45% of births in the US are now induced, but evidence supporting this intervention is severely lacking. Case controlled studies show no benefit to inducing for postdates. Routinely inducing for postdates is based on 3 uncontrolled, retrospective studies showing 1/1000 less stillbirths while ignoring possible increases in brain damage to newborns as a result of induction.
It's hard to compete with 20 billion years of evolutionary selection, but the current medical management of the birth of the fetus and the placenta attempts to do just that, albeit rather unsuccessfully. For eons, all animals including humans passed on genes and habits that ensured delivering a live healthy newborn without bleeding excessively or dying of postpartum hemorrhage at birth.
In 2009, while 99.3% of US women delivered in hospital, 0.7% delivered at home. In response to this slight rise in homebirths, The American Journal of Obstetricians and Gynecologists issued a warning to all doctors and midwives to refuse to attend homebirth under all circumstances. In the absence of respected medical research showing planned homebirth to be unsafe, their recommendation is based on a single maternal death reported in the Daily Mail.