For any set of data indicating women have a problem, there are conservatives claiming the data doesn’t exist or should be ignored. Regarding the Save the Children report (pdf link) (see my post below this one for info), Hindrocket of the conservative blog Power Line writes:
Hindrocket says “I always like to check the actual data,” but he doesn’t actually provide any data showing that the US does any better in world rankings if we measure “on a per-baby basis.” The World Health Organization’s website includes 1995 data (in Word format) for both “Lifetime Maternal Mortality Risk” (the measure Save the Children used) and “Maternal Mortality Ratio” (risk assessed “on a per-baby basis,” as Hindrocket prefers).
So is Hindrocket right? Nope. By the WHO’s measures (which include more countries than Save the Children’s – I think Save the Children was right to exclude some of those countries, as the data-gathering accuracy is questionable), the U.S. is 23rd in the world for “maternal mortality ratio,” and 26th in the world for “lifetime maternal mortality risk.” That is, the U.S. actually does just about the same regardless of which method is used. And according to the WHO’s numbers, even measured the way Hindrocket prefers, Sweden (#10) and Switzerland (#12) both are doing better than the United States.
(Note: In my first version of this post, I made some errors with the numbers in the above paragraph, which I’ve now corrected).
So Hindrocket – and his cheerleader Instapundit, who in a classic example of conservative “deny the problem” thinking writes “Save the children — from bogus claims!” – are both wrong on the facts. Whichever way you measure maternal mortality, the US isn’t doing well compared to countries like Sweden and Switzerland.
Hindrocket is also wrong in theory. Save the Children used “lifetime maternal mortality risk” not because they’re conspiring to make the US look bad (conservative paranoia is endless, isn’t it?), but because that is a standard figure used internationally, by the World Health Organization and many other groups.
Hindrocket says “assessing the risk on a per-baby basis… would be logical,” but he doesn’t support this bald statement with any reasoning. I think Hindrocket is mistaken; Save the Children’s experts made the correct and logical choice.
The purpose of Save the Children’s report is to measure factors that contribute to mother’s and children’s well-being. One important factor is whether a country’s health care system provides reproductive control for mothers, through education, access to birth control, and (although Save the Children diplomatically avoids discussing this element) access to abortion. In countries in which the health care system doesn’t provide women the ability to control when and how many children they have, mothers have birth earlier and more often, and both mothers and infants are more likely to die in the process.
Measuring only “per birth” maternal mortality, as Hindrocket suggests, is illogical because it can grossly underestimate the risk mothers face due to childbirth, and because it fails to measure the additional risk put on mothers when they don’t have access to reproductive control. In relatively low-birthrate countries (such as Sweden and, for that matter, the U.S.), this may not make much of a difference. But in high-birthrate countries, a lifetime risk assessment provides a much more realistic idea of how much risk mothers actually face.
Update: Maybe a model will make things clearer. (Models are oversimplified compared to reality, but they make some relationships clearer).
Imagine two countries, Access and Noaccess. Access encourages family planning with education and easily-available birth control. Noaccess is run by fundamentalists who have outlawed all forms of family planning. As a result, the average mother in Access has 1.2 children, whereas the average mother in Noaccess has 12 children.
As it happens, both countries have equally good “per birth” maternal mortality rates. But since Noaccess mothers give birth 10 times as often, they are ten times more likely to die in childbirth.
Under this model, which measure of maternal mortality provides a better idea of how well off mothers are? According to Hindrocket’s preferred “per birth” statistics, mothers are equally well-off in both countries. There is no reason, if we measure as Hindrocket wants us to, to say that Noaccess is doing any worse than Access.
In contrast, measuring the incidence of maternal mortality over a lifetime, as Save the Children does, correctly shows that mothers in Noaccess are ten times more likely to die in childbirth.
Okay, now let’s consider this logically.
- The purpose of Save the Children’s report is to measure differences between countries.
- Hindrocket’s measure utterly ignores an essential difference between countries like “Access” and “Noaccess.” “Lifetime maternal mortality risk” measures the same difference accurately.
- When measuring differences between countries, a statistic that measures differences is superior to one that ignores differences.
- Therefore, Save the Children is correct to measure using “lifetime maternal mortality risk,” rather than using Hindrocket’s method.