility to fall with trend improvement in health. This will be reinforced by public health programmes that effectively target children at greatest risk.
Investigating trends for the 31 cohorts in 1970-2000, we find that, each decade, the correlation between child mortality and mother´s height falls by between 20% and 30%. This looks fairly impressive, especially in view of the record of limited income and education mobility. But breaking down the numbers by continent reveals that Latin America is alone in showing a consistent improvement in health mobility, with rates of decline twice as large as the average rates. Asia shows no significant trend and Africa shows a worsening trend, especially for mortality in the first month of life. So, in the last 30 years, while much economic progress has been made, there has been no amelioration of the tendency for children born to relatively unhealthy mothers in relatively poor regions in Africa and Asia to start life in poor health.
What can be done about it?
The natural question for public policy is how to narrow inequalities in child health at birth, or limit the persistence of health disadvantage within families so that children start life with more equal opportunities.
We find that each of economic growth, public health provision and women’s education act to weaken the tie between mother and child health. Economic growth and education are most effective in producing gains for children of mothers in relatively poor health, by virtue of which they lead to an improvement in the distribution of health. Turned around, these results imply that children are more likely to bear the scars of poor maternal health if they are conceived or born in adverse socio-economic conditions, such as recessions or period of civil strife associated with lower incomes and disruption to public health services. This is all the more important in view of the evidence that early childhood health predicts life expectancy, productivity and living standards in adulthood.
- Bhalotra, S. and S. Rawlings (2013). Gradients of the Intergenenerational Transmission of Health in Developing Countries. Review of Economics and Statistics, forthcoming (either May or July).
- Bhalotra, S. and S Rawlings. Intergenenerational Persistence in Health in Developing Countries: the Penalty of Gender Inequality? (2011), Journal of Public Economics, April, 95 (3-4), 286-299.