Much as I hate doing a response writeup, the argument that Jennifer presents here is fairly misleading. That's not really surprising, because most of the research on the safety of hospital births as opposed to home deliveries is fairly poorly publicized. In many ways, for a low risk pregnancy, a hospital delivery can actually be at least as dangerous, both for the baby and the mother, as a home birth attended by a competent midwife. This holds true in both the Third World and the First, though for fairly different reasons. Now, for high risk pregnancies, on the other hand, say breech deliveries or twins, hospital care is invaluable, and I think you'd be hard-pressed to find anybody to dispute that. The issue here is that what's good for a certain subset of women and babies is actually a poor choice for many others.
In America, the rate of babies who die in home deliveries attended by direct-entry midwives is about two out of every hundred (n.b. this is different from the infant mortality rate, which is calculated per live births; this figure includes stillbirths). This is exactly the same rate as in hospitals, and appears to be about as low as it can possibly go. The rate of epidurals at home deliveries: 0%, of course. The rate of mothers transported to hospitals for Caesarean Sections: about 5%, as opposed to normally 15%, and as high as 30 to 40 at some hospitals. Nobody knows what effect the anaesthetics used on mothers have on developing children, but we do know that they pass directly into the baby's bloodstream through the umbilicus.
It's true that the rate of infant mortality decreased drastically around 1900, which was also the time that doctors began handling childbirths. But, as we all know, correlation does not imply causality. Actually, the two single developments which are probably most responsible for this are better nutrition (which drastically lowers the rate of hemorrage and sepsis, more on this in a bit), and the advent of the germ theory of disease, which was adopted both by doctors and midwives, and led to regular hand-washing, and better sterility on the part of all.
Now, in the Third World, on the other hand, the rate of infant mortality remains very, very high. Somalia's a bit of a special case, seeing as it has no functioning government and relatively little contact with the outside world these days, but the statistics that Jennifer quoted are true, or close to true, for a number of other countries as well. Again, the single biggest culprit in this is malnutrition. When the mother's not properly nourished, she becomes exponentially more vulnerable to hemmorage, infection, and a variety of other conditions, and the health of her baby is endangered as well. Now, it's safe to say that transport to hospitals will have no effect on this malnutrition - it may even make the conditions worse, because many hospitals forbid women from eating during the labor period, and feed them sugar water through an IV instead.
One particular problem with hospitals in the third world is that sanitary conditions are often atrocious, and this drastically increases the risk of infection, both for the mother and the child. So much so that many NGOs and other foreign-aid groups are actually finding it more cost-effective and life-saving to provide basic medical training (on sanitation, and how to deliver CPR) to local traditional midwives, than to spend the money on more hospitals.
Home birth isn't the best answer for every mother, but neither is delivery in a hospital. There's a mounting body of evidence that for a fairly large population, home birth really is the best answer.