Let the Little Boys Die – Reaction to the 2012 World Development Report

The 2012 World Development Report has a stat that the World Bank is mighty proud of. I’ll let Bank President, Robert Zoellick tell the story:

Imagine if a city of almost four million people disappeared every year. A Los Angeles, Johannesburg, Yokohama. It would be hard to miss.

 Yet it goes largely unnoticed that almost four million girls and women “go missing” each year in developing countries.

It’s a shocking statistic. For comparison, AIDS and TB each kill around 1.7 million people a year – malaria a million. So why are so many women missing? What’s happening to them? And what does the Bank want to do about it?

Burrow into the report and the total drops a bit – to 3.882 million. A third of the ‘missing’ are from China, 30% from Sub-Saharan Africa, and 22% from India. The two big rising powers and the countries of the world’s poorest region clearly have some questions to answer.

The initial analysis follows a well-trodden path. According to the Bank, the largest group, 37%, are ‘missing at birth’. This is largely a problem for China and India (95% of missing baby girls). Many parents in these countries want sons rather than daughters, and are prepared to use ultrasound and abortion to make sure they get them.

It’s when we move onto infant mortality that the WDR gets into trouble. 617,000 of the missing (16% of the total) are girls who die before the age of 5, it reports. These girls die in much larger numbers than their brothers because they are neglected by their parents and are starved of healthcare by the prejudiced societies into which they have the misfortune to be born. Right?

Well no, not at all, as it happens.

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’.

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.

So why are too many African and Indian girls dying?

  • Are parents discriminating against girls? No – “there is little association between excess mortality among girls and disadvantages in vaccination, differential use of medical care, or differences in childhood nutrition as measured through the heights and weights of boys and girls.” In many countries, more boys than girls are “stunted, wasted, or underweight.”
  • What about doctors? Presumably, they simply can’t be bothered to treat females properly. Again, no. Researchers watched 30,000 medical consultations in Afghanistan, Burkina Faso, India, Mozambique, Paraguay, Rwanda, and Uganda and failed to find any discrimination at all. “In all seven countries, doctors spent the same time, asked the same questions, and completed the same number of examinations regardless of the sex of the patient.”

So what’s going on? The answer is slightly hard to follow: girls are significantly less likely to die of infectious diseases than boys, but they are even less likely to die of perinatal conditions (in or just after childbirth). So while all children benefit as infectious diseases are tackled, girls benefit slightly more than boys.

In other words, there’s no discrimination in play and this section of the report is based on a statistical artefact. Or a red herring. Or a tendentious attempt to beef up a press release (and if the result implies the health of poor girls matters more than that of poor boys – well, hey ho).

The WDR gets back onto much firmer ground when it moves on to adult women. Again, too, there is no evidence that discrimination in the health system is having a significant impact. The Bank cites Brazil, the Dominican Republic, Paraguay, Peru, South Africa, Egypt, Ghana, India, Indonesia, and Kenya as examples of countries where women get more healthcare than men.

But women in developing countries suffer horrendous levels of maternal mortality, due to large family sizes, early pregnancies, and poor, or missing, health care. Women are forty-three times more likely to die in pregnancy or childbirth in Africa than in the developed world. This is clearly not a problem men suffer from. Women also account for 52% of people living with HIV worldwide, and 60% of infections in Africa (where people are most likely to succumb to AIDS).

According to the Bank, these two factors are mainly responsible for another 1.3m missing women of child-bearing age and another 334,000 missing women in their fifties.

But, buried away in the report, we find that there are excess male deaths too – countries where men die even faster relative to women than in developed countries. Apparently there are 1 million ‘missing men’ – although we don’t get much of a breakdown of this figure. The Bank gives these deaths short shrift, putting them down to men’s need to take too many risks and an excess of “manliness” in contemporary societies. Don’t worry, in other words, these deaths can be discounted because they are men’s fault.

So, all in all, what does the “4 million women go missing every year” figure add up to?

First, families – mothers included, presumably – prefer sons over daughters in some parts of India and China. It’s a big problem with no obvious solution (the Bank suggests a crackdown on abortions and the re-education of families). And one that has already been intensively researched. The Bank’s added value is zero.

Second, children die far too often in poor countries. Boys and girls. Whatever the Bank insinuates, there’s no ‘correct’ gender balance when it comes to children’s births. How can anyone learn that an African boy is 10% more likely to die before the age of 5 than an African girl, and come away thinking – “if there was any justice in the world, that gap would be wider”? Especially when no evidence is presented that discrimination is a factor.

Third, women are more likely to die in childbirth (and of AIDS in some countries), and men take too many risks, especially when they’re young. Well, no shit Sherlock. Maternal health is one of eight MDGs for a reason, and the 1990 WDR did a much better job of banging the drum for ‘safe motherhood’ than its 2012 successor. This report doesn’t take the debate on women’s health forward an inch – and its discussion of men’s health is so thin as to be risible.

My conclusion: the 4 million figure is an advocacy stat of the worst kind. Lazy in its execution. And borderline dishonest in its presentation – especially for those who read the op-ed, and fail to find the detail buried in the report.

And it’s not even  an original approach. Amartya Sen said that 100 million women were missing in 1990. Stephen Klasen calculated a similar figure a decade ago. The Economist led with gendercide in 2010 – and did a better job of explaining the issue in a thousand words than the Bank does in 450 pages.

Securing women’s rights continues to be a massive challenge (and one with the potential to bring huge economic and social benefits). But the Bank does nothing for the cause. The analytical core of its report is rotten and its need to impress the media seems to blame.

The WDR website may be cock-a-hoop that it garnered “156 news stories published by lead print news outlets across the world, in just a week.” But I don’t think a blitz of favourable media for the Bank is what the WDR should be all about…