The WDR is based on a measure of ‘excess mortality,’ with methodology drawn from this 2010 paper by Anderson and Ray (which is, itself, inspired by Amartya Sen’s work on missing women from more than thirty years ago).This indicator assumes that men should die faster than women at a rate set by the developed world, where female life expectancy is nearly 7 years higher than for males. Any deviation from this, at any age, is evidence that women are ‘missing’.He concludes saying that addressing women's rights and missing girls is important. There is no doubt, but ignoring the impact that is on both genders can be problematic. A better statement would be there there are a given number of children who 'go missing' each year in developing countries.
Take Nigeria. Of every 1000 boys that are born, 146 die before their fifth birthday. The figures are slightly lower for girls, but still eye-poppingly high: 135 deaths. In developed countries, only 8 boys and 7 girls out of every 1000 die before the age of 5.
For the Bank, though, this is evidence not of excess mortality among children, but of excess mortality among girls. Seven fewer girls (or eight more boys) would need to die in Nigeria for the country to provide girls with the ‘correct’ health advantage. The same pattern holds across Africa and India: more boys die than girls, but the gap is not as wide as the Bank thinks it should be.
Since determining these measures and numbers is not my strength, I hope that the boys at the World Bank Development Impact blog will take to looking at Steven's concerns. There may be a good reason why the World Bank came up with the 4 million number.